Statistical profiling of the unemployed: a literature review and the case of Denmark

Description

Accelerated by technological developments in AI and big data, statistical profiling tools to identify the risk of long-term unemployment have proliferated in the last decade across the OECD. However, many of such tools also seem to fail due to a number of issues such as public controversies, lack of accuracy, and tension with caseworker discretion. In this talk we address this topic in two ways. First, we present the current research landscape, a landscape divided into several disciplines and with limited knowledge of how the legitimacy of profiling tools is maintained and challenged. In the second part, we attempt to fill this research gap by presenting the (preliminary) findings of a study of the legitimation, controversy, and (ultimately) failure of a statistical profiling tool developed by the Danish Ministry of Employment and used by the jobcentres and unemployment insurance funds from 2015-2022. The study highlights the dominant ‘repertoires of evaluation’ used to justify as well as criticize profiling tools.

Period	4 May 2023
Held at	CEDIC Research Centre for Digitalisation of Public Services and Citizenship, Norway
Degree of Recognition	International