Neufassung der Arbeitsmarkttypisierung im SGB III

"As part of its controlling activities, the Federal Employment Agency (BA) uses regional classifications on the basis of a specially designed two-stage classification approach established by the IAB. With these regional classifications, the employment agency districts are split into different clusters called comparison types. Agencies that belong to one and the same comparison type are similar to each other with regard to labor market conditions that – in addition to the actions of the agencies – are co-determinants for the achievement of labor market policy objectives, such as the integration of unemployed persons into employment or the filling of vacancies. In the Federal Employment Agency’s system of management by objectives, the comparison types of the IAB represent an important basis for the structuring of regional target level controls. These are only appropriate if it is ensured that the agencies whose target level values are being compared are those with similar context conditions. Agency groups of this kind are provided with the comparison types. The IAB has developed a new version of the Social Code III (SGB III) classification for the BA. This new version was necessary because changes in some of the regional labor market conditions were likely since the last classification, calculated in 2017 with data from 2016. In addition, the new classification needed to be aligned with the modified SGB III system of management by objectives of the BA comprising two target figures, namely the SGB III integration rate as before and, additionally, the vacancy filling rate. The SGB III integration rate relates the number of cases of integration into employment subject to social security contributions or self-employment to the total population of SGB III clients (registered unemployed and participants in labor market policy measures). The vacancy filling rate indicates how many vacancies were successfully filled, taking into account the regional establishment structure. Like earlier classifications, the new one is based on a two-stage approach. In the first stage, labor market conditions that are relevant for the target figures are identified and their relative importance is determined. In the second stage, agency districts with similar labor market condi-tions are classified into comparison types. As a starting point, a catalog of more than 40 variables is used to quantitatively map the context conditions in regional labor markets. These include various aspects of labor market conditions, economic and settlement structure, and the situation on the labor market. The relevant context conditions are selected with the help of the LASSO procedure, a machine learning algorithm. The procedure optimizes the prediction quality of the target variables and, at the same time, reduces the number of predictors. The selected variables are included in a cluster analysis to form different comparison types. Be-forehand, the variables are standardized and weighted to ensure that context conditions with higher explanatory power are taken into account to a greater extent. The result of the classification is 13 comparison types. Based on the unemployment rate and the settlement structure, these can be summarized in four groups consisting of the comparison types Ia to Id, IIa to IIc, IIIa to IIIb and IVa to IVd. Even within these groups, the individual comparison types differ in terms of unemployment rate and settlement density. Furthermore, there are differ-ences in the size structure of establishments, the structure of industries and the employment-to-population ratio. Group I includes, on the one hand, rural, mostly eastern German agency dis-tricts with above-average unemployment (Ia and Ib), and on the other hand, rural or urban dis-tricts throughout Germany with average to slightly above-average unemployment (Ic and Id). In contrast to group I, group II concentrates on western Germany and is characterized by agency districts with below-average unemployment. Due to their particularly low unemployment rates and (very) pronounced seasonal dynamics, the comparison types IIIa and IIIb, all but one located in Bavaria, differentiate from this group. The fourth group consists of four (large) urban compari-son types that differ from one another in terms of unemployment rate and establishment size structure. Overall, the present SGB III classification 2024 shows that the concept of comparison types, which has been expanded by methodological innovations, can be successfully adapted to changes in the BA target system." (Author's abstract, IAB-Doku) ((en))