Neufassung der Arbeitsmarkttypisierung im SGB II

Abstract

"This documentation presents the redesigned SGB II labor market typology for the 2027 planning year. Driven by shifting regional dynamics and the introduction of an updated performance indicator System - comprising integration rates for women, as well as for long-term and non-long-term beneficiaries - a comprehensive conceptual overhaul was required. Methodologically, the typology employs a two-stage classification approach. In the first stage, a machine learning algorithm (LASSO procedure) was utilized for the first time to select the 47 most statistically significant context variables from over 170 potential factors. In the second stage, 392 job centers were categorized into 15 comparison types using cluster analysis (Ward’s and K-means methods). These categories are further consolidated into three "super-types"based on their respective shares of long-term unemployment. The new typology provides a robust, fully data-driven foundation for fair performance comparisons between job centers by accounting for regional factors beyond their immediate control. Due to the fundamental methodological shift, direct comparisons with previous typologies are not advisable." (Author's abstract, IAB-Doku) ((en))

Cite article

Haas, A., Miele, K., Moritz, M. & Dauth, W. (2026): Neufassung der Arbeitsmarkttypisierung im SGB II. Nürnberg, 25 p. DOI:10.48720/IAB.GP2607.01