ABSTRACT
In this chapter, I show how the investigation of racialization in datafied applications can be done through an instrumental, epistemological, and ontological approach to datafication and that the results of each approach do not necessarily match. By analyzing the attempted implementation of a tool aimed at measuring the composition of personnel in terms of migration background called Barometer Culturele Diversiteit (BCD) at Utrecht University in the Netherlands, I show how the tool is using ideas about race (instrumental), shaping knowledge through colonial politics (epistemological), and producing race (ontological) simultaneously. Aided by this analysis, I will advocate for an understanding of the use of race-ethnic data for affirmative purposes in terms of strategic essentialism, making epistemic imperfection regarding race warranted only in antiracist data systems working toward their own obsolescence.
