PhD defence by Asbjørn William Ammitzbøll Flügge

Portrait of Asbjørn


Where the Practice is - a study of humans and machines in public employment services


Public services are increasingly digitized, and at the forefront of this transformation is the implementation of predictive algorithms into varied welfare domains. In this dissertation I investigate how predictive algorithms in public employment services for profiling new job seekers' risk of long-term unemployed, are used in municipal job centers. Researching within the traditions of Human-Computer Interaction (HCI) and Computer-Supported Cooperative Work (CSCW), this dissertation takes an ethnographic approach for studying how these technologies are being used in practice by frontline workers, caseworkers. In the dissertation, and the research articles written as part of it, I find that the predictive algorithms that were used at the time of the dissertation did not provide caseworkers the expected value. A fundamental argument of the dissertation is that to know predictive algorithms, we need to observe their use/non-use at the hands of practitioners.

Due to Covid-19, and the following societal lockdowns, which significantly impacted the opportunities for conducting fieldwork, this dissertation also provides the first in-depth study of caseworkers' transition from co-located/in-person consultations with unemployed persons, to remote consultations and to the hybrid format they are currently working in, at the time of finalizing the dissertation (March 2023). Having remote consultations yield benefits and downsides for caseworkers and unemployed persons. However, what is clear is that only having remote consultations is not enough to support caseworkers' need of information to understand the situation of unemployed persons.

In light of the newly elected government’s political ambition to dismantle the job centers, this dissertation provides highly relevant insights from digital public employment services, thus dissemination of the research has been an integrated part of the dissertation. 

Assessment Committee

Professor Pernille Bjørn, University of Copenhagen

Professor Luigina Ciolfi, University College Cork

Professor David Randall, University of Siegen


Principal Supervisor: Associate Professor Naja Holten Møller, Dep. Of Computer Science, University of Copenhagen

Co-Supervisors: Professor Thomas Hildebrandt, Dep. Of Computer Science and Henrik Palmer Olsen, Faculty of Law, University of Copenhagen

Leader of the defense: Tariq Osman Andersen

For an electronic copy of the thesis, please visit the PhD Programme page