Zum Hauptinhalt springen

RWI Projektberichte

2010

Ronald Bachmann, Daniel Baumgarten, Hanna Kröger, Sandra Schaffner, Matthias Vorell, Michael Fertig

Study on various aspects of labour market performance using micro data from the European Union Labour Force Survey

Contract No. VC/2009/0123
Final Report – October 2010
Research Project for the European Commission – DG Employment, social affairs and equal opportunities

With this study, we, the consortium of RWI (Rheinisch-Westfälisches Institut für Wirtschaftsforschung, Essen) and ISG (Institut für Sozialforschung und Gesellschaftspolitik, Köln), submit the Final Report of the „Study on Various Aspects of Labour Market performance using micro data from the European Union Labour Force Survey (EU-LFS)”. The study aims at fully exploiting the richness of the EU-LFS micro data to examine in detail the current situation and past developments of the labour markets in the EU. The following research topics are investigated in detail: Task 1: Labour market participation, full-time/part-time employment, and number of hours worked; Task 2: Duration of unemployment and the job search methods of the unemployed; Task 3: Labour market transitions; Task 4: Temporary employment; Task 5: Education and training; Task 6: Intra-EU mobility and migration. In this endeavour, all EU Member States will be covered and a specific focus will lie on similarities and differences between Member States, as well as developments at the level of the European Union. The target population consists of individuals aged older than 15 years which will also be investigated at a sufficiently disaggregated level by taking the skill-, gender-, and age-dimension of the research topics into account. The LFS data for the years 1998-2008 forms the basis of the empirical analysis, which explicitly considers the gender perspective throughout all working steps. In general, these steps proceed as follows. In the introduction to each task, we discuss the importance of the topic under investigation, as well as the most important academic literature. The first step of the technical analysis contains the descriptive evidence computed from the EU-LFS data base. For the variables under investigation, we thus present averages for the dataset as a whole, as well as for different worker groups (e.g. according to age or education), countries, and years. A particular emphasis is placed on the computation and presentation of pivot tables and Markov transition matrices (for labour market transitions). Where we present graphs in the document, the corresponding pivot tables are contained in the appendix. For all the tasks, the econometric analysis is conducted in a second step. Here, different econometric tools are used in order to establish the statistical relationship between the variables of interest. The explanatory variables are made up of two distinct sets. The first set of variables consists of individual and household characteristics, a particular strength of the EU-LFS data base. In this case, the analysis takes place completely at the level of the individual worker. The second set of variables relates to indicators at the macro level with respect to either economic conditions (e.g. the growth rate of GDP, the level of unemployment), or to the institutional framework (e.g. the importance of firing costs, expenditure on active labour market policy measures, etc.). However, it should be pointed out that the results using the second set of macro variables should be interpreted carefully. This is so because, despite the fact that we are using a large data set on individual workers, the analysis using the indicators at the macro level only captures cross-country variation and thus relies on a very restricted number of observations. For each task, we summarize the most important results in the concluding section of every chapter. Finally, the study includes a chapter on practical issues concerning the use of the EU-LFS data. In particular, we discuss issues of data quality (especially with respect to household variables), data access, and the coding of variables.