The Politics of Inaccuracy: The Racial Contract and Algorithmic Risk-Scoring in Dutch Welfare Bureau

Higher Seminar in Media and Communcation Studies with Gerwin van Schie, Assistant Professor of New Media and Digital Culture at Utrecht University.

In the last decade, multiple new algorithmic systems in Dutch governance caused discriminatory and inaccurate work practices, often victimizing marginalized and racialized Dutch citizens. While attention has been paid to how the discriminatory logic of these algorithmic systems is related to Dutch cultural norms of whiteness, the inaccuracy of these systems is often either not taken into account, or seen as a separate problem. In this lecture, we will take a closer look at two welfare fraud risk-scoring algorithms (from the city of Rotterdam and the Dutch Tax and Customs Administration) of which the inaccuracy is either known by their creators or not tested at al. We will discuss how it is possible that, within a popular discourse that stresses the efficiency and accuracy of data and algorithms, inaccurate and untested systems are implemented in welfare infrastructures. By analysing these systems and their faults through the lens of racial contract theory (Mills 1997), I will argue that their inaccuracy is intricately linked with the racial logic of their design. By making the surveillance and prosecution of Dutch citizens focussed more on their identity than their behaviour, a situation is created in which relying on welfare is inherently risky for racialized citizens. The always present potential for prosecution, which I will refer to as “prosecutability”, functions as a form of biopolitics enforcing the Dutch racial contract. As a result, racialized people are deterred from applying for welfare, just for existing in the Netherlands.

It will be possible to attend the seminar both live on campus and remotely via Zoom. For more information, please e-mail miriana.cascone@sh.se