Decision support and algorithmic support: the construction of algorithms and professional discretion in social work

ABSTRACT

The promise of artificial intelligence and algorithms to make bureaucratic processes more efficient and minimise ‘subjective’ judgement and, hence, professional discretion in case management has recently resulted in the development and implementation of different kinds of algorithms in social work. In this paper, we analyse the ideas, developments and use of three ‘decision-support’-algorithms, developed for Danish municipalities within social work with vulnerable children and families. These algorithms are meant as support of professionals in their decision-making in respect of referrals. Building on recent studies on the effects of digitalisation on frontline work and professional discretion, we analyse the intentions, designs, and workings of decision-support algorithms in social work, specifically in casework with vulnerable children and families. We analyse what is meant by professionals in need of ‘decision support’ and how these algorithms reconfigure ideas about professional discretion. Through the analysis, we find that the algorithms are intended to support decisions by standardising casework, objectivising basis for decision, and by eliminating bias. Furthermore, we document the close relationship between the intentions of algorithms and the conception of professional discretion.

ABSTRAKT

For nyligt har forestillingen om, at kunstig intelligens og algoritmer i fremtiden kan effektivisere bureaukratiske processer og minimere den ‘subjektive’ vurdering og derfor også det professionelle skøn i sagsbehandlingen resulteret i udviklingen og implementeringen af forskellige former for algoritmer i socialt arbejde. I denne artikel analyserer vi ideer, udviklinger og brugen af tre såkaldte ‘beslutningsstøtte’-algoritmer, der er udviklet for de danske kommuner til brug i det sociale arbejde med børn og familier. Meningen med disse algoritmer er, at de skal støtte de professionelle i deres beslutningsprocesser i forhold til underretninger. I vores analyse trækker vi på science and technology-studies og en forståelse af teknologier som apparater, der former og selv bliver formet af praktiske processer og hverdagssituationer, og vi analyserer intentionerne, designet og arbejdet med beslutningsstøtte algoritmer i socialt arbejde, specifikt i sagsbehandlingen af sårbare børn og familier. Vi analyserer, hvad der menes med, at professionelle har brug for ‘beslutningsstøtte, og hvordan disse algoritmer rekonfigurerer ideers om det professionelle skøn. Gennem analysen af tre vigtige temaer, standardisering, objektivisering og at eliminere bias, dokumenterer vi den tætte relation mellem de analyserede algoritmer og det professionelle skøn og det professionelle skøns ambivalente rolle.

KEYWORDS:

• Digitalization
• professional discretion
• algorithms
• referrals

NØGLEORD:

• Beslutningsstøtte
• professionel skøn
• algoritmer
• underretninger
• STS

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Notes on contributors

Marie Leth Meilvang

Marie Leth Meilvang is a Assistant Professor in the research programme Municipal Management and Practice at UCL University College. She holds a PhD from the Department of Sociology at the University of Copenhagen. Her current research focusses on the relationship between professions, organisations, and politics, among other things the methods, tools, and technologies in welfare professional work, and questions of legitimacy and morality in (inter-)professional work. She has published in journals such as Journal of Professions and Organization and The Sociological Review.

Anne Marie Dahler

Anne Marie Dahler is an Associate Professor and leader of the research programme Municipal Management and Practice at UCL University College. Her current research focusses on technologies and normativities in the work of welfare professionals, social inequality and politics, and research interventions. She has been involved in a range of projects on technology, competences, and welfare professional work and has previously published in journals such as Rehabilitation Research and Practice and Disability and Rehabilitation.