🚀 go-pugleaf

>>     pcollac@pyrnova.mis.pyramid.com (Paul Collacchi) writes:
>>|>   thompson@atlas.socsci.umn.edu (T. Scott Thompson) writes:
>
>>|> The most successful forecasting models do not attempt
>>|> to model causal mechanisms directly, and therefore are rather
>>|> uninformative about what the "fundamental forces" are driving the
>>|> business cycle.
>>|> --
>
>>I assume then that there are "analytic" models as well -- one's in which
>>tons of little micro-economic beings are "added up" to form a macro-economic
>>whole.  You say these don't do too well?
>
>Yes.  These models are known as "micro-simulation" models.  They have
>a pretty good track record at predicting relatively stable things like
>demographic change, at least over horizons of a few years.  They are
>notoriously bad at predicting aggregate economic variables like GNP
>over any horizon.
>
>[stuff deleted...]
>
>we _always_ rescaled the output of these models to be in conformance
>with the macroeconomic aggregate projections produced by our macro
>forecasting division.... partly because ...  we had no faith whatsoever 
>in the aggregate projections that came out of the microsimulation models.

In a nutshell then, micro-simulation models are appealing since they 
build models from collections of forces which are known to exist in
the economy, but they fall short at predicting, and the macroeconomic 
models are nicer at predicting but don't give any revealing
glimpses of the underlying forces at work in the economy.  Consequently,
the predictive models don't provide any suggestions regarding policy.
Is that fair?

I would think economists must be a little uneasy about this.  You know, I
think that Planck "fudged" some numbers to get the black body radiation
formula and Schroedinger kind of faked it with the wave equation, but in
both cases "backing into" an equation which described the data points did
contain some synthetic deduction based upon an understanding of actual
components that appeared in the system.

Am I right in understanding that a 200 point GDP curve doesn't nearly
present enough data points to make any sound correlations at all between
policy events and their consequences?  If so, then it might be the case
that to date political intervention into economic realms is virtually 
irrelevant: the oscillating mechanism is, as yet, unknown and untouched, 
and the data are and have always been consistent with the normal unfolding 
expression of a healthy and mysterious underlying system.

Can you imagine what it must be like to be a president?  We idolize them
when they ride the wave of a positive growth economy and say about them
that their actions were knowledgeable and "caused" the positive.  And we
heap shit upon them when they crash in a wave of negative growth and remind
them of their ignorance and "blame" them for the negative oscillation.

I certainly don't wish to offend anyone who may be in pain as a result of this
"recession", but right now I see many really positive, corrective economic
behaviors ocurring.

Paul Collacchi