Return to training and atypical study paths: determinants and valorisation on the labour market.

Summary Lifelong learning is an important issue in our societies. After an initial exit from the education system, it can be characterized by a return to study or by access to continuing education. In France, study paths seem to be less and less linear, and young people are encouraged to take advantage of these interruptions to diversify their experiences. They are then, throughout their career, encouraged to train in order to have "the freedom to choose their professional future" as proposed by a recent law. In view of these observations, the purpose of this thesis is to identify the determinants of these returns to training, and to study their value on the labor market. The preliminary chapter, based on a review of the literature, examines the reasons for returning to school or accessing training, as well as the theoretical conditions for their valorisation on the labour market. The analyses in the first empirical chapter are based on Céreq's Génération 98 survey, which interviews graduates of initial training who have been followed for 10 years. Longitudinal knowledge of their educational and professional career paths makes it possible to evaluate the effect of non-linear educational paths on the wages of young graduates, using the instrumental variables method. The latter, despite having a higher than average educational and social profile and sometimes better endowed with social skills, receive a lower salary than the others, all other things being equal. This seems consistent with the hypothesis of a negative signal sent to employers. In a second chapter, these results led us to examine in greater depth the question of how atypical educational paths are valued on the labour market. Using the vignette method, we interviewed recruiters in order to better understand their perception of these paths. More than a thousand fictitious CVs were evaluated by recruiters (according to the probability that they would offer a job interview). Our econometric analyses show that the fact of not specifying the experience lived during temporary interruptions of studies seems to penalize, ceteris paribus. Faced with a lack of information, recruiters would probably not take the risk of hiring a candidate who had experienced a period of inactivity or unemployment. But this strong signal effect is not observed when candidates specify that they have interrupted their studies to travel abroad or to perform civic service.The third chapter seeks to better understand the determinants of reinvestment in training, but this time by employed workers. We use data from the Céreq Defis survey, which is based on a double survey (employees and companies). Our results show that very few contextual variables seem to affect the probability that the employee will want to train. Similarly, access to non-mandatory training, although more influenced by company characteristics, seems to depend mainly on individual variables and job characteristics and significantly increases inequalities in access to training. The use of a double-difference estimator method with matching shows that this training does not seem to be valued in the form of wages.