Lessons from mixed-method evaluations—An example from labor market research[Instrumental Variables and the Search for Identification: From Supply and Demand to Natural Experiments]

This article deals with the following questions: which approaches are suitable for program evaluations in the context of labor market research and why are mixed-method designs often most promising? The strengths and weaknesses of different approaches suggest that summative and formative as well as quantitative and qualitative elements should be combined with each other. We use the case of active labor market policies—an evaluation of qualification measures for the unemployed—as an example to illustrate mixed-method evaluations in research practice. The results of the evaluation show that the scientific gain of the formative part often depends heavily on the careful selection of the ‘right’ persons for expert interviews. The findings can even lead to further research projects that deal with some of the questions raised in the formative part in more detail. The summative part of the evaluation—a quantitative impact analysis—is based on extensive quantitative data sets. It is a methodological innovation to combine survey and administrative micro-data in this particular context. We use propensity score matching to analyze program effects on the participants that are positive and substantial.