* Robert J. Chassell ([EMAIL PROTECTED]) wrote:
>   * What would be the current GDP and median per capta US at the
>     growth rate that Republican administrations achieved historically?
>     Presume they were the only administration in power since 1948 (or
>     whatever is the base year) and that they succeeded economically as
>     well as they did.
> 
>   * What would be the current GDP and median per capta US at the
>     growth rate that Democratic administrations achieved historically?
>     Presume they were the only administration in power since 1948 (or
>     whatever is the base year) and that they succeeded economically as
>     well as they did.

This sort of extrapolation is likely to be more dangerous than helpful,
because it generates a big number to give one a false sense of security
greater than justified from the data.

The goal of any such analysis is likely to be to formulate a future
policy. But the data can support any number of "explanations" which
could lead to dramatically different policies.

For example, the data could equally well "explain" that there is an
economic "sweet spot" that the US has historically averaged close to
over the past 70 years, but the average was slightly on the conservative
side (oscillating back and forth around a bias point slightly on the
conservative side of the sweet spot).  Therefore, whenever Democrats get
power they will move the US closer to the sweet spot, thus improving
economic growth, and the Republicans move the country away from the
sweet spot. But if the Democrats had been in power the whole 70 years,
we likely would have far overshot the sweet-spot on the liberal side and
thus had much slower growth.

The data cannot distinguish between that, and "the Democrats policies
will consistently result in stronger growth if applied continually". Or
any other of a large number of similar "explanations". So the conclusion
you are looking for from such an ambitious extrapolation will be likely
to lead to a false sense of security.


--
Erik Reuter   http://www.erikreuter.net/
_______________________________________________
http://www.mccmedia.com/mailman/listinfo/brin-l

Reply via email to