Bosses of Their Own: Are the Children of Immigrants More Likely to be Self-Employed than their Parents?

Abstract

Using a generational cohort method, this paper compares the self-employment rates of immigrant parents and their children when they were in the same age range of 25 to 44. The focus is on three questions: (1) Are the children of immigrants more or less likely to be self-employed than their parents? (2) Are the children of immigrants more or less likely to be self-employed than the children of Canadian-born parents? (3) Is the generational change from immigrant parents to their children in the self-employment rate different to that from Canadian-born parents to their children? The results show that Canadian-born male children of immigrants had a lower self-employment rate than their fathers. The decline in the self-employment rate was not unique from immigrant fathers to second-generation men. It was also observed from Canadian-born parents to their children. For both groups, the decline was related to changes in life-course events—longer schooling, fewer marriages and fewer children. Similar to the difference among their fathers, the second-generation men had a higher self-employment rate than the third-and-higher-generation men. Among women, the self-employment rate increased from immigrant mothers to their daughters and from Canadian-born mothers to their daughters.

Keywords:

• Immigrants
• Second Generation
• Third-and-Higher Generations
• Self-Employment
• Canada

Notes

1. For this reason, we include individuals aged 25 to 44 from the 2006 Census so we can compare young adults in the same age range but 25 years apart, although strictly those who were aged 0 to 18 in the 1981 Census should be at the 25 to 43 age range in the 2006 Census. The addition of individuals aged 44 from the 2006 census would slightly increase the unadjusted self-employment rate of the 2006 sample, since age is positively associated with the likelihood of self-employment. However, their inclusion has little effect on the adjusted self-employment rate, as age is controlled in the multivariate analysis. Alternatively, we could restrict the sample from the 1981 Census to age 25 to 43. Both approaches produce similar adjusted self-employment rates.

2. Education levels contain six categories: no high-school certificate, high-school certificate or diploma, non-university certificate or diploma, bachelor's degree, graduate degree and degree in medicine, dentistry, veterinary medicine or optometry. Years of potential experience are estimated as ‘age minus years of schooling and 6’. Marital status is coded as married (including common-law) vs others. Mother tongue is coded as English/French=0, other =1. Housing tenure is coded as owners vs renters. Ethnic/population groups are based on the combination of visible minority status and ethnic ancestry variables. Visible minority status is used to identify sub-groups within the visible minority population, while ethnic ancestry is used to identify ethnic groups within the non-visible minority population. The identified 15 groups are Chinese, South Asian, Black, Filipino, Korean/Japanese, other visible minorities, British/French/Canadian, German, Italian, Ukrainian, Dutch, Polish, Jewish, Portuguese, Other European origins. In the model, British/French/Canadian is used as the common reference group, the combination being used to maintain historical comparability. From the 1981 Census to 2006 Census, the share of the population reporting their ethnic origin as Canadian increased significantly, the majority of whom reported themselves as British or French (Bonikowska and Hou 2010). Geographic regions are grouped into 13 categories: Montreal, Toronto, Vancouver, and ten provinces (excluding the three largest metropolitan areas in their respective province).

3. The conventional Oaxaca decomposition uses either β^f or β^s to compute the ‘explained’ component. The results may differ if the differences in β^f and β^s are strongly correlated with differences between x_f and x_s. We tested the robustness of our results to the choice of coefficients and found some small differences that do not affect our conclusions.

4. When we use potential years of work experience, derived as ‘age minus years of schooling minus 6’, to replace age in the model, the impact of changes in education levels is minimal, but the impact in changes in potential years of work experience is large.