Les transitions sur le marché du travail: une perspective longitudinale

Résumé

La précarité de l'emploi caractérise de plus en plus la culture contemporaine du travail. Les individus sont appelés par conséquent à subir des transitions sur le marché du travail. Certaines sont volontaires mais plus de 80 % sont involontaires et parmi les personnes qui subissent de telles transitions une première fois, plus de la moitié subiront une transition subséquente. Nous nous attardons ici aux seules transitions entre l'entrée et la sortie en emploi, qui touchent plus du tiers de la population active à chaque année au Canada, d'après les données recueillies par Statistique Canada auprès de plus de 65 000 personnes au cours des années 1986 et 1987.

Abstract

Work precariousness is becoming a major characteristic of the contemporary work culture. This means that individuals are likely to experience a relative mobility in the labour market, what will be called here the "transitions". These embrace numerous aspects: lay offs, return to work after an unemployment spell, mobility between different employers, etc. Such changes are considerable: nearly hait of the active population in Canada experiences a transition, whatever it's type, at least once a year. However, the underlying processes are not well known. The data analysed have been collected for the LMAS project ("Labour Market Activity Survey"), a joint effort of Statistics Canada and Employment and Immigration Canada. The longitudinal nature of the data collected takes into account the dynamic aspect of reality, of situational changes in the labour market. Once the distinction is made between three main types of transitions, special attention is paid to the one likely to be the most revealing of work precariousness, that of the transitions between the entry and the exit in employment (v.g. employment-unemployment). This type affects more than one-third of the active population each year.

The "transition rates" and the "inter-transitions intervais" are analysed in the perspective of "Survival Analysis", one of the most important steps of the "Event History Analysis" method. The "transition rates" are probabilistic measures that indicate how much chance the survivors (those who havent experienced a transition over the period) have of experiencing a transition in the course of the ensuing two years. The "inter-transitions intervais" indicate the time spent between two transitions and are thus used as a homogeneous computation base to evaluate the proximity in time of the transitions. This proximity in time is used here as an indicator of precariousness, at least within some pre-defined limits. Conceptually, it is apparent that this proximity in time is linked to the occurence of transitions and hence, to work precariousness.

The most striking result lies in the relatively high proportion of individuals likely to experience, after the occurence of a first transition, a subsequent one. This proportion very slowly decreases over time while remaining at a quite high level, always between two-fifths and three-quarters of non-survivors up to the fourteenth transition, up to which point the data still meet an acceptable level of significance.

As expected, the transitions are concentrated in the younger age groups: the relationship between transitions and age is quite strong. In the case of the first transition, three groups can be singled out. The younger, up to 34 years old, are the more severely affected; persons between 35 and 54 years are the most "stable"; and finally, a slight increase appears with persons over 55, among whom are those about to retire. As for other transitions, the more there are of them, the greater is the concentration among younger age groups.

As it happens, there is a very limited relationship between individuals' gender and the transition rate. Women are slightly more numerous in experiencing a first transition than men, who themselves are slightly more numerous in experiencing multiple transitions (6 and over), but differences are small and are more linked to age. Contrary to expectations, relationship with schooling is not linear. The extremes ("0-8 years" and "university") are more "stable" than the intermediate categories ("high school completed or not completed", "post-secondary diploma" and specially the median, "post-secondary not completed").

This examination of sub-groups is carried out with a statistical test based on inter-transitions intervais. The fore-mentioned results generally persist but some distinctions appear as the transitions are cumulated. The test allows to establish, for example, that after a certain number of transitions, the transiting probabilities are not different for the younger compared to older age groups (after the first three transitions), for the moderately schooled individuals compared to those at the extrêmes (after the first two transitions), etc., except that the probabilities effectively met at the occurence of these first transitions, significantly higher, have strong consequences. A large number of persons are sent through the vicious circle of precariousness, as the likelihood of a subcohort experiencing another transition ranges between two-fifths and three quarters of the initial sub-cohort. Consequently, the younger sub-groups and the moderately schooled ones are more affected by work precariousness.