A
C
R
O
S
S
             
 
  Investigating educational issues across Europe (ACROSS Base)
Roundtable
Proceedings of ECER 2003 Roundtable on methodological approaches in European projects
Home

Introduction

Access by 
> theme
> approach
> method

Index of
> projects
> experts
> references

Search

Roundtable
 

BACK TO CONTRIBUTIONS

Comparative analysis of transitions from education to work in Europe

Issues for discussion

Moderator

David, you have been involved in the CATEWE project, which stands for 'comparative analysis of transitions from education to work in Europe'. Your methodological objective in this project was to develop proposals to harmonise existing school leavers' surveys in the particular countries. What was the challenge in this harmonisation?

David Raffe

While our approach was quantitative, it actually raised similar issues to Anja's, but in completely different ways. The problem of quantitative research is not that you are negotiating meaning within your subjects, but that you are confronting meanings already embedded in the data that you are analysing. This is the case in the project CATEWE, which is engaged in a secondary analysis of existing data.

What the project was trying to do as a whole was to conceptualise and identify what we call transition systems: What are the main features which distinguish different countries' arrangements - institutional, social, cultural etc. - for the transition from vocational education to work, and how do those transition systems shape the processes? For instance, in different transition systems you might find different types of relationships between educational qualifications and employment or unemployment; you might find different patterns of gender or class or other inequalities. So those are the kinds of questions we were addressing. 

We looked at two data sets. One was the European Union labour force survey, which had the advantage of formal comparability - and I stress formal comparability -, because you have really got the definitions of both national and EUROSTAT statisticians built into these data sets. But on the other hand, they didn't ask a lot of questions about transitions; it wasn't at least at that time a longitudinal survey. We also tried to link data sets from five national school leavers surveys, which had contrasting strengths and weaknesses. The surveys individually were all constructed for nationally specific purposes; they used nationally specific definitions; the frames of reference were very different. On the other hand, they did provide more details of transition from education to work. So we had these two rather complementary types of data.

Our methodological objective was to learn from this experience and to try to make recommendations on what future data sets in Europe or generally cross-national research on transitions might look like. To conclude, I would like to identify two key issues that confronted us. First,  there were practical, tangible problems with our data sets: for example, they might lack a particular variable, - in France it was almost impossible to ask questions about ethnicity, and in a lot of countries surprisingly there were not questions asked about the social backgrounds of children and the actual information we had available varied very much from survey to survey. But also, more importantly, underlying all this there were different conceptualisations of some of the key categories and key terms of transition. 

A very simple example of that: We at least pragmatically defined our survey as secondary school leavers surveys. Underlying that is the fact that 20 years ago one tended to conceptualise the transition from education to work as a single event: One day you are in school, the next day you are in a job, or at least three days later. Now it's increasingly a complicated event which in fact moves backwards and forwards between different state. Of course, the way in which this longitudinal complexity works out varies from country to country; the actual stages of activity you go through don't follow as the same pattern in each country. 

In that context it is very hard to define a common conceptual frame of reference that would apply equally to all countries. If you look for example at the OECD indicators or the EUROSTAT key data they tend to be based on this rather simplified model of a single transition. So the key event is either the date when you complete education or the date when you enter the labour market. Now, trying to define either of those terms in an unambiguous way that makes sense in different countries is actually impossible. So our recommendation was a very conceptually open model based on age cohorts. This would be very difficult, probably very costly to follow through, but would actually be the model that lay behind the PISA proposals which have still unfortunately not won a lot of support from countries.

The other issue was a more general one about comparative methodology. The approach we were following was to try to construct a single data set on individual young people that covered thousands of people in five countries from school leavers surveys analyses. That assumes that one can actually construct common variables across those five countries. An alternative approach which a lot of other comparative work on transitions has taken uses the comparison at a higher level of abstraction. Firstly you make analyses within countries, following a country specific set of concepts using variables that might or might not have a direct equivalence in other countries, and then rely upon a higher level of abstraction, or theorisation perhaps, to make the comparisons across countries. So I think the question that I wanted to raise at the end is: given these two approaches to comparative methodology: which is actually the more productive one in the long term? Maybe the answer is - one needs a bit of both.

BACK TO CONTRIBUTIONS

Top of the page
© WIFO