Data sources, methods and definitions

View the most recent version.

Data sources

Data from the 1991 Census of Population file, which was linked to the Longitudinal Worker file (LWF), were used in this study. The LWF combines several administrative files, including the T1 Personal files, and is a 10% random sample of the population receiving a T4 Information Slip or filing a T1 Income Tax Return. Individuals who were at least 25 years old on the 1991 Census were linked probabilistically to the LWF with a success rate of more than 75%; however, only 75% of the original sample was maintained in the linked file (based on random selection). The individuals who were successfully linked and maintained in the file were very similar to those in the broader sample of 1991 Census respondents based on several socio-economic characteristics, including highest level of educational attainment.

This study examined 15,166 men and women who, according to the 1991 Census, held a college certificate or a bachelor’s degree, were aged 26 to 35 at that time, were born in Canada or came to Canada before age 18, had not attended school in the previous nine months, and appeared in the T1 files in 18 of the next 20 years. The study population consisted of 2,796 men and 3,140 women with a bachelor’s degree, and 3,634 men and 5,596 women with a college certificate. The T4 wages and salaries of individuals were tracked in the LWF from 1991 to 2010. Individuals who do not appear in the LWF in a given year were assigned zero T4 wages and salaries.

Although it is possible that some members of the study sample obtained higher educational credentials after 1991, the number is likely small. According to March and September data from the Labour Force Survey (LFS), of all Canadian-born individuals aged 26 to 35 who had a postsecondary credential in 2006 or 2007, 37.6% of them had a bachelor’s degree, and 11.3% had a credential above a bachelor’s degree. Six years later (in 2012 or 2013), the percentage of individuals who were six years older (but otherwise similar) and had a bachelor’s degree was 34.1%, and the percentage with a credential above a bachelor’s degree was 15%. The percentage with a college certificate remained steady over the period.

Thus, according to the LFS, about 9% of 26- to 35-year-olds had upgraded their education. However, the current study is restricted to individuals who had not attended school in the nine months before the 1991 Census. In the LFS, 11% of 26- to 35-year-olds who held a bachelor’s degree in 2006 and 2007 were still attending school. It is possible that many of them were the people who eventually obtained higher educational credentials, although this cannot be determined because the LFS data are not longitudinal.

Methods

Medians and percentiles are the only statistical techniques employed in this analysis. No regression adjustments were necessary, because the individuals who were compared across and within fields of study shared many important characteristics: they had the same sex; they were about the same age; were born in Canada (or immigrated before age 18); and had the same highest level of education in 1991.

Definitions

Cumulative earnings: This refers to the sum of T4 wages and salaries plus net self-employment income earned from 1991 to 2010, expressed in 2010 constant dollars based on the annual Consumer Price Index (CPI) 2011 basket (CANSIM table 326-0021).

Bachelor’s degree: A university degree at the undergraduate level. It excludes university certificates above or below a bachelor’s degree, as well as first professional degrees (medicine, dentistry, veterinary medicine, or optometry). This study also excludes graduates of law programs, because a Bachelor of Laws (L.L.B.) is generally considered to be a first professional degree. Non-L.L.B. programs such as legal studies were also excluded, because the specific programs could not be distinguished.

College certificate: A certificate awarded by either a college, CEGEP, or other postsecondary non-university institution (excluding registered apprenticeships or trades certificates). To be consistent with the bachelor’s degree category, law programs were excluded.

Field of study: The field of study was identified by the major field of study code used in the 1991 Census. The fields are categorized into nine major groups:

• Education: Includes Educational, Recreational and Counselling Services.
• Fine and Applied Arts: Includes Fine and Applied Arts.
• Humanities: Includes Humanities and Related, as well as No Specialization.
• Social Sciences: Includes Social Sciences and Related.
• Business Administration: Includes Commerce, Management and Business Administration.
• Life Sciences: Includes Agricultural, Biological, Nutritional and Food Sciences.
• Engineering: Includes Engineering and Applied Sciences and Science Technologies and Trades
• Health: Includes Health Professions and Related.
• Mathematics and Physical Sciences: Includes Mathematics, Computer and Physical Sciences.