Investigating educational issues across Europe (ACROSS Base)
Proceedings of ECER 2003 Roundtable on methodological approaches in European projects


Access by 
> theme
> approach
> method

Index of
> projects
> experts
> references




David Raffe

Comparative analysis of transitions from education to work in Europe (CATEWE project)

Extracts from:
Walter Müller and David Raffe 
Paper to European Research Network on Transitions in Youth, Florence, 2002 (available on-line)
NOTE: CATEWE papers are available on the project web site: www.mzes.uni-mannheim.de/projekte/catewe

The CATEWE project aimed to develop a more comprehensive conceptual framework of education-to-work transitions in different national contexts, and to apply this framework to the empirical analysis of transition processes across European countries. It used data from two complementary sources: the European Union Labour Force Survey (EULFS), supplemented by analyses of national LFS data for Italy and Spain; and the national school leavers’ surveys of France, Ireland, the Netherlands, Scotland and Sweden. 

The main research question underlying the project was: ‘How do national transition systems shape transition processes and outcomes?’. The project also had a methodological objective: ‘To develop proposals to harmonise existing school leavers' surveys in the participating countries; and to encourage the extension of comparative transition surveys to other European countries currently planning surveys of school leavers’. 


The EULFS has several strengths as a source of data on comparative school-to-work transitions. It provides regular, standardised data on all sections of the labour market in all EU member states; it therefore enables us to study and compare the full diversity of national contexts across Europe, to study trends over time, and to contrast the situation of recent entrants to the labour market with that of more experienced workers. The EULFS contains detailed information on employment outcomes, allowing us to examine their relationship with level and type of education. However, the cross-sectional nature of the survey means that we cannot directly examine the transition process itself. The CATEWE project constructed synthetic labour force entry cohorts (based on ‘typical’ age of graduation from different levels and types of education) in order to examine patterns across different cohorts. These pseudo-cohort analyses have proved to be illuminating when comparing national systems, but they do not allow longitudinal analysis at an individual level, for example to study the effects of interventions on the employment outcomes of those who participate in them. The EULFS provides only limited data on individuals’ educational experiences and their movements within the labour market, and it provides virtually no information on their social background. Analysis of the data is further handicapped by inadequate documentation, which makes it difficult to assess the comparability of the data, and by the restricted access to micro-level data (Couppié and Mansuy 2001, Müller and Gangl 2001). The CATEWE project recommends that Eurostat provides improved documentation of, and access to, EULFS micro-data, including the new transition module added to the EULFS in 2000. This module aims to address some of the limitations of the EULFS for the study of transitions, and it has been evaluated by former members of the CATEWE project team (Iannelli 2002, Müller et al. 2002, Kogan and Müller 2002). 

The school leavers’ surveys have strengths which complement the EULFS. They provide micro-level longitudinal data which allow us to observe individual pathways rather than the aggregate patterns described by the EULFS. They provide more detail on educational experiences and on the processes of transition, and they are sensitive to country-specific institutional arrangements. On the other hand, the surveys are conducted only in a limited range and number of countries, some are infrequent or irregular, and some cover only post-secondary transitions. They do not allow for comparisons between young entrants to the labour market and more experienced workers. Above all, their cross-national comparability is limited. The surveys are designed to reflect national priorities, needs and circumstances. The national surveys vary widely with respect to the variables collected and the way they are coded and classified, the number and timing of sweeps, and the sample design. All five surveys analysed by the CATEWE project are secondary school leavers’ surveys, but each has a different operational definition of a ‘secondary school leaver’. This is a conceptual problem as well as a technical one. Comparisons of different event cohort surveys can only be valid if the event in question – leaving secondary school – has the same sociological significance in each country. But education-work transitions increasingly comprises a series of transition events, whose nature, role and sequence typically vary across national systems. It is not possible to define a single transition event which has the same significance in each country, and which can therefore be the basis for comparison (Brannen and Smyth 2001, Raffe 2001a). 

The CATEWE project defined a set of ideal requirements for a data and indicator system on comparative education-to-work transitions. When current data sources (school leavers surveys, EULFS and administrative data) are evaluated against these criteria, none fully satisfies all the requirements, and none can easily be adapted or modified to do so. In particular, there is limited scope for harmonising national school leavers’ surveys, which differ widely in design and content, and which serve distinct national purposes. The project proposed criteria for the partial harmonisation of these surveys, to increase comparability and promote ‘best practice’. Strategies for new data-collection should aim to fill gaps in existing sources, including: 

  • data on the demand side of the labour market (for example, employers’ strategies and behaviour for recruiting, selecting and recruiting young workers);
  • the collection of ‘equity’ variables (gender, social background, ethnicity/nationality) on a consistent basis: 
  • data on itineraries within the education and training system; 
  • subjective data collected prior to key decision points; and
  • ‘substantive comparability’ between countries: that is, comparability not based on a single transition event. 
The project recommends the initiation of a European-wide survey based on a prospective age cohort design starting at about fifteen years, followed over a period of about ten years. This would enable us to examine decision-making processes among young people at the point of leaving compulsory education and their subsequent trajectories through the education and labour market systems. The PISA longitudinal survey provides a model and a possible basis for such a survey (Raffe 2001a). 

Further reading:

David Raffe, University of Edinburgh
In: Alain Degenne, Jean-François Giret, Yvette Grelet and Patrick Werquin (eds) Les Données Longitudinales dans l’Analyse du Marché du Travail: 10es Journées d’études Céreq-Lasmas-Institut du Longitudinal, Université de Caen, 21-23 mai 2003. Documents Séminaires no 171, Céreq, Marseille


Top of the page