4 ms·

This is an interesting argument but I don't know any proof that it is correct. It seems very vulnerable to "adding up" or aggregation problems, where local opt

This is an interesting argument but I don't know any proof that it is correct. It seems very vulnerable to "adding up" or aggregation problems, where local optimization clearly can work, but when you have lots of independent local optimization, the aggregate effect is bad. Have the Austrians proved that large number of actors each doing the best they will reliably aggregate up to global optimality? (That would actually be darned impressive given the well known problems of getting stuck on local optima.) If not then there's no clear argument against government action to prevent or deal with bad local optima. And without any strong account of how and when markets fail, the scope of government action needed is potentially broad. By the way, Shrekli is a pretty good example of "profit as a reward for helping everyone."

devereaux 11y ago
It is indeed very interesting, however it has not been studied enough. Nothing has been proved. Too much effort goes in mainstream economics, while complex systems and agent based models are where the good stuff is, to answer question such as yours. Aggregation problems are at the very core of Hayek reflection. I would not say it's bad. Look at his excellent 1945 paper: http://www.econlib.org/library/Essays/hykKnw1.html http://www.econlib.org/library/Essays/hykKnw1.html There seems to be a lot going in favor. The global optimality depends on what one think is best. Like, maybe Betamax was really better than VHS, but we humans settle on VHS. They went away with DVD. Then, rinse and repeat with Blueray. Maybe Betamax and HD-DVD were much better, and maybe in an alternate reality they won. But saying they should have won and imposing that by force would be unfair to all those who chose otherwise in this reality. And if we argue on a coordination problem, I would disagree- all it takes is something much better, like DVDs were to the VHS, or as streaming is to Blueray. So being "stuck" on a local optima depends on a) the existence of an alternative only slightly better and b) lack of progress, ie never ever having a clearly better alternative to overcome the smallness of the difference or c) disregarding the opinions or the costs to whoever prefers the alternative. We may not have proofs, but there is a strong argument. But maybe I am biased, as I would consider myself an Austrian economist. Personally, I think the best way is leaving everyone free, not forcing anyone to do what they don't want to do, as I don't like the concept of force or violence. I don't care about government action as such - if it was optional, as in a "recommendation". As in saying cigarettes are bad. Not in banning or taxing them.
AnimalMuppet 11y ago
> If not then there's no clear argument against government action to prevent or deal with bad local optima. There may not be an economic argument, but there's a political one. In theory, government can move you from a local maximum to a global one. In practice, though, government moves you to a maximum likelihood of [them] getting re-elected, not to a maximum of any economic utility function. More bluntly: You worry that the free market is going to produce a sub-optimal solution? Okay; that's a valid concern. But you think Congress is going to make things better? Have you looked at Congress lately? (And, yes, of course I know that the Ayn Rand unfettered free market is not the answer. We really do need regulation. I get that. But be wary of the assumption that government can do better than the market at optimizing aggregate results. Government is a very blunt hammer, not a fine adjustment.)

devereaux 11y ago

It is indeed very interesting, however it has not been studied enough. Nothing has been proved. Too much effort goes in mainstream economics, while complex systems and agent based models are where the good stuff is, to answer question such as yours. Aggregation problems are at the very core of Hayek reflection. I would not say it's bad. Look at his excellent 1945 paper: http://www.econlib.org/library/Essays/hykKnw1.html http://www.econlib.org/library/Essays/hykKnw1.html There seems to be a lot going in favor. The global optimality depends on what one think is best. Like, maybe Betamax was really better than VHS, but we humans settle on VHS. They went away with DVD. Then, rinse and repeat with Blueray. Maybe Betamax and HD-DVD were much better, and maybe in an alternate reality they won. But saying they should have won and imposing that by force would be unfair to all those who chose otherwise in this reality. And if we argue on a coordination problem, I would disagree- all it takes is something much better, like DVDs were to the VHS, or as streaming is to Blueray. So being "stuck" on a local optima depends on a) the existence of an alternative only slightly better and b) lack of progress, ie never ever having a clearly better alternative to overcome the smallness of the difference or c) disregarding the opinions or the costs to whoever prefers the alternative. We may not have proofs, but there is a strong argument. But maybe I am biased, as I would consider myself an Austrian economist. Personally, I think the best way is leaving everyone free, not forcing anyone to do what they don't want to do, as I don't like the concept of force or violence. I don't care about government action as such - if it was optional, as in a "recommendation". As in saying cigarettes are bad. Not in banning or taxing them.

AnimalMuppet 11y ago

> If not then there's no clear argument against government action to prevent or deal with bad local optima. There may not be an economic argument, but there's a political one. In theory, government can move you from a local maximum to a global one. In practice, though, government moves you to a maximum likelihood of [them] getting re-elected, not to a maximum of any economic utility function. More bluntly: You worry that the free market is going to produce a sub-optimal solution? Okay; that's a valid concern. But you think Congress is going to make things better? Have you looked at Congress lately? (And, yes, of course I know that the Ayn Rand unfettered free market is not the answer. We really do need regulation. I get that. But be wary of the assumption that government can do better than the market at optimizing aggregate results. Government is a very blunt hammer, not a fine adjustment.)