University graduates with lower levels of literacy and numeracy skills

by Darcy Hango

[Release from The Daily] [Full article in PDF]

Skip to text

Text begins

Start of text box

Overview of the study

This article examines the share of adults aged 25 to 65 with a university degree who were in the lower range for literacy skills, numeracy skills, or both, and the factors most likely to be associated with lower levels of literacy or numeracy among university graduates. In this article, individuals in the lower range for literacy and numeracy are defined as those who scored at level 2 or below (out of 5 levels) in tests administered to survey respondents who participated in the Programme for the International Assessment of Adult Competencies (PIAAC).

  • Of all Canadian adults aged 25 to 65 in 2012, 49% were in the lower range for literacy proficiency, 55% were in the lower range for numeracy proficiency, and 43% were in the lower range for both literacy and numeracy.
  • University graduates were less likely than other educational attainment groups to be in the lower range for skills. For example, 27% of university graduates were in the lower range for literacy skills, compared with 88% among those with less than a high school diploma.
  • Foreign-born university graduates were more likely to be in the lower range of literacy and/or numeracy. For instance, 45% of them were in the lower range for literacy, compared with 16% of Canadian-born university graduates.
  • Among the Canadian-born university graduates, the proportion of those in the lower range for literacy and numeracy varied across various factors, such as age, field of study, and number of books at age 16 (used as a proxy for cultural capital).
  • Canadian-born university graduates who were classified in the lowest skill proficiency levels were as likely to be employed as those who were in the highest levels, but those who were employed were less likely to work as professionals or managers.

End of text box

Introduction

Higher education is typically associated with greater benefits, especially in the labour market, both in the short-term and over longer term.Note 1 For instance, a recent study based on a cohort of individuals in the early stages of their career in 1991 suggested that Canadian men with a bachelor’s degree earned $728,000 more on average than their counterparts with a high school diploma over a 20-year period. For women, the difference between high and low education levels was less, but still substantial at about $442,000.Note 2 Meanwhile, other research, measuring proficiency skills more directly, suggests that factors such as literacy may also have a substantial impact on earnings.Note 3 Other studies found that over one-half of the education effect on earnings can be explained via cognitive factors (such as literacy).Note 4 More recent evidence also finds that higher cognitive skills, such as literacy and numeracy, raise earnings—an effect present in over 20 countries.Note 5

While it is well-known that education is closely tied to the labour market, it is also increasingly being discovered that cognitive skills, separately and in combination with education, are important determinants of labour market success.Note 6 Therefore, knowing the level of skills, such as literacy and numeracy, in conjunction with level of education, can contribute to our understanding of labour market success. To this end, this study will examine those who graduated from university, yet were in the lower range of literacy and/or numeracy scores. Recent results from the Programme for the International Assessment of Adult Competencies (PIAAC) suggest that even though the university-educated have higher literacy and numeracy skill levels than their less-educated counterparts, a significant proportion of university graduates in Canada are at level 2 or below in terms of literacy or numeracy proficiency. In 2012, around 27% of university graduates aged 25 to 65 were in the 2nd proficiency level or below for literacy, while the figure was 31% for numeracy.Note 7 

Those who are highly educated but find themselves at the bottom of the skills distribution represents individuals who, beyond the required credentials, may not have the necessary skill set to succeed in the labour market. As university degree-holders in Canada continue to increase in numbers, employers may value other, perhaps less tangible, qualities, such as literacy or numeracy skills. Thus, differences in literacy and numeracy skills could help explain why individuals with a similar educational background sometimes differ in their labour market outcomes.

In this article, new data from the Programme for the International Assessment of Adult Competencies (PIAAC) are used.Note 8 PIAAC is a large survey that collected information on the information processing skills of youth and adults between the ages of 16 and 65 in Canada and 23 other countries in 2011 and 2012 (See Data sources, methods and definitions). This article begins by providing descriptive statistics about the proportion of individuals who were in the lower range for literacy, numeracy, or both. The second part of this article examines the characteristics of university graduates with lower levels of skills, with a focus on the Canadian-born population. The third section of this paper examines the extent to which lower levels of skills could be linked to lower employment outcomes among university graduates.

Higher levels of literacy and numeracy skills for university graduates

In PIAAC, each survey respondent had to answer questions aimed at measuring their information processing skills, and obtained scores ranging from 0 to 500. In the case of literacy, respondents were measured for their ability to engage with written texts (print-based and digital). In the case of numeracy, respondents were measured for their ability to engage with mathematical information “in order to manage the mathematical demands of a range of situations in everyday life”.Note 9 On the basis of such scores, respondents can be classified across levels (ranging from below level 1 to level 5). The higher the level, the more respondents are capable of processing more complex information and understanding complex representations.

Since the focus of this study is on university graduates, who typically have higher average levels of literacy and numeracy, it is appropriate to qualify individuals who are at level 2 or below as those who are in the “lower range” for literacy or numeracy skills. University degree holders below level 3 may not have mastered the minimum foundation of literacy needed to attain higher levels of performance. Conversely, those at level 3 and above generally have positive economic, social and educational outcomes, and thus may be better equipped to occupy professional or managerial jobs that normally require a university education.Note 10

With respect to literacy, individuals at level 2 or below are less likely to undertake tasks that involve the integration of information across multiple sources and more likely only to have the ability to undertake tasks of limited complexity such as locating single pieces of information in short sections of text.

For numeracy, individuals at level 2 or below are less likely to be able to perform and understand complex mathematical information and work with mathematical models. They are also less likely to use problem solving strategies, and more likely only to have the ability to perform simpler mathematical operations (See Data sources, methods and definitions for a description of each level).Note 11

Across the entire population aged 25 to 65, 49% had a literacy score at level 2 or below, while 55% had a numeracy score at level 2 or below (Chart 1). About 43% of Canadians aged 25 to 65 scored at level 2 or below on both literacy and numeracy.

Chart 1 for the Insights on Canadian Society article number 14094

Description for Chart 1

As might be expected, the level of education is positively correlated with skills.Note 12 Not only do individuals gain more skills and enhance their competencies while in an educational program, these programs are also selective of skills and competencies in the first place. For example, recent work has highlighted the link between greater reading proficiency at age 15 and university attendance by age 21.Note 13 Furthermore, university-educated individuals are more likely to work in occupations that require the use of more complex skills.Note 14

This same relationship was found in PIAAC as well. For example, 27% of PIAAC respondents aged 25 to 65 with a university degree had literacy skills in the lower range (level 2 or below). This increased to 48% for those with a non-bachelor PSE diploma, to 60% for those with a high school diploma and to 88% for those with less than a high school diploma. A similar trend across educational attainment categories was found for numeracy, although the proportions (at level 2 or below) were higher in each category (from 32% among university graduates to 91% among those with less than a high school education). Lastly, the proportion of individuals aged 25 to 65 who were in the lowest levels of literacy and numeracy also varied across the education gradient, from 22% among university graduates to 85% among those with less than a high school diploma.

University-educated immigrants are more likely to be in the lower range of literacy and numeracy

Because a relatively large proportion of university graduates in Canada are foreign-born (about 4 in 10), the scores obtained by immigrants may influence the results of university graduates as a whole. It is also noteworthy that PIAAC tests were administered in French or English to all survey respondents (including immigrants), meaning their results could be influenced by their proficiency in the test language (81% of foreign-born university graduates had a native tongue other than English or French).

Among university graduates, about 45% of the foreign-born were in the lower range for literacy (level 2 or below), while 16% of the Canadian-born were in the same situation (Chart 2). The results for numeracy were similar: 46% of immigrants had numeracy in the lower range (level 2 or below), while the figure was 23% among the Canadian-born. Furthermore, 36% of foreign-born university graduates had both numeracy and literacy in the lower range, compared with 12% of the Canadian-born university population. The differences between immigrants and the Canadian-born, however, were much reduced when looking only at immigrants who completed their studies in Canada (see Results by location of study and immigration status).

Chart 2 for the Insights on Canadian Society article number 14094

Description for Chart 2

Because PIAAC has been administered in other countries, some international comparisons are possible, notably with the United States. Comparisons indicate that in both countries, the shares of university graduates who were in the lower range for numeracy and literacy skills were relatively similar (Chart 3). Of all university graduates in the United States, 22% were in the lower range for literacy (compared with 27% in Canada), 31% were in the lower range for numeracy (32% in Canada) and 18% were in the lower range for both (22% in Canada). The gap between immigrants and the native-born, however, was smaller in the United States—the result of relatively higher rates among those who were born in the United States (compared to those who were born in Canada) and relatively lower rates among immigrants in the United States (compared to their Canadian counterparts).

Chart 3 for the Insights on Canadian Society article number 14094

Description for Chart 3

Understanding why immigrants score lower in numeracy and literacy would require an examination of, among other factors, language, ethnic origin, and location of study. In the near future, results for immigrants (and the Aboriginal population) will be examined more extensively in other thematic reports. In the rest of the present analysis, only Canadian-born respondents are included, since the factors outlined above may not be as relevant in explaining why some Canadian-born university graduates have lower levels of literacy and numeracy.

Characteristics associated with lower levels of literacy and numeracy among Canadian-born university graduates

As previously observed, a significant portion of the Canadian-born university-educated scored at level 2 or lower in PIAAC tests (16% in literacy and 23% in numeracy). This section examines the characteristics associated with the Canadian-born university graduates with lower levels of literacy or numeracy. Personal characteristics are examined, as well as family background variables and educational background variables (Table 1).

Table 1
Proportion of Canadian-born university graduates aged 25 to 65 at level 2 or below in skill proficiency levels across various socioeconomic characteristics
Table summary
This table displays the results of Proportion of Canadian-born university graduates aged 25 to 65 at level 2 or below in skill proficiency levels across various socioeconomic characteristics Level 2 or below, literacy and Level 2 or below, numeracy, calculated using percentage units of measure (appearing as column headers).
  Level 2 or below, literacy Level 2 or below, numeracy
percentage
Gender  
Men (ref.) 13.6 16.8
Women 16.2 26.8Note *
Age group  
25 to 29 12.3 19.5
30 to 34 11.3 17.4
35 to 39 (ref.) 8.6 17.4
40 to 44 13.7 21.2
45 to 49 14.2 22.0
50 to 54 17.9Note * 26.8
55 to 59 23.8Note * 29.0Note *
60 to 65 23.9Note * 29.3Note *
First language spoken (mother tongue)  
English (ref.) 12.7 20.6
French 18.5Note * 24.0
Other 23.8Note * 30.3
Province of residence  
Newfoundland and Labrador 15.7 23.6
Prince Edward Island 15.1 24.5
Nova Scotia 15.1 22.6
New Brunswick 17.6 27.5
Quebec 19.4Note * 25.2
Ontario (ref.) 13.0 20.0
Manitoba 17.3 28.6
Saskatchewan 13.1 20.3
Alberta 13.0 20.4
British Columbia 12.3 21.0
Parental education  
Less than a high school diploma (ref.) 26.6 34.7
At least one has a high school diploma 18.8 26.8
At least one has a PSE below bachelor 11.7Note * 18.7Note *
At least one has a university degree 11.5Note * 18.0Note *
Number of books in home at age 16  
10 or less (ref.) 30.9 39.1
11 to 25 20.5 30.5
26 to 100 17.0 24.0Note *
101 to 200 14.4Note * 20.8Note *
More than 200 8.6Note * 15.4Note *
Field of study  
STEM (ref.) 9.4 11.5
Humanities, languages and arts 17.9Note * 29.4Note *
Social science, business and law 13.1 20.9
Teacher training and education science 22.2Note * 29.1Note *
Health and welfare 15.5 26.5Note *
Educational attainment  
Bachelor's degree (ref.) 16.0 23.4
1st professional degree (medical, veterinary medical, dental, optometry, law, divinity) 13.1 19.3
Masters or Doctorate 13.2 19.9

Personal characteristics

Some personal characteristics that may be related to lower levels of literacy and numeracy among Canadian-born university graduates include gender, age, language, and province.

First, the proportion of those who had a level 2 or below in numeracy was higher among women (27%, compared with 17% among men).Note 15 However, similar proportions of men and women had a level 2 or below in literacy (14% for men and 16% for women).Note 16

Differences could also be found across age groups, even among university graduates.Note 17 The proportions of those at level 2 or less were lowest among those aged 35 to 39, (with percentages of 9% for literacy and 17% for numeracy). Conversely, the highest percentages were found among those aged 55 and over—about one-quarter of this population had literacy in the lower range and nearly one-third had numeracy in the lower range.

Since this analysis is restricted to the university-educated Canadian-born, the majority had either English (66%) or French (27%) as their mother tongue, while the remaining 7% had another first language. Of these individuals with a mother tongue other than English or French, three-quarters had parents who were born outside Canada; that is to say second-generation Canadians. In this population, 24% were in the lower range for literacy, compared with 13% among those who reported English as their mother tongue. At 19%, individuals with French as their mother tongue also had a significantly higher rate than individuals whose mother tongue was English. The differences in the results for numeracy, however, were not significant between the three groups.

The pan-Canadian PIAAC report had found that, among all people in Canada aged 16 to 65, average levels of literacy and numeracy were highest in Alberta and lowest in Newfoundland and Labrador.Note 18 However, with respect to the probability of having literacy or numeracy in the lower range for the Canadian-born university-educated, there was little variation across provinces. The one exception was Quebec, which had a higher proportion of university graduates in the lower literacy range than Ontario (19% versus 13%).

Family background

As indicated above, the acquisition of numeracy and literacy skills is a complex process, resulting from the interaction of complex life experiences, including work experiences, education decisions, individual abilities and family background. The educational attainment of parents is one aspect of family background collected in PIAAC.

The links between parental education and that of their children are well-founded.Note 19 Better-educated parents may have higher literacy and numeracy skills to pass on to their children. The benefits may also be witnessed among university graduates: using numeracy as an example, more than one-third of university graduates whose parents had less than a high school education were at level 2 or below in numeracy skills, compared with less than one-fifth among those whose parents were university graduates. Similar results were found for literacy.

Another important family background characteristic is access to cultural capital at a young age. It is believed that exposure to various sources of intellectual stimulation in childhood has a positive impact on academic outcomes.Note 20 In PIAAC, these sources of cultural capital can be proxied with the estimated number of books in the home when respondents were in high school. The availability of reading material can be beneficial for developing and building skills, especially reading skills, as youth enter adulthood. Even among the Canadian-born and university-educated, not everyone had access to a critical mass of books at home when they were 16. About 32% of Canadian-born university graduates aged 25 to 65 said they had over 200 books in their home when they were 16, while 6% had access to 10 books or less.

As expected, Canadian-born university graduates who had 10 books or less at home at age 16 were more likely to be in the lower range for literacy (31%). This proportion declined over successive groups, to 9% for university graduates with over 200 books. The same general relationship was found with respect to numeracy: the proportion who were in the lower range among those who had 10 books or less was 39%, a rate that was significantly higher than those who had 101 to 200 books (21%) and those who had more than 200 books (15%).

Educational background

It may be that individuals gain more skills in some specific educational programs, and/or that students with better skills may be concentrated in certain types of programs. As a result, the type of educational experience may also be a determinant of literacy and numeracy skills among adults. The results are compared across five types of instructional programs.

For both literacy and numeracy, the lowest proportions of those at level 2 or below were found among those who graduated from science, technology, engineering, mathematics, and computer science (STEM) programs (9% for literacy and 12% for numeracy). With respect to numeracy, differences between STEM graduates and other types of instructional programs could be expected, since most STEM programs typically involve more complex mathematical techniques, and also because STEM graduates may be more likely to work in jobs allowing them to maintain such skills. But even for literacy, the proportion for those who graduated from STEM programs was significantly lower than graduates from teacher training and education science programs (22%), and graduates from humanities, languages and arts programs (18%). This finding could perhaps highlight the more selective nature of STEM program participation, especially as it pertains to skills requiring greater mathematical ability.Note 21

Lastly, no significant differences were found between graduates with a bachelor degree and those with a masters or doctorate, or with those in professional degree programs (such as medical, dental or optometry programs).

Some of the variables discussed above are related. For example, parental education is strongly correlated with the number of books at home at age 16. However, when the variables are considered together in logistic regression models (with a dependent variable equal to 1 if the respondent is at level 2 or below, and 0 otherwise), differences by gender, age group, number of books at home at age 16, and field of study remained significant. Conversely, differences by first language spoken and parental education lost their significance. In the latter case, this is likely because of the close association between parental education and other control variables, particularly the number of books at home at age 16, and age (i.e., older respondents are less likely to have university-educated parents).

Employment outcomes of university graduates in lower skill levels

In this section, the links between lower skill levels and employment outcomes are examined among Canadian-born adults aged 25 to 65 with a university degree. Specifically, two questions are examined: (1) are university graduates with lower levels of numeracy or literacy less likely to be employed than those at higher skill levels?; and (2) among those who are employed, are those at lower literacy and numeracy skill levels less likely to employed in high-skilled occupations?

Both questions can be answered by using multivariate models. In the employment models, the dependent variable is equal to 1 if the respondent is employed and equal to 0 otherwise. Three sets of explanatory variables are then applied. In the first set, only one variable is included, indicating whether the respondent has a level 2 or below for literacy (or numeracy). In subsequent models, demographic and education-related variables are included to account for the other factors (other than skills) that may influence the probability of employment. Separate models are applied for literacy and numeracy.Note 22

Results from the first model indicate that having a level 2 or below for literacy was associated with a 10 percentage-point reduction in the probability of being employed, compared with those who were at level 3 or higher (Table 2). Put differently, university graduates with a level 2 or below had a 79% probability of being employed, compared with 89% among those with a level 3 or higher. The results for numeracy were similar, as those with a level 2 or below had an 8 percentage-point reduction in the probability of employment (or 81%, compared with 89% for those with a level 3 or higher).

Table 2
Relationship between skill levels and the probability of employment, all Canadian-born university graduates aged 25 to 65
Table summary
This table displays the results of Relationship between skill levels and the probability of employment Model 1, Model 2 and Model 3 (appearing as column headers).
  Model 1 Model 2 Model 3
Controls for skill levels only Add demographic factors Add education-related factors
Marginal effects of being employedNote 1 of table 21 percentage point
Literacy (ref.: level 3 or above)  
Level 2 or below -9.8Note * for table 2* -4.1 -3.5
Predicted probabilities percentage
Level 3 or above 88.9 91.3 91.4
Level 2 or below 79.0 87.2 87.9
Numeracy (ref.: level 3 or above) percentage point
Level 2 or below -8.0Note * for table 2* -3.9 -3.2
Predicted probabilities percentage
Level 3 or above 89.2 91.5 91.6
Level 2 or below 81.2 87.7 88.4

However, when other demographic or educational variables were included in the model, the difference between lower- and higher-skilled individuals shrank to the point of becoming non-significant. This suggests that, as far as employment is concerned, lower levels of literacy and numeracy would have little influence on the probability of being employed—at least among Canadian-born university graduates.

Similar models can be used within the employed population to determine whether skill levels are associated with a lower probability of being employed in occupations typically requiring a university degree ('professional' occupations) or in management occupations. In 2012, 85% of Canadian-born university graduates who had a job were employed in such occupations.Note 23

Results from the first model (with only the skills variable as a covariate) indicate that among university graduates who were employed in 2012, those with a level 2 or below in literacy or numeracy were about 8 or 10 percentage points, respectively, less likely than those with higher skills to be employed in a professional or managerial occupation (Table 3). When these results are translated into proportions, 82% of employed university graduates with a level 2 or below in literacy were employed in a professional or managerial occupation, compared with 90% among employed university graduates who had a level 3 or above. For numeracy, the results were 82% among those with a level 2 or below and 91% among those with a level 3 or above.

