Order / planning etc (old 'about' page)

Just sticking this here for the record as I'm gonna update the 'about' page.

In the early 20th century, Ludwig Von Mises started off the 'socialist economic calculation debate'. He claimed that a planned economy was not only undesirable, but logically impossible. Friedrich Hayek - another Austrian economist - took up the argument and ran with it.

Hayek argued that society was just too complex to plan. Human minds - smaller, less complex systems - could never grasp the intricacies of the whole. For Hayek, this meant that all planning was the road to serfdom.

In the past ten to fifteen years, many people working on computation and society have found Hayek's writing prophetic. Whether or not you agree with his politics, Hayek seemed to have an understanding of society as an evolving process, and of people as bounded in a 'sensory order' from which they must get all their information. He was, perhaps, an agent-based thinker - one who could see the limitations of economic theory at the time, but lacked the tools we have now to investigate these ideas robustly.

An absolutely terrifying amount of work is being done on agent-based modelling, artificial/synthetic societies and the agent-based and cognitive foundations of economics. But that old 20th century question - what, actually, is the possible scope of planning? - might be a good way to focus all this. Is society really too complex to (for example) carefully plan for UK housing, or plan for a food market that protects the most vulnerable while defending local markets from being undercut by state action? Or was Hayek right - any attempt must be anti-market, and anything anti-market must be oppressive? And, anyway, don't we already plan? And, also, can modelling really answer this kind of question?

Or there's a totally different way to look at it. Stephen Lansing, in his book Priests and Programmers, tells the story of the Balinese water temple system: thousands of small temples co-ordinate water flow, in an overlapping patch-work of local choices, from volcano crater to sea. With no central oversight, a fine balance between pest control and rice-growing is struck. But then top-down Green Revolution methods came along and broke it - yields dropped as pests partied.

Lansing worked with ecological modeller James Kremer. The model they built made visible something that had long been invisible to policymakers: just how the system coordinated water flow and fallow periods. They helped reverse major institutions' views of the traditional system, which has since been reinstated.

The Balinese water temple system seems closer to Hayek's view of the market as a 'spontaneous order' than it does to a planning regime. Both are effective because they are decentralised; both can be broken by top-down imposition. So the really important question might be: can you plan for spontaneous order? Can it be grown - to meet a specific goal - like a vine on a trellis? When should we rely on this kind of decentralised system, and when might good ol' central command be better? And should we stop fetishising the price system, since Bali shows there are other ways - with less working parts that can go wrong - to make a productive distributed system?