Abstract
In the Netherlands, data about the spatial distribution of Dutch citizens structured on the basis of their “migration background” is available in governmental open data sets. In this dissertation, such data are called “race-ethnic” for the combined racial and ethnonationalist connotations of the categorizations and labels. I will argue that
... read more
the appropriation of race-ethnic data for new purposes enables a process of racialization, understood as the selection of “particular human features for purposes of racial signification” (Omi and Winant 2015, 110). To understand how processes of racialization are not only historically and socially determined, but also result from specific technologies, I propose the concept of technologically mediated racialization (TMR). I use TMR in the investigation of the development of past technologies used for counting and clustering people in the Dutch census between 1899 and 2011. I show how race-ethnic categorization is intimately tied to sociocultural ideas about which human characteristics matter, as well as to the available technologies to record, store, and distribute that information. I will trace the historical connection between contemporary Dutch categories and clusters, such as “person with a migration background”, and older categories and clusters, such as “niet-Westerse allochtoon”, and “vreemdeling”. This will allow me to show that contemporary categorization practices are still informed by Dutch colonial history and migration history. As a result, label changes are both unable to solve racializing practices, and, simultaneously, make historical connections between race-ethnic categories and colonial history harder to access. This process can be understood as a form of colonial aphasia: “a loss of access that may verge on active dissociation” (Stoler 2016, 12). In my first two case studies, I show how race-ethnic categories are made available by the data infrastructure of Statistics Netherlands (CBS) and are used in data applications such as the Allochtoon-o-meter of the alt-right weblog Geenstijl, and the Leefbaarometer of the Dutch Ministry of the Interior. While these two systems seem very different, at least in terms of their politics, these data applications inherit the same embedded perspective and normative assumptions about Dutchness already present in the data infrastructure of CBS. Therefore, the results of these applications are no accidental instances of TMR but rather materializations of a technicity of race that was already embedded in both the available technologies and related race-ethnic categories. Furthermore, especially in the case of the Leefbaarometer, the colonial and racialized connotations of the provided information have become obscured as a result of the decontextualizing nature of the process of datafication, worsening the aforementioned colonial aphasia. My third case study, the Crime Anticipation System (CAS), a predictive policing system of the Dutch National Police, shows how even a system that does not use any explicit race-ethnic categories can produce TMR. In a data assemblage that includes both CAS, a system that aims to exploit information about societal inequalities to achieve operational “efficiency”, and an organizational culture that has only minimal protections against ethnic profiling, racialized and racializing results are no accident but an inherent feature of the system. In the conclusions, I show how the concept of TMR can be used in the planning, evaluation, and design processes of current and future governmental practices that are aided by data technologies. In this way, I hope that the results of this investigation are not only aiding in the production of knowledge within the scholarly fields of critical data studies and postcolonial studies, but also in guiding the Netherlands along a process towards a more accountable and just datafied society that values all people equally.
show less
