Who Is the Problem Represented to be? Subjectification and the Social Division of Automated Welfare Surveillance

Automated welfare surveillance has become increasingly more popular in public sectors. Sweden is no exception; on the 1st of January 2024, a new Swedish authority was launched. The UBM, short for Utbetalningsmyndigheten, is a public agency that will utilize system wide data analysis in order to detect, forestall and prevent inaccurate welfare payments. Meaning that the authority will screen welfare recipients for fraud using algorithms and data analysis. Previous cases of algorithmic welfare fraud schemes indicates that there are significant risks of harm when states utilize automation when conducting welfare surveillance.

Sweden is commonly recognised as a social democratic welfare model, meaning that the state provides welfare benefits to all citizens. The Swedish welfare state makes a difference between allowances, deductions and subsidies. While this division formally takes shape as an organisational difference, the division also has moral and social implications. It is argued that tax reductions and tax reliefs tend to require far less documentation and control than social welfare, a notion that very closely can be connected to the idea of the criminalisation of poverty.

This paper seeks to engage with the issue of who is considered to be the problem when the Swedish state establishes an automated welfare surveillance scheme. With the WPR framework (short for what’s the problem represented to be?), this paper will answer the research questions: Who is the target population in the establishment of the UBM? And how has this problem representation come about? By making use of public reports and official documents, this paper highlights who the presumed welfare fraudster is, and how the idea of who is presumed as the “problem” has changed over time.