Acceso usuarios
Kinan Alajak, Merve Burnazoglu, Koen Leurs, Gerwin van Schie
This article critically investigates the unintended and perverse effects of digital public technologies on refugees’ fundamental rights and socio‐economic inclusion. As a case study, we examine the GeoMatch algorithm, a recommender system implemented by the Dutch government to automate employment search and matching processes for refugees across its 35 labour market regions. As data and methods, we used close reading techniques to analyse a set of disclosed documents obtained through Freedom of Information (FOI) requests, drawing on the practices of investigative journalism. Contrary to official claims of effectiveness, economic impact, and objectivity, our findings suggest that GeoMatch’s algorithmic system prioritises aggregate optimisation over individual opportunities, with a disproportionate risk of discrimination on the basis of ethnicity, gender, or marital status. The findings further indicate a diminished capacity for both refugees and reception officers to contest automated decisions, threatening refugees’ human dignity and self‐determination. We therefore argue that the deployment of GeoMatch should be reconsidered until these ethical concerns are adequately addressed. The article provides an empirical case supporting concerns raised in the literature on the role of algorithmic systems in social and economic stratification. Methodologically, our contribution endorses the emerging approach of combining FOI and close reading to study opaque technological systems and automated policy domains.
