12.12.2022

How “social” are algorithms in welfare?

Automated processes promise more services and less costs. In sensitive areas such as welfare, the use of algorithms could bring disadvantages for people. In a new project, eight EU countries including Austria compare automation in their welfare state systems.

Bild mit drei Personen, die miteinander sprechen
Auto-Welf analyses input of the people involved in the automation process (unsplash.com)

Data-based infrastructures are becoming increasingly important in public administration. In the welfare sector, partial automations are expected not only to reduce costs, but also to distribute welfare state services more effectively. Algorithms and the analysis of citizen data are supposed to help. But how does this affect the relationship between citizens and the state? And how will the welfare state change as a result?

Transparency and social justice in the first place

"In Austria and other European countries, decisions on the allocation of social benefits are increasingly being partially automated. Many of these systems have failed because they reinforce social inequalities. Others are still in use although their effects are completely non-transparent to the public", says Doris Allhutter from the ITA of the Austrian Academy of Sciences about the European efforts to date to achieve improvements in performance in this area.

The project Automating Welfare - Algorithmic Infrastructures for Human Flourishing in Europe, or Auto-Welf for short, is researching the effects of welfare automation in the area of core welfare state services, such as employment services, health care and the provision of social services. But it also explores the idea of community welfare in the context of so-called smart city and smart village initiatives. Eight European countries, specifically Austria, Denmark, Estonia, Germany, Italy, Poland, Portugal and Sweden have different welfare models and initiatives to automate social services. "Auto-Welf is the first research project to investigate automated decision-making in different European welfare systems. We analyse how it affects social justice, equity and the well-being of citizens in Europe."

AUTO-WELF project page