Pred-Pol-Pov: Visibility, data flows and the predictive policing of poverty

Predictive and data-driven policing systems continue to proliferate around the world enticing police forces with promises of improvements in efficiency and the ability to offer various ways of addressing the future to pre-empt, predict, or prevent crime. As more of these systems become operationalised in England and Wales, this paper takes up Duarte's (2021) observation of a lack of description as to what such systems actually are. It adapts a social network methodology to explore what a data-driven policing system is. Using a police force in England, UK, as a case study, we provide a visualisation of a data-driven policing system based on the data flows it requires to operate. We show how a disparate network of affiliate organisations act as collators of specific data types which are then used in a range of policing applications. The paper makes visible how data travels from its source through various nodes and the different potential points of translation that occur. We uncover how certain groups are made visible to the police due to the data being shared within a network of institutions and organisations, and as others have argued before us, how the data points that make this network visibility possible are often proxies for poverty. This makes certain groups and sections of society highly visible to the digital system whilst other groups become less visible – and potentially even hidden.