Making Data Count

An important task of the modern social scientist is to push back against the notion that techno­logy and technocratic solutions are inhe­rently politically neutral, a task made even more difficult in a world where technological capabilities increase at light speed. The Uncounted by Alex Cobham is a valuable book that must be read by all those who believe in this project, for it provides an important analysis of how existing biases and structures of power influence the raw material on which much of today’s technological int­erventions are built: that of data. Cobham’s work is more focused on the question of inequality, but his arguments hold relevance for all forms of socio-­economic data.

Cobham’s thesis is deceptively simple: data on inequality systematically leave out information on two sets of indivi­duals from large-scale socio-economic databases, those at the bottom of the distribution—whether seen in terms of income or possessing other characteri­stics deemed to lie at the bottom of a ­socially constructed hierarchy—and those at the very top.
I use the words “deceptively simple” because at first glance, these occurrences may be explained away simply due to the exigencies of sample construction—the very rich tend to understate their incomes and/or those with truly large levels of income are not surveyed, and it may be hard for surveyors to fully enumerate all those at the bottom of social pyramids due to ­logistical difficulties.