Step 4-4

Work package 7: Criterion validation of ESeC using the ECT

The validation of the prototype ESeC will consist in verifying that the classification measures work position and market situation, since it has been built for that purpose.


The data used will be the last two enquêtes sur les conditions de travail (working conditions survey, hereafter: ECT), conducted in France in 1991 and 1998, which are a supplement of the 1991 and 1998 enquêtes emploi (French Labour Force Surveys, hereafter: EE), and the annual EE themselves, from 1990 to 2002, and if necessary, from 1982 to 1989. The EE gives information on the market situation (type and duration of the contract, working hours) that will allow us to test that persons classified in each of the different classes of the ESeC share the same type of position in the labour market. The ECT questionnaire makes it possible to measure the autonomy enjoyed by each employee in organising his or her working time, the degree of autonomy in the actual execution of the work, and finally to identify the constraints and the power that individuals can be obliged to exert on other employees of their company. The availability of the data at several points of the 1990s will enable us to test the consistency of the classification over time.


Ideally, a classification measuring the variety of employment relationships should minimise the diversity of employment relationships within the classes and maximise the differences between the various classes. Therefore, our method for testing the validity of the ESeC will consist in comparing the explanatory power of an alternative set of ESeCs on employment relationships and market situations, with discriminant analyses (see Goux and Maurin, 2001). This technique will allow us to assess the ‘best’ classification among the group of potential good ESeCs.