Revisiting Hanson’s 'Long-Term Growth as a Sequence of Exponential Modes'

Abstract

In “Long-Term Growth as a Sequence of Exponential Modes,” Robin Hanson models economic growth over the history of the genus Homo. The paper shows that, across 2 million years, the data series for gross world product (GWP) can be modeled as a series of a few exponential growth episodes (“modes”). The paper's pre-ferred model uses the constant-elasticity-of-substitution (CES) function form to smooth the transitions be-tween modes. The fit implies that after the transition to the current mode is complete, GWP will double every 6.3 years. If a fresh mode commences, on historical par with the industrial revolution, it could reduce the doubling time to weeks or days. I reconstruct the paper, revise and extend the GWP series, and add new anal-ysis, principally by modelling GWP per capita in addition to GWP. I find that the CES model fitting only ap-pears to converge because of the limitations of finite-precision math. The tendency of the fitting to infinitely extend the transition from the agricultural and to the industrial mode shifts weight to the view that the global growth acceleration since the Black Death has been gradual.

Keywords: macroeconomic history, gross world product

JEL Classification: C13, E10, N00, O11, O40