送交者: mangolasi 于 2005-4-11, 14:58:17:
回答: 好奇地问 由 boxer 于 2005-4-11, 14:44:01:
Thus I think it's a healthy direction in economics to emphasize methodology and empirical evidence. Real world data is dirty, but have to live with it.
But some theory are not only elegant, but also seamlessly logical, and unfortunately, very difficult to test and difficult to apply (because of strict assumptions). My favorite example is Arrow's impossibility theorem to prove there must be a dictator in the society (not the dictator in the usual sense)--UNDER SOME CONDITIONS.
One bad thing in economics, computability asides, is that theorectial people relies too heavily on assumptions, and economics theory are contrast against the real social phenomenons which don't take place in lab conditions, so some assumptions must be violated. And which assumptions to be chosen are not entirely free from ideologies. Like assuming everyone is rational may make sense to some while senseless to the others. But assuming surface free of friction make sense to all physics, and if it doesn't, they can make experiment to see whether they can reach agreement.
Very few experiments are possible in economics, though some people is doing that.
Babbling from an apprentice. Insiders don't laugh, please.
So my feeling is that the trade is a way of making a living. Though I can proudly announce that econometrics can be counted as hard science.