Endnotes

View the most recent version.

1 The unemployment rate for high school dropouts (for the age group 15 years and over) oscillated between 12% and 17% while the national unemployment rate fluctuated between 7% and 11% during the period. Source: table 282-0004 in Cansim, Statistics Canada.

2 Social costs are also important. A study by the Conference Board of Canada (Lafleur, 1992) estimated that 140,000 dropouts – or roughly 5,000 more than the number of Canadian dropouts aged 20 to 22 in 2001 – cost near $4 billion over their working lifetime in 1989, which, put in current dollars, corresponds to $58.7 billion in 2005.

3 Calculated by the author using the first two cycles of the Youth in Transition Survey, Cohort B, of Statistics Canada. This proportion excludes the individuals who were enrolled in PSE or had already completed a PSE certificate in 1999. When these individuals are not excluded from the calculations, as calculated by Zeman, Knighton and Bussière, the return rate reaches 27%.

4 The higher return rates among young women probably explain partially their lower dropout rate at the age 20.

5 The total population for the 20 to 24 year-olds corresponds to the population for the calendar year and not the academic year. In the calculations, the academic year-based data are divided by the corresponding full year. For example, the dropout rate uses the number of dropouts in the academic year 1990-1991 divided by the total population for 1990.

6 The definition of a dropout in this section corresponds to individuals from groups A and B in Figure A.1, those without a high school diploma not attending school.

7 Bowlby (2005) presents dropout rate also based on the Labour Force Survey. His dropout rate is based on the highest educational attainment (highest diploma or degree achieved) and thus excludes those without a high school diploma who have completed a PSE diploma from the numerator of the rate. Approximately two to three percent of all 20 to 24 year-olds do not have a high school diploma and do have a PSE diploma. As such, the numbers presented here are higher by two to three percent than those of Bowlby.

8 The Labour Force Survey gathers information about school attendance, whether attendance was fulltime or part-time, and the kind of school attended. The respondent has a choice of primary or secondary school; community college, junior college, or CEGEP; university; and other.

9 The largest increase has been in university programs, most likely university certificates targeted to mature students. Most of the increase took place since 2000/2001 for women and 2003/2004 for men, with the increase being strongest for women. It is unclear if this trend will hold. It should be further noted that employer-sponsored courses would also be captured in the data.

10 This section discusses dropping out and returning to school from the economics literature perspective. Dropping out has been described as a "fading out" process. See Gilbert, S., L. Barr, W. Clark, M. Blue and D. Sunter "Leaving School: Results from a National Survey Comparing School Leavers and High School Graduates 18 to 20 years of age." Human Resources Development Canada and Statistics Canada, no 81-575E. 1993.

11 Bushnik et al, (2004) show that future dropouts at the age 15 had significantly lower results on PISA for reading, a standardized test from the OECD, than students who later graduated. Male dropouts exhibit even lower scores than female dropouts.

12 Additional modifications can be brought to the human capital investment model to allow for changing preferences over time. This modification permits to account for social pressures and other environmental factors that may lead youth to drop out.

13 The "leftover" schooling to complete also applies for an individual choosing to pursue a college degree as s/he will most likely have to go through an academic upgrading program based on what s/he is missing.

14 The concept of economic region is defined by Statistics Canada and is used for the analysis of the labour force. It covers the geography of one or many census divisions depending on the population density.

15 The number of quarters out of high school is measured as of December 1999 and includes the summer quarters.

16 When those ignorant of their parent's education are excluded, the proportions of men and women whose parent has a PSE diploma are 21% and 23% respectively.

17 The decisions to work and to go to school are most likely contemporaneous. The difficulty here is to determine whether these decisions are done ahead of time or not. A much more elaborate decision model would be required to explore these issues, but is unwarranted for the testing of the present hypothesis of involuntary leavers returning. It should be noted that instrumentation of working or not in the fall of 1999 was tested using the unemployment rate following Parent (2006) along with provincial dummies. Other attempts used the share of primary industries in the local labour market, calculated using the number of people employed by industry over the total employed in the 2001census. Both types of instruments turned out to be ineffective.

18 This was tested by running one model for both sexes in which gender was interacted with PSE aspirations. The interaction term is not statistically significant with a z-statistic of 0.35 (calculated by bootstrap).

19 The census also includes non-permanent residents. It is unclear whether there is a higher incidence of non high school graduation among non-permanent residents than among the rest of the population. Hence it is not possible to evaluate how the inclusion of this group in the calculations affects the dropout rate.

20 Ideally, only those enrolled at the secondary level should be excluded from the calculations if dropping out is defined as meaning groups A and B. If a dropout implies only individuals from group A, then all attendance should be excluded from the calculation of the dropout rate.

21 Some authors in the education literature will make further distinctions. For example, individuals attending regular high school after dropping out are not considered dropouts but those attending adult school or diploma equivalency program are still counted as dropouts. In such a case, the dropout rate would fall between the non-completion rate and the dropout rate as defined here.

22 There are 189 high school dropouts who were enrolled in or had completed a postsecondary education diploma in December 1999. These individuals are eliminated from the sample as several pre-return characteristics were not collected. Referring to the Figure A1 in Appendix A, the sample is made exclusively of group-A individuals.

23 The re-weighting took into consideration personal, familial and geographic characteristics. The reweighting was done at the provincial level. Quick comparison of the dropouts found in Cycle 1 only and those found in both cycles reveals very little difference, with the exception of not knowing the last grade completed (less likely to have responded in Cycle 2). See Raymond (2003) for details about the reweighting procedure done by Statistics Canada to adjust for the attrition.