Technology, institutions, and wealth inequality: What do we learn from comparable estimates over 12 millennia?

Comprehensive data on individual income or wealth became available only with modern tax- based polities. This has limited the quantitative comparative study of economic inequality across institutions – e.g. slavery, multi-ethnic empires, communal non-state societies, modern democ- racies – and technologies – e.g. collecting food rather than producing it, substituting non-human animal work for human labor. We expand the range of institutions and technologies under study by providing distributions of likely levels of wealth inequality for 431 sites and dates over the past 12 millennia that we have made (insofar as possible) representative of the underlying population and comparable across differing asset types, population sizes, and ownership units. In this data set we find that both technology and institutions matter for wealth inequality, but not entirely in the ways that many would have thought: a) despite vast increases in scale (population) and wealth (and therefore the surplus over subsistence), there is effectively no trend in inequality over the past 6 millennia; b) inequality among hoe-based farmers is not greater than among sedentary hunter-gatherers; c) inequalities are greater among capital-intensive (plow-based, e.g.) farmers than hoe-based farmers; d) societies without states of any kind are less unequal than state-governed societies, a finding that not an artifact of differences in scale; e) economies based on enslaved labor are more unequal; f) early modern Europe’s †little divergence†in the growth of per capita output was preceded by an equalization of wealth in the north and west, apparently driven by a enduring regional divergence in institutions in response to temporary labor shortages in the aftermath of the 14th century plagues; and g) wealth inequality is not less in democracies than in non-democratic (and non-slave) states.