Re: poly: World Economic History

From: Robin Hanson <>
Date: Fri Jun 05 1998 - 09:30:29 PDT

[I screwed up one number in my post on this topic. This
 post fixes that. Futher revisions will appear at: RH]

I got hold of a data series estimating total World Product (WP)
from 1 million B.C. to present. While a similar series on
population history plausibly fits a hyperpolic model, I find
that the WP history looks more plausibly like a sum of
exponentials. I interpret this as saying that the world
economy has switched between exponential growth "modes".

I get a close fit with these 5 exponential terms:

Doubling Time Dominant Period DT factor WP factor
------------- ------------------------- --------- ---------
500,000 yrs 1M B.C. to ~300K B.C. ? 9
140,000 yrs. ~300K B.C. to 5000 B.C. 4 6
    860 yrs. 5000 B.C. to 1700 155 196
     58 yrs. 1700 to 1900 15 11
     15 yrs. 1900 to 2000 4 37

It's not clear the first two periods are really different
(the data is bad then). If you merge them you get a doubling
time there of ~190K years, and a WP factor of 51, with the rest
of the model basically unchanged.

For each term, I've listed the doubling time and period when it
dominated. I also list the factor by which the growth rate
increased from the previous period, and the total WP growth factor
for each period.

The big questions to ask are: how soon might the world economy jump
again to a faster growth mode, and how much faster might that be?

If the current mode were to last through a WP factor within the
range of the previous modes, it would have lasted from 70 to 150
years. Since it has actually lasted 100 years, we could have
another 50 years to go or we could be overdue. The new doubling
time might be as slow as four years or as fast as one month.

These conclusions seem remarkable to me.

Compare the model and data in this graph:
See the whole analysis in this spreadsheet:

Robin Hanson
RWJF Health Policy Scholar, Sch. of Public Health 510-643-1884
140 Warren Hall, UC Berkeley, CA 94720-7360 FAX: 510-643-2627
Received on Fri Jun 5 16:36:30 1998

This archive was generated by hypermail 2.1.8 : Tue Mar 07 2006 - 14:45:30 PST