Andie:
>Just an anecdote, but I think a representative one.
[WS:] If we are telling anecdotes, here is another one. Back in the 1960s American astronauts found that their ball point pens did not work in space in a zero-gravity condition. To fix the problem, NASA spent a few million $ to invent one that does. The Russians faced the same problem, of course, but instead of pens, they used pencils.
>From another angle, the Russian AK-47 is a rather backward design comparing to the US models. But it has one advantage over its more advanced competitors - it is simple, easy to fix and reliable, and that is what made it a weapon of choice around the world.
The lesson from these anecdotes is that every innovation must be looked in the proper socio-historical context. The fact that we (as people) can make someting, does not mean that we should, at least under a particular set of circumstances.
Another aspect of innovation is path dependence - the basic argument here is that following a pre-set path often offer significant competitive advantages (transaction cost savings) over innovating. So from the rational actor's point of view following a path is more advantageous than trailblazing. Brian Arthur http://www.santafe.edu/~wbarthur/ developed some fancy-schmantsy statistical models demonstrating that, but it is really common sense, when you think about the nature of human social interaction. People imitate one another 99% of the time, and that is what increases thier chances of survivals.
Going back to central planning - it was a significant innovation at the turn of the century (before even Keynes became fashionable) to solve a particular socio-economic problem - quick development of a backward economy. And it worked really well, far much better than the markets would do under a similar set of conditions. But that, of course, does not mean that central planning would work well under a different set of circumstances (in fact we know it did not work in Africa, as African experiments with socialism cf. Tanzania were not very successful), let alone under *any* circumstances.
To push things a bit further, central planning was very good at mobilizing local resources for investment, while effectively controlling negative social and economic aspects of severely "overheated" economy (inflation, price gauging, profiteering from rent-seeking behavior, urban poverty, inadequate social safety net, etc.) . However, it was not a panacea for domestically driven development. Specifically, it did not work very well in situations where establishing centralized control was not socially feasible - as the case of African socialism seems to suggest.
If this interpretation is correct, it follows that central planning would not work very well to prevent global financial meltdowns for similar reasons - establishing global centralized control of is not feasible, and even if it were, it would not guarantee that what s good for the global system as a whole, would be good for its individual parts (I believe Krugman makes that point quite explicitly in _The Reurn of Depression Economics_). The bickering over stimulus policies in the EU is an empirical manifestation of that. But of course the unequal geographical distrubution of costs and benefits a big problem of global free markets as well.
So the bottom line is that we (as people) are now facing a problem for which nobody really has any convincing solutions. All previously tried approaches (markets, planning, Keynesianism) that used to work in the past failed this time, and nobody really has any good ideas what to do next - even Krugman (the conclusion of his _Depression Economics_ is disappointingly weak.)
I personally think that the idea of a self-regulating system is a bunch of crock - a social science version of the perpetuum mobile myth. Its main appeal is to government-hating contumacious libertarians and yuppies. At the same time, any effective planning system carries very significant (if not prohibitive) transaction cost, which grows exponentially with the complexity of the system. If central planning of medium-sized, relatively homogenous European countries was quite challenging, then doing the same with a global economy borders on impossibility. I do not think democracy would survive in such a system.
I can only conjecture that creation of a global parliamentary government responsible for setting global Keynesian policies and equipped with powers to enforce them may offer a solution (it worked in the US, which is pretty big and diverse) - but I do not think such a government is possible any time soon. At this point in time, the most likely "solution" is muddling through the crises with makeshift bandaid measures amidst speculations of the chattering class what makeshift bandaid measure is the solution.
Wojtek