Complex systems and Keynes approach

It is often said that Adam Smith (1723-1790) is the first economist to have proposed a systemic view of Economics. Indeed he clearly saw that a system is at work when economic agents produce, exchange and consume (and save, invest, etc.). He said that "It is not from the benevolence of the butcher, the brewer, or the baker, that we expect our dinner, but from their regard to their own interest". And he introduced the concept of "Invisible hand". These are definitely "complex dynamic systems" ideas.

In fact, Quesnay (1694-1774) already had some systemic understanding of Economics, despite the fact that he said that only farmers produce wealth. Quesnay made the first deliberate attempt at constructing macroeconomic measures of the economy of a country, and at modelling their evolutions (Quesnay zig-zag design). One can even distinguish some systemic thinking in the reasoning of the mercantilists (for instance Colbert (1619-1683)) who advocated launching state manufactures to collect gold and finance the state.

Yet Smith, like all his contemporaries, had a poor intuition about the workings of these systems. He thought that the economic system converges toward stability and that the stable state is optimal. Trying to prove this under various sets of hypotheses has kept economists busy for two centuries. (But the most elaborate models don't use money !)

One of the culprit hypotheses that leads to the mistaken feeling that economic systems ought to converge toward stability and optimality is the hypothesis of decreasing efficiency of agent's resources. Everyone who has worked a bit in the real world of firms and competition knows that this hypothesis is wrong. (The hypothesis seemed natural to XVIIIth century economists because they were still impressed by the ideas of the Physiocrats who said that only land produces wealth, and the law is realistic when working land with the techniques of that century.) Let's take the simple model of a little community with various agents and in particular two bakers. If for some reasons one of the bakers is pushed out of work, his efficiency to start over again producing bread is not higher than the baker who remained in business but much lower. So he will have a tough time reentering the trade. And conversely the marginal efficiency of the remaining baker is very high, not very low. So the remaining baker will easily take over the business of the pushed out baker. What is true is that the pushed out baker can either become an employee of the remaining baker, and for Marxist economists this is an illustration of relationships of strength and exploitation of man by man, or he can do something new. The naive supporters of liberal systems say that it is the obvious solution. But usually they are teachers in universities with tenured positions, and have no experience whatsoever of what it is to start over anew. And of course so long as the pushed out baker hasn't got back a new well paid activity (that is an activity which generates a substancial flow of money in and out of his accounts) it is the whole community that loses, because he can no longer purchase what he used to purchase. All the other agents who supplied him also see their own revenue decrease. These considerations only scratch the surface of complex dynamic systems.

Keynes (1883-1946) is the first economist to have clearly understood that Smith system does not converge to an optimal state. He stressed that the dire unemployement level of the Great depression was not the result of too high wages, and that lowering the wages even more would not solve the crisis. (One of my grand fathers, Amaury Lebon, who was a wealthy architect in the 20's, car, spacious house, etc., lost most of his work in the 30's, and became an employee of the reconstruction agency after the war.) To Keynes the system was endogenously producing the crisis, because of other imbalances. He searched for causes in the management of the money supply. He opposed Pigou's ideas about the "real balance effect" that should also lead back to an optimal stable economic state.

Keynes proposed a bold solution : the State should substitute for the failing industrial sectors in investing and henceforth start over again the economic machine. He explained that there would be a multiplicative effect, etc. (Note that any expenditure has a multiplicative effect, not only those made by the State.)  Pres. Roosevelt applied Keynes ideas in the 30's, for instance in the Tennessee Valley Authority project. Keynes ideas dominated western economies from 1930 until 1980.

Of course the idea of State intervention did not appear with Keynes ; there are many examples from the past. In fact before large market economies emerged after, say, 1600, a part of the activity of society members was organised by State authorities to construct public works, cathedrals, city walls, palaces, etc. But this was mostly achieved not through market economics.

The Mercantilists of the XVth and XVIth century had strong (and biased) views on state responsibility to bring wealth. But Keynes goes one step farther than the Mercantilists, since he proposes to intervene within the domestic economy, and not only set up exporting manufactures to import gold.

It is still debated whether it is the keynesian economics of the 30's or the more brutal war economics of 1939-1945 that helped relaunch the US economy. It has also been pointed out that Keynes policy is a step toward central planification, which has dramatically shown its inadaptation to human nature.

Whatever may be true, it is sure, from the viewpoint of complex systems, that Keynes recommandation does not take into account the subtle effects on the economic machine of changes in public policy. And that will have to be investigated with the new tools that will be constructed in the field of complex dynamic systems.

For instance when one hears on the radio that "the government is planning to finance more agencies to be in charge of helping unemployed people to find back a job" one knows that this ignores entirely the dynamics of complex systems, and without any ideological slant one can predict that it will be a failure.

Between "no government" that leads to suboptimal economic states, and "big government" that leads to economic disruptions what is the way of the future ? Most probably the same as for organisms : the organisation of local communities of various sizes with a large degree of autonomy, with some central "nervous system".

Another issue that Economics has not resolved is this : is it possible to create a satisfactory economic system without a positive selfless contribution from its members ? It's been four centuries that economists grope for proofs that, within a certain legal framework, i.e. rules of the game (which already imply the notion of "a community" and of "an ultimate power" to enforce the rules), it is possible - without success.

André Cabannes

29 November 2004

References :
1) Political economy, 1740 - 1860  by Stewart Crehan
2) Spontaneous order Wikipedia article (a wonderful example of community construction with a positive selfless contribution from its members).
3) For some thought provoking ideas : Energy economics
4) For a discussion of the Invisible hand from classical mathematics, as opposed to the mathematics of complex dynamic systems