Saturday, June 25, 2016

Complexity Theory and the Study of History: Overproduction of Elites, Food Shortages, and Power Laws

I have written quite a bit about cliodynamics, the scientific approach to studying history, in the past. Peter Turchin has done quite a bit of work in this area, writing most recently about the inevitable collapse of the European Union.

Turchin's overall thesis is that an overproduction of elites results in inequalities that lead to the collapse of empires. And, as Egypt suggests, not just empires. However, one of the leading complexity theorists, Yaneer Bar-Yam, has developed a complexity model that suggests societal collapse comes about due to food prices increasing. More, network theory predicts power law distributions of revolutions/political collapse.

These are not necessarily conflicting views. Certainly the argument that there is a power law distribution of historical events like revolutions could easily complement the models of Turchin's and Bar-Yam's. It is also not impossible that the same dynamics that result in food prices increasing could result in the overproduction of elites and inequalities to increase.

Bar-Yam's thesis is particularly interesting, though one does wish he understood economics a bit better, especially around "speculation," which actually helps to smooth out prices and make them less volatile. At the same time he's right about the complete idiocy behind turning food directly into fuel, which unnecessarily drives up food prices, especially around anything involving corn (not just for direct consumption, but as feed for chickens, pigs, and cows, driving up meat prices). And he leaves out various farm subsidies, which politicians argue keeps prices down, but which in fact drive prices up (when New Zealand eliminated its subsidies for sheep farmers, wool and mutton prices dropped even as farmer incomes increased).

It would be interesting to map the food price-driven patterns of unrest with Turchin's patterns to see what overlaps there are and to see if one could find similar causes resulting in these two effects. I also don't know that Turchin has looked for power law distributions, but he certainly should. I would certainly be surprised if Bar-Yam's version didn't result in power law distributions. 

No comments: