As developed countries cross the threshold from massified to universal post-secondary education (Marginson, 2016; Trow, 2007), they are becoming more circumspect about the upper limits of benefits to be obtained, socially and individually, from such levels of participation.[1] It is no longer readily assumed that more education leads automatically to rewards in the labour market. Instead, researchers have begun to consider more specifically what kinds of credentials and skills are most likely to yield employment.

A recent Statistics Canada study, “Do postsecondary graduates land high-skilled jobs?” is an example of these more nuanced approaches to the relationships between education and employability. In this work, Frenette and Frank (2017) shed new light on credentials and occupational attainment recorded in the 2011 National Household Survey by linking these  outcomes to specific skills profiles. The findings uphold the basic relationship between higher level credentials and higher skills used in graduates’ subsequent jobs, but also suggest troubling patterns of underemployment for graduates from a number of educational fields.

The starting point for the study was a taxonomy of skills developed by the US Occupational Information Network. O*NET, a comprehensive, publicly accessible on-line database, uses this taxonomy to create occupational profiles for occupations catalogued according to US Standard Occupational Classifications (SOC). The researchers used factor analysis to reduce the 35 skills in the O*NET to a more manageable set of nine skill clusters: reading comprehension; writing; mathematics; science; process, complex problem-solving and system skills; technical design and analysis; and resource management (Frenette & Frank, 2017, p. 9).

Frenette and Frank created concordances between SOC classifications, and those of National Occupational Classification (NOC) used in Canada’s National Household Survey. This allowed them to map the O*NET skills classifications on to NOC occupations. Skills classifications could then be used to examine links, draw from NHS data, between:

  • skills used in the occupations people reported in the census
  • educational fields;[2] and
  • highest obtained educational qualifications.

Through this process, the researchers were able to generate data that shows what skills recent Canadian graduates are using on the job. The sample for this study, drawn from census data, comprised 247,781 men and 196,302 women, all of whom worked 30 or more hours a week. Via a series of regression analyses, the authors break out skills used in occupations by incumbents’ 1) level of qualification; 2) field of education; and 3) gender.

Study Findings

First, the analysis confirms what is generally understood to be a positive relationship between occupational skill level and level of education. Numerous variations on this theme that overall would seem to reflect longstanding structural features of the labour market. For example "technical operation and maintenance skills" are highest in trades-related occupations regardless of educational attainment, and lower math skills required in occupations held by those with professional credentials are accounted for primarily by the inclusion of the large category of legal professionals in this category.

Second, the authors find that skill requirements vary significantly by fields of study, with occupations typically linked to STEM fields calling for both more skills, and higher levels of those skills among incumbents.[3] These are troubling findings for students and graduates in fields like education, fine arts, and humanities. At both college and university levels, this study finds that graduates from these fields occupy jobs that in most all cases require lower level skills, including reading comprehension, writing, and critical problem-solving skills. One conclusion that might be drawn here is, as the authors note, that graduates in these fields may be suffering more mismatching and underemployment (p. 14).[4]

Another interpretation, less kind to these disciplines, is one considered by Usher (2017): that these “soft” discipline graduates are in fact less skilled than their STEM counterparts. However, this raises questions about how “skills” themselves are measured and valued in workplaces. Given that O*NET skills are filtered in part through the lenses of job incumbents themselves, it is fair to wonder whether, for example, “problem-solving” as a skill is interpreted quite differently in different occupational sectors. Although O*NET appears to be well-vetted for its purposes, it will most certainly be aligned to a degree of status-quo thinking in labour markets and labour market sectors; that is, it reflects orthodoxies. In other words, while on the surface this finding would seem to uphold the tenets of human capital theory, it leaves out much information.

Finally, gender differences are observed across the board. As before, some of these differences constitute labour market structures because they are entrenched and have proven difficult to dislodge. For example, higher skills in the workforce for females with undergraduate degrees in health sciences may be readily accounted for by the predominance of women in the large occupational field of nursing, while low use of technical and maintenance skills by females across educational levels reflects their lower presence in trades and STEM related fields.


Overall, the report findings suggest that educational credentials, skills used in employment, and remuneration are reasonably harmonious. There is an overall pattern of matching higher education with higher skilled occupations. This finding is not surprising, making it easy to read as a confirmation of a broad tenet of human capital theory — namely that the labour market fairly reflects some inherent measure of value in a credential. This is the line of thinking that informs the thinking of commentators who are quick to diminish the value of those skills and credentials that do not have ready transaction values in the labour market, and then to call out universities for failing to adequately prepare grads.

It is important to note, however, that Frenette and Frank do not speculate on the causal relationships between credentials and outcomes; they simply offer useful criteria, in the form of the nine skills clusters, for finer interpretation. Usher’s brief analysis does not question the structure of the labour markets from which the skills definitions used in this study emerge, but there are hints in the study to suggest that these are very important. Nursing and engineering, for example, are two large, well-established occupational fields with clearly defined pathways; they also consistently provide full employment and good remuneration. Highly positioned skills may reflect the stability, size and occupational closure of occupations that use these skills, rather than clear cut demand in the labour market for the skills themselves. Similarly, indicators of gendered skill sets may be said to reflect gendered occupations and sectors in labour markets over which education has had little impact over time.

It is more useful to consider the educational outcomes measured in this study — that is level of education and field of education — from critical perspectives that do not take labour market structures as neutral or given variables against which the value of skills are measured. Indeed, this study cannot speak to the labour value of many skills because it excludes significant portions of young workers. Graduates working in non-standard jobs are not included in the sample; thus significant forms of labour force participation, including the skills these use, are left out of the analysis. As young adults turn or are forced increasingly into part-time work, contracts, entrepreneurship, and other forms of precarious labour, the exclusive alignment and recognition of skills only within the contexts of standard labour must be seriously questioned.

Summarily, the Frenette and Frank study is a very useful methodological contribution to the study of the relationships between qualifications and credentials, but significant questions about the meanings of these relationships cannot be answered with what is only a partial representation of graduate employment and employability.

Contributor: Laura Servage


Frenette, M., & Frank, K. (2017). Do postsecondary graduates land high-skilled jobs? Statistics Canada. Analytical Studies Branch Research Paper Series. Retrieved from

Marginson, S. (2016). The worldwide trend to high participation higher education: dynamics of social stratification in inclusive systems. Higher Education, 1–22.

Trow, M. (2007). Reflections on the transition from elite to mass to universal access: Forms and phases of higher education in modern societies since WWII. In J. Forest & P. Altbach (Eds.), International Handbook of Higher Education (pp. 243–280). Dordrecht: Springer.

Usher, A. (2017, January 26). An amazing StatsCan skills study. [Blog.] Retrieved from

End Notes

[1] Trow (2007) discusses the transition during the second half of the 20th century from elite higher education restricted to 15% of the university age-going population to mass higher education where up to half of the relevant population attends university. A universal system is one in which over half of the population pursues higher education.

[2] Fields of study were drawn from the Classification of Instructional Programs (CIP, 2011).

[3] Similar patterns are noted in college education. The authors find that male math, computer science and information sciences graduates hold occupations that require the highest skill levels in all areas. However, they also find that the "soft" fields (humanities, social sciences, arts and communication) "fare somewhat better in a relative sense among college graduates" (p. 24).

[4] As the authors observe, "Graduates from a given discipline may possess very high skills (e.g., in reading comprehension); however, this does not guarantee that they will find jobs that require them to fully utilize these skills" (p. 14).