Measuring Socio-Spatial Justice: From Statistics to Big Data – Promises and Threats

Résumé En

Whether to be praised or vilified, Big Data is often presented as the coming of a new statistical world, where causality and scientific method will be replaced by powerful statistical algorithms, able to infer probabilistic predictive models from correlations between the mass of our daily numeric tracks. In this paper, as an Economist, we discuss, within the framework of socio-spatial justice measurement, the normative pros and the cons of the 'Old World' of traditional, bottom-up and deductive statistics and those of the 'New World' of Big Data statistics, decentralised and inductive. In the end, Big Data is neither white nor black magic: a social construct, it is only another statistical tool that we can use for a better understanding of the world. As such, it can and should be scrutinised by social scientists. From the consequentialist point of view of Economics, for which measuring well-being is crucial to conceiving social justice, Big Data looks like a genuine improvement on the data scarcity from before… provided that we are able to organise, from a procedural point of view, its regulation.