This dissertation presents an interesting contribution to the emerging field of critical data and algorithm studies. Machine learning, data platforms and big data are often associated with social media and tech companies such as Google, Amazon, and Microsoft. These technologies have however gained ever increasing interest in the Norwegian public sector in the last years. Policy documents and consultancy reports promise that data-driven systems will make public administration more efficient, seamless, and adaptive. At the same time social scientists have pointed out various issues associated with the arrival of datafication. By providing the state with the ability to compile and link ever increasing amounts of data, datadriven public administration might enhance the state’s possibilities to surveil, understand, predict, and control citizen activities. In other words, it has the potential to seriously affect citizen-state relations and power.

The idea of “data-driven public administration” is central to this dissertation. I define public administration datafication as being made up of two interwoven processes. Firstly, the use of more and different data in administrative processes. The Norwegian welfare state has produced and managed large amounts of data on its citizens for decades. Data are now imagined fueling all aspects of public administration. At the same time, we can observe an ever-increasing interest in complex technology such as machine learning and automated decision making to recirculate and link data both within and outside of the public sector.

This dissertation provides a unique insight into the inner workings of public administration datafication. I make use of ethnographic field work, interviews, a survey among public sector employees and document studies of policy and strategy documents. The dissertation focuses on what practitioners, policy makers, and institutions do, say, and imagine in relation to datafication, rather than measure its immediate effects on society. What is the citizen’s role in this administrative reform? What is problematized by practitioners and policy makers and what issues of datafication are often silenced? Why is datadriven technology portrayed as inevitable, necessary, and apolitical? Can the challenges that many practitioners experience in working on datadriven systems tell us something about alternative ways of imagining the future of the welfare state?

Through four papers I invite the reader to critically reflect on the interplay of data, technology, power and politics and the role of datadriven technology in everyday life.