The challenge of explaining the rapid increase in self-employment over the last few decades lies not only in understanding what is attracting more people to this type of work, but also in the ability to determine those conditions which support the retention of individuals who venture into this type of work. This article presents a first empirical evaluation of the factors that foster the continuity of self-employment experiences in Canada and contribute to an explanation of its growth. Particular attention is paid to the distinct cases of women and of men. We propose that the decision to proceed further with a self-employment experience at a given point in time depends on the worker’s capacity to meet or go beyond minimal economic results. Given the type of services offered by the self-employed worker, this expected minimal result is easier to achieve during a economic upswing or during a period of economic expansion and also when the self-employed worker has the human capital, the financial resources and the necessary means of production to take full advantage of this favourable economic situation or to compensate for insufficient economic profitability during an unfavourable economic situation. The model suggests that the continuity of the experience may also be favoured by the particular hopes and preferences of the worker who is willing to accept a lower economic result since he or she finds this employment status a more satisfactory response, especially with regards to autonomy and to creativity within work, variety among tasks to accomplish and challenges to be met or a greater use of his competencies. In order to evaluate the influence of the variables in our theoretical framework on the continuity of self-employment, the authors used the Cox proportional risk regression model and estimated the probability of a worker leaving, within a specific month, self-employment status for a salaried job or for a full-time search for such a job (unemployment). The regression model was applied to a sample of some 3,000 persons ranging in age from 16 to 65 from the first panel of SLID (Survey of Labour and Income Dynamics) from Statistics Canada. The model was also applied in a distinct manner to the 1,595 men and 1,237 women in this sample. The follow-up on their businesses was carried out on a monthly basis and only deals with the first experience undertaken by these individuals between January 1, 1993 and December 31, 1998. The average length of these self-employment experiences was 14.4 months for women, and 17.6 months for men. The survival analyses also reveal that after the first year’s experience, the probability of women leaving is higher than for men. However, when one takes into account the set of relevant variables with the Cox regression model, the risk of leaving self-employment for salaried employment or for a job search is 16 percent lower for women. In keeping with the hypotheses found in the theoretical framework, the regression analyses reveal that the risks of leaving self-employment increase when market conditions and specific conditions of the worker’s experience deteriorate. The self-employed worker’s occupational group also allows to predict the precariousness or longevity of his experience depending if salaried employment prospects seem to be relatively attractive or rather limited for his or her occupation. Self-employed workers who work in a rural setting or in small towns also have a lesser risk of leaving their enterprise because of the fewer salaried employment prospects than in larger municipalities. The human capital of the self-employed worker as seen in educational background and workplace experience has a rather mitigated effect on the continuity of his or her enterprise. His financial resources seem, however, to contribute more greatly to his success, particularly investment income and owning one’s own home. Finally, sociodemographic variables turned out to be secondary predictors and difficult to interpret. These imprecise results do not allow to confirm a link between the career aspirations of sociodemographic groups and the longevity of their experience. This continuity of self-employment model applies more to men than to women. Several predictors of the length of self-employment for men seem less relevant when predicting the continuity of women’s experience. Education, experience and investment income for female self-employed workers are resources which are slightly or not at all associated with the permanence of their employment status, while a dependence on government transfers increases the precariousness of this employment status. The conditions of their self-employed work experience also play a less important role. The explanation for the longevity of their enterprise seems to lie elsewhere: a spouse’s financial support, self-employment experience or business management experience; fewer alternatives to self-employment which are the result of reduced employment market mobility and increased concentration in certain occupational groups; career aspirations different from those of men, and a smaller and less profitable job market integration which results partly from their difficulty in reconciling the demands of family responsibilities with those of work