Table 3
Relationship between skill levels and the probability of employment in professional and managerial occupations, employed Canadian-born university graduates aged 25 to 65
Table summary
This table displays the results of Relationship between skill levels and the probability of employment in professional and managerial occupations Model 1, Model 2 and Model 3 (appearing as column headers).
  Model 1 Model 2 Model 3
Controls for skill levels only Add demographic factors Add education-related factors
Marginal effects of being employed in skilled occupationsNote 1 of table 31 percentage point
Literacy (ref.: level 3 or above)  
Level 2 or below -8.4Note * for table 3* -8.3Note * for table 3* -7.2
Predicted probabilities percentage
Level 3 or above 90.5 91.5 93.2
Level 2 or below 82.0 83.2 85.9
Numeracy (ref.: level 3 or above) percentage point
Level 2 or below -9.6Note * for table 3* -9.5Note * for table 3* -7.5Note * for table 3*
Predicted probabilities percentage
Level 3 or above 91.3 92.3 93.7
Level 2 or below 81.7 82.8 86.2

However, and somewhat contrary to the employment results, the effect of skill does not entirely disappear once other covariates are included in the second or third model (at least in the case of numeracy). Using the results of the third model as an example, university graduates who had a level 2 or below for numeracy had a probability of employment in skilled occupations that was about 8 percentage points lower than individuals with skills at level 3 or higher, even after taking other demographic and educational variables into account. The results for literacy however, indicate that once educational factors are considered in Model 3, the impact of literacy skills on the probability of working in a skilled occupation becomes non-significant.

Such results indicate that university graduates of different skill levels are just as likely to be employed, but not necessarily as likely to occupy the same type of work because those in the lower range of skills (at least in the case of numeracy) are significantly less likely to work in managerial and professional occupations than their higher-skilled counterparts.Note 24 These results reflect associations between variables, and not necessarily cause-and-effect relationships. In other words, the fact that employed individuals with lower skill levels are less likely to be employed in skilled occupations does not necessarily mean that such individuals had lower skills in the first place. The type of job may also have an impact on skills; for example, if the skills normally acquired at university are not subsequently used by workers on a regular basis, a certain degree of skill loss may occur. More research will be needed to understand the dynamics between lower skills and low-skilled occupations.

Conclusion

Skills such as literacy and numeracy are important factors for the successful economic integration of university graduates, therefore knowing whether graduates with similar levels of education have varying degrees of success in the labour market becomes an important area of concern not only for potential employees, but also for employers and society in general. In 2012, about 16% of Canadian-born university graduates aged 25 to 65 scored at level 2 or below for literacy (out of five categories determined by PIAAC) and 23% scored at level 2 or below for numeracy.

With regard to the characteristics most associated with lower skill levels among Canadian-born university graduates, three findings are of particular significance: older individuals were more likely to have lower levels of skills than their younger counterparts; the field of study of university graduates was an important factor, since graduates from STEM fields had higher levels of skills; and the greater the number of books in the home when the respondent was age 16, the less likely he or she was to end up with a lower level of skill as an adult. These findings suggest that those who benefit from a higher level of cultural capital during their youth may end up having higher literacy and numeracy skills, even within the population of university graduates.

Even though literacy and numeracy skill levels did not seem to affect the employment probability of university graduates, it did have a relationship with the type of occupations held by those who were employed. More precisely, those with lower skill levels (especially numeracy) were at least 8 percentage points less likely to be employed in managerial and professional occupations, even after accounting for other demographic and educational factors. Such results do not provide clear answers as to why some university graduates end up having lower skills as adults, since proficiency scores reflect a vast array of education, work, and personal experiences. They may, however, help explain why some university graduates are less well-matched with their current occupation for their given level of education.

Darcy Hango is a researcher in the Centre for Education Statistics of Statistics Canada.

Notes


Related material for this article

Date modified: