Transforming Welfare Infrastructures in Europe. The Use and Abuse of Algorithms in Decision Making
Across Europe, the welfare sector is facing increasing demands and shrinking resources. Calls to increase efficiency and effectiveness suggest introducing data-driven decision-support and enhancement with AI. In various countries, public employment services have been one of the main domains that attempt making use of the “knowledge in data” to profile job seekers and support active labor market policies. Austria's public employment service, for instance, is introducing a semi-automated assistance system to calculate the future chances of job seekers. Based on past statistics, it looks for connections between job seekers' characteristics and successful employment. The objective is to allocate resources for further education to those for whom the support measures are most likely to succeed. An analysis of the tensions and biases of the system calls into question the objectivity of data claims, as well as the role of algorithmic managerial practices in the transformation of the welfare state to an "enabling state" that aims at mobilizing citizens' self-responsibility. My talk explores the extensive implementation of automated decision-making in the welfare sector across Europe. It suggests that data-based infrastructures for public administration are shaping welfare systems and transforming citizen-state relations, raising questions of human agency imbricated in complex socio-technical systems.
Speaker: Doris Allhutter, senior scientist at the Institute of Technology Assessment, Austrian Academy of Sciences, where she leads the Austrian team of an international comparative study on Automating Welfare funded by Chanse. Her research focuses on the implicit normativity of computing practices in machine learning under the lens of how these practices are entrenched in power relations. She was a visiting researcher at Vassar College New York, UC Berkeley and Lancaster University, and a Fellow at the Paderborn Research College ‘Data Society’, and the Digital Curation Institute, University of Toronto. Her publications include “Algorithmic Profiling of Job Seekers in Austria: How Austerity Politics Are Made Effective in Frontiers in Big Data”, and “Of ‘Working Ontologists’ and ‘High-quality Human Components’: The Politics of Semantic Infrastructures”.
Introduction: Antonio Vetrò (Lecturer in Data Ethics and Data Protection, Politecnico di Torino)