Our data is derived from a 1994 survey of young French persons in employment who had left school 4
years previously with a low level of educational level1. Information was collected via an interview
questionnaire. In addition to socio-economic information (parents’ occupation, marital status, number of
children, etc.), the survey covers education and gives details of the previous occupation (wage, type of
contract, sector, function in the company, etc.). It also includes recall questions covering these 4 years
and allowing us to build up some historical variables on employment status (experience, etc.). Our
sample consists of 7188 individuals.
We consider as migrant an individual declaring that he or she had moved from one French
département2 to another at least once during the previous 4 years. It is reasonable to suppose that
these migrants will have changed their local labour market, although “border” effects will of course
exist in certain instances.
Wage determinants may be divided into four categories: personal characteristics such as sex and age;
“human capital”, i.e. migration, educational level and professional experience; job characteristics, i.e.
skill level, function, work contract and weekly work time; company characteristics, i.e. sector, size
and location. The decision to migrate is essentially governed by human capital (educational level and
driving licence) and personal characteristics (sex, marital status, parents’ job situation). We also
introduce variables to take into account wishes concerning job (wish to change job, job search).
Section 4: Empirical results: Wage determinants
In our sample, 43% of those interviewed changed residence but less than half (19% of the total
sample) migrated over a long distance. Our sample covers a generally low educational level but a
distinction may be drawn between two principal categories: persons possessing a professional
qualification, i.e. vocational training certificate (CAP), technical education certificate (BEP), and those
without professional qualifications. The former category accounts for 75% of the sample. In view of
the fact that women tend to be “better educated” than men in France, men are over-represented in the
sample (69% of the total sample). In 1994, 57% of the interviewees were single. With regard to the
parents of these young persons: 68% have fathers in employment, 51% have mothers in employment;
less than 5% have unemployed fathers or mothers; fathers are retired in 16% of cases, mothers
inactive in 37% of cases. For the sample as a whole, the average wage is 5938 FF. The difference
between migrant and non-migrant wage is very small and is not significant: migrants earn on average
1.7% more than non-migrants3.
Comments on the decision to migrate
1. Individuals dropping out during the final year of a short vocational cycle, or in the 6th form (or earlier) of a
secondary school
2 A départemenits an administrative geographic subdivision, roughly equivalent to the British county. In France,
there are about 100 département.s
3 Complementary descriptive statistics on explanatory variables in the Appendix.