Kuznets' tides: An archaeological perspective on the long-term dynamics of sustainable development

Abstract

Understanding the relationship between inequality and economic growth is a critical science problem that hinders sustainable development. In 1955, Simon Kuznets hypothesized that rising economic growth raises inequality, which levels off as that growth continues. Kuznets' "curve," which is a cornerstone of development economics, was based on data from a small sample of rich capitalist economies. Here, we draw on the GINI database, which includes area measurements of 53,464 residences from 1,176 settlements dating from 21,000 BC to the present, and published data from the Spatial Analysis in Maya Studies (SAMS) group, to radically reevaluate Kuznets' curve. We use Gini coefficients of residential disparity, a proxy of inequality, and mean residence area, a proxy of productivity, to investigate past sustainable development in the Bronze Age Interaction Zone (BAIZ), the Mundo Maya, and Britain prior to, over the course of, and after the Roman conquest. We interpolate spatial patterns across each zone to statistically evaluate changes in inequality and economic growth. We find a recurring pattern in which phases of sustainable development, a rise in productivity without a rise in inequality, gave way to increasing inequality. These patterns resemble those Branko Milanovic termed "Kuznets' waves," albeit at timescales better described as "tides," which began after the introduction of weight metrology, an early form of economic governance associated with long-distance exchange. We posit that past sustainable development was predicated on balancing reciprocity from the bottom-up with mechanisms like early weight metrology but was repeatedly forestalled as inchoate elites co-opted these mechanisms.

Keywords: archaeology; economic growth; inequality; productivity; sustainable development.

Fig. 1.

Workflow for comparing economic growth to inequality using changes in residence area, [ML_TAH], and residential disparity, [Gini], within zones of interaction as proxies.

Fig. 2.

A global visualization of the GINI database’s SiteGiniLevel table. The kernel density of site locations is weighted by [Gini] in those locations. Clusters marked above the visualization, and the mean [Gini] of clusters is compared to that of the total sample. The map was prepared using QGIS 3.32 and a basemap from naturalearth.com projected using the Eckert IV coordinate reference system.

Fig. 3.

A comparison of Δproductivity and Δinequality during different phases of Kuznets’ tide within each zone. A–C present spatial interpolations for Britain, D–F present spatial interpolations for the Mundo Maya, and G–I present spatial interpolations for the BAIZ. The maps were prepared using QGIS 3.32 and a basemap from naturalearth.com projected using the Eckert IV coordinate reference system.

Fig. 4.

Kuznets’ tides plotted by years before present and by years since the advent of weights and measures. Lines depict the correlation between Δproductivity and Δinequality across space in different timeslices from each zone. Plot A depicts the correlation in years before present and Plot B depicts the correlation in years elapsed since the emergence of weights.