Post #321,642
2/17/10 8:32:58 PM
|
I read it a little differently.
It seemed to me that he was mainly arguing for more care in the use of language to describe, and thinking about, economic models.
He said:
I'd say that, in a standard general equilibrium model, informational efficiency does imply allocational efficiency. So to the extent that one accepts such models as guides, one should believe that, for practical purposes informational efficiency implies allocational efficiency.
The problem is that approximate informational efficiency does not imply approximate allocational efficiency.
IOW, in a model with perfect information, the allocation results do follow. However, in the real world we don't have perfect information, so we have to be more careful. Mathematics tells us that we cannot assume that a little noise in the information gives a little noise in the allocation results. It doesn't follow, no matter how much we might like it to. When talking about a "good approximation" it's important to be clear about what meaning of the term is being discussed.
It seems like a reasonable point to me.
FWIW.
Cheers,
Scott.
|
Post #321,644
2/17/10 9:09:22 PM
|
like I always said about climate models
climatology is a lot like economics, we dont know enough about the peripheral effects. We see carbon rising slightly behind temperature for good reason but the pieces we dont understand are huge. Models are built using imperfect information with noise that can be adjusted the wrong way or suppressed. One fellow thought in the next 30 years we are looking at 650ppm c02 in the atmosphere with intense effort to curtail it starting now with no decent prediction as to what 650ppm actually means to the climate and ultimately the weather
If we torture the data long enough, it will confess. (Ronald Coase, Nobel Prize for Economic Sciences, 1991)
|
Post #321,645
2/17/10 9:32:52 PM
|
The analogy is bad.
The Earth's climate is a complicated black box. We know much of the science. We know much of the physics; we don't know all the various pieces and how they interact with enough precision. But we're getting better.
Economics models are an attempt to explain large-scale human behavior in an economy. It's not physics - it's herd mentality. Fundamentally, economics attempts to apply mathematics to one of the social sciences. It's much less deterministic than a physical system.
Even though the climate models are still young, they're actually very good. And getting better all the time. What economic model from even 5 years ago would have predicted the Great Recession? What economic model can predict the state of the world economy in 30 years with any confidence? I'd have much more confidence in the prediction of the state of the climate from present models.
You can even play with some of them on your own PC or Mac (courtesy of your nemesis James Hansen's group ;-) - http://edgcm.columbia.edu/software2/
FWIW.
Cheers,
Scott.
|
Post #321,646
2/17/10 10:03:07 PM
|
What makes economics very different from physical sciences
Elements in the system (people) are aware of the "rules" of the system and consciously alter their behavior in response to what they believe the rules to be.
--
Drew
|
Post #321,650
2/17/10 10:37:31 PM
|
welcome to quantum physics
If we torture the data long enough, it will confess. (Ronald Coase, Nobel Prize for Economic Sciences, 1991)
|
Post #321,656
2/17/10 11:21:31 PM
|
That's not how quantum physics works.
There are strong, well-understood rules and equations that will give you the answer to a quantum mechanics problem. The answer includes a probability.
http://en.wikipedia....B6dinger_equation
As Drew said, Economics tries to apply mathematics to Calvinball. The players are always changing the rules.
Cheers,
Scott.
|
Post #321,658
2/17/10 11:25:44 PM
|
Re: That's not how quantum physics works.
yes it is, the answer depends on the observation. Is the cat dead or not?
If we torture the data long enough, it will confess. (Ronald Coase, Nobel Prize for Economic Sciences, 1991)
|
Post #321,659
2/17/10 11:37:31 PM
|
Economics is not falsifiable
--
Drew
|
Post #321,669
2/18/10 11:07:50 AM
|
What? ever hear of ponzi?
If we torture the data long enough, it will confess. (Ronald Coase, Nobel Prize for Economic Sciences, 1991)
|
Post #321,660
2/17/10 11:44:17 PM
|
It's both.
Observation collapses the wave function. It doesn't change the physics.
More: http://en.wikipedia....%B6dinger%27s_cat
Cheers,
Scott.
|
Post #321,648
2/17/10 10:35:39 PM
|
climatology isnt herd mentality? bwahahahaha
as for your link? you have to use their model no peaking at the source code. Microsoft for climate prediction anyone?
If we torture the data long enough, it will confess. (Ronald Coase, Nobel Prize for Economic Sciences, 1991)
|
Post #321,653
2/17/10 10:49:16 PM
|
So run it against history.
Treat it as a black box. Does it do a reasonable job with the historical data or not?
http://www.grida.no/...c_tar/wg1/308.htm
If you really want it, the source is public domain - http://edgcm.columbi...port2/faq/#faq003
HTH.
Cheers,
Scott.
|
Post #321,655
2/17/10 10:57:10 PM
|
thanx, downloaded will review
If we torture the data long enough, it will confess. (Ronald Coase, Nobel Prize for Economic Sciences, 1991)
|