right. They don't predict very well at all. And they don't predict
better than competing models do. The whole "positive economics
predicts" business is like the bullfrog who puffs himself up to look
fiercer, to attract females and deter competing males. But I was
focussing on the basic problem with the whole story: the assumptions
themselves are poor predictors, but the model can be "saved" by adding
new poor assumptions.
On 1/29/07, Bill Lear <[EMAIL PROTECTED]> wrote:
And just how do these so-called models predict anything? What do they
predict? What is the accuracy? For how long into the future do
they predict things? How does accuracy tail off as time increases?
Friedman himself flamed out with outrageous predictions of rising
inflation in the '80s ("10 to 11 percent by the end of the year"
type predictions).
There is a confusion here between scientific, testable models, and a
gaggle of guesses dressed up as science.
Bill
On Monday, January 29, 2007 at 12:48:50 (-0800) Jim Devine writes:
>On the Methodology of The Methodology of Positive Economics
>By James Devine
>at http://maxspeak.org/devine.html
>on January 29, 2007.
--
Jim Devine / "Doubt is uncomfortable, but certainty is ridiculous." -- Voltaire.
