Deutsche Übersetzung des Titels: Kausale Effekte aktiver Arbeitsmarktpolitik in Europa : Grundlagen und Empirie

Vorschau

PDF, Englisch Download (574kB) | Nutzungsbedingungen

Übersetzung des Abstracts (Englisch)

Over the last two decades most European countries have been subject to a strong and persistent increase in their unemployment rates, leading these countries to attempt to combat obdurately high unemployment by exercising measures of Active Labor Market Policy (ALMP). These measures can be classified into three types of labor market programs: (i) training programs, such as classroom training or on-the-job training, (ii) incentive schemes, such as wage subsidies, and (iii) direct job creation in the public sector. The question whether these policy instruments are effective clearly implies the necessity to evaluate any such program, in order to identify a causal effect of the labor market program on the desired outcome measure. This thesis assesses causal effects of European ALMP from a threefold perspective. First, it gives a thorough discussion of methodological issues that arise in the venture of causal inference. This analysis includes an overview of the predominant procedures to model causation in the empirical sciences. It subsequently focuses on a particular statistical model - formulated in terms of counterfactuals - and the causal queries that can be asked, and answered, within the model. Second, it places the undertaking of combating unemployment by means of ALMP in a European context. Following from disillusioning diagnostics on the state of European labor markets, it shows that a large variety of active labor market programs has been implemented across countries in Europe. The analysis proceeds to investigate the findings of scientific evaluation research on the effectiveness of these measures, and concludes with a set of implications for economic policy. Third, it presents two microeconometric studies on ALMP effectiveness. These studies analyze the causal effects of three labor market programs in Poland using the statistical method of matching, a specific nonexperimental variant of the causal model laid out in the first part.