Evaluation und Qualitätsmanagement in der Arbeitsmarktpolitik : einige systematische Vorüberlegungen und praktische Ansätze zur lokalen Umsetzung (Evaluation and quality management in labour market policy : some systematic preliminary considerations and practical approaches for local implementation)

"The first part of the treatise deals with the relationship between evaluation and quality management in labour market policy. The considerations on this point refer mainly to the fact that one-dimensional evaluation approaches based solely upon the results of integration balances are problematic both scientifically and socio-politically and especially as regards practical knowledge. They are dubious in scientific terms because in the studies they deal only with intended effects but not in any way with unintended effects; as far as social policy is concerned they are counter-productive as the exclusive look at entry rates in the selection of participants for measures leads to 'recruitment of the best' - in other words the allocative efficiency runs suboptimally; in terms of practical knowledge they are often not very useful since here there are usually hardly any indications of why integration processes have run more or less well. Against the background of this the paper argues in favour of a polyvalent evaluation concept which, especially in the sense of quality management, does not look only at results/products (e.g. labour market entries), but examines processes (e.g. gaining of self-confidence among unemployed individuals), looks at structural conditions (e.g. links) and above all makes the quality of the process a subject of the analyses - in other words how and with what techniques and procedures the tasks are dealt with in the projects/measures for the unemployed. The second part of the treatise then relates the preliminary considerations to practical experience. Here a model project (in accordance with §10 of the Social Code III) is presented in which the labour and social administrations tried out individual made-to-measure methods of development and placement assistance (DPA) for long-term unemployed people with supplementary receipt of social assistance, in order to open up new prospects for this group, too. From the wealth of the various results from the evaluation for the model project, which was evaluated using an experimental and control group design following the polyvalent approach, the paper selects two areas for the following discussions: firstly it looks (conventionally) at the product quality, i.e. at the integration balance, and this is done in comparison with results obtained from two control groups, parallel to the model project, which had deliberately not been given the opportunity to take part in the measure. The survey results in the integration results, which diverge from the original study hypothesis - i.e. the relatively poor entry results in the model project - are then in the paper the reason for looking into an additional assumption: this is the consideration that a suboptimal quality of the procedures and methods in the project could possibly explain the comparatively low level of labour market success. Accordingly this is then followed by a presentation of some results from the examination of this quality dimension, which were the subject of multi-dimensional analyses in the polyvalent evaluation concept of the study. An assessment in this respect of the instruments used by the project organisers for the development and placement assistance in the model, which results in largely poor assessments from independent experts in the overall evaluation leads to the assumption that there could still be quite considerable optimisation possibilities in the methodology of the work with hard-to-place longterm unemployed people. The concluding reflection of the paper then deals with these optimisation possibilities." (Author's abstract, IAB-Doku) ((en))