Discretion, automated decision-making and public values: Background and test of an approach for unpacking human and technological agency

Abstract

This study aims to develop a theoretical and analytical approach for studying discretion, automated decision-making (ADM) and the consequential public values. This is achieved through our proposed approach, an overview of literature and an empirical test. The context of our empirical test is decisions made about economic support in social work. The research questions are as follows: 1) What are the relevant components in a theoretical and analytical approach with this specific aim? 2) How does human, non-human and joint, ‘hybrid’, agency influence digital discretion and the consequential public values in social work? 3) What are the usefulness and prospective problems with the approach? Our approach enhances the understanding of the ground for discretion in ADM, which is seen as an emergent routine in the form of knowledge about process, and the details of human and non-human actors involved in relation to consequential public values. To develop its usefulness, the approach should primarily be applied in multiple case studies of ‘ín-between’ contexts, such as social work, to generate a theory of the role of human and non-human agency in the consequential public values of ADM.