Objectives To examine the association between education-to-job mismatch and work injury.
Methods Cross-sectional data from the 2003 and 2005 Canadian Community Health Surveys (n=63 462) were used to examine the relationship between having an educational level that is incongruent with occupational skills requirements and the risk of sustaining a work injury requiring medical attention or a work-related repetitive movement injury (RMI). The effect on injury of the interaction of overeducation with recent immigrant status was also examined. Models were stratified by sex and adjusted for possible confounders. Occupational physical demands were conceptualised as a potential mediating variable.
Results After adjustment for covariates, over-education was associated with work injury and RMI for both sexes. Adjustment for occupational demands attenuated the impact on work injury but did not eliminate the effect on RMI among men. The interaction of over-education and recent immigrant status resulted among men in a fourfold increase in the odds of work injury compared with non-recent immigrants who were not over-educated. After adjustment for occupational demands, over-educated recent immigrant men still had more than a twofold increase in the odds of injury.
Conclusions The risk of sustaining a work injury is higher among those whose education exceeds that of job requirements. These findings highlight the need to address barriers to suitable employment, particularly among recent immigrants.
- work injuries
- public health
- occupational, gender
Statistics from Altmetric.com
Education-to-job mismatch occurs when a worker's level and/or type of education does not correspond to the requirements of their job. Research in this area has almost exclusively focused on overeducation,i a situation whereby education level surpasses that of job requirements. In Canada in 2008, one in four workers aged 25–54 years was over-educated.1 However, the prevalence has been shown to vary across the labour market and according to sociodemographic characteristics. It is higher among young workers and immigrants,1 2 and in certain fields where the rise in higher education degrees has not been matched with concomitant increases in skilled jobs.3 4 The reverse situation—having a lower level of education than is required by one's occupation—tends to be somewhat less common, although few Canadian data exist.4
The impact of over-education on workers has traditionally been looked at by examining economic returns, occupational mobility and job satisfaction. In recent years, an increasing number of studies have examined the implications for workers' health, with many focusing on immigrant populations. Objective and subjective measures of over-education have been associated with poor general mental and physical health5–8 and with chronic diseases9 10 after adjustment for a range of factors including age, gender, educational level, occupation, income, period of immigration, ethnicity and language. Others have shown an impact on health for workers' spouses and children because of stress and material and social deprivation.8 11 On the other hand, having a lower educational level than required has been linked to cardiovascular disease among men.12
While a study by Kerr and colleagues found that feeling over-educated relative to others in a similar job was associated with reports of work-related back pain,13 to our knowledge the effect of education-to-job mismatch on the risk of work injury has not been examined. Previous studies have found that formal education is inversely associated with work injury14 15 even after adjustment for job characteristics.16–18 Education may have a protective effect on work injury risk through representation in less dangerous tasks, improved training opportunities, better knowledge acquisition and problem-solving skills, or safer attitudes and behaviours towards risks. However, it is unclear whether education offers a protective effect on work injury risk when working in an occupation for which one is over-educated. While some have suggested that over-education may decrease the likelihood of work injury,19 others have proposed that the risk may be higher because various determinants of work injuries may be similarly associated with over-education. Thus, over-educated workers may be disproportionately found in dangerous work, lack knowledge and experience that are relevant to their job,20 be less likely to receive training21–23 and lack colleague or supervisor support.24
This study therefore seeks to examine the association between education-to-job mismatch and the likelihood of sustaining a work injury. We hypothesise (1) that over-education will be associated with a higher risk of injury than when no mismatch is present; (2) that the impact of over-education will be greatest among labour market participants with the least Canadian experience—namely, recent immigrants; and (3) that occupational physical demands are a mediator in the relationship between mismatch and work injuries.
This study used secondary data from the 2003 and 2005 Canadian Community Health Surveys (CCHS). Each CCHS uses a multistaged, stratified sampling frame to select a sample of households from which to target individuals aged ≥12 years, excluding people living on Indian reserves or Crown lands, residents of institutions, full-time members of the Canadian armed forces and residents of certain remote regions. The sample of households was selected using an area frame, a list frame of telephone numbers and a random digit dialling sampling frame. The household response rate to the 2003 CCHS was 87%, with the selected person response rate being 93%.25 For the 2005 CCHS the household response rate was 85%, with the selected person response rate being 94%.26 To remove barriers in conducting interviews, Statistics Canada recruited interviewers with competencies in 23 different languages. When necessary, the interviewers conducted the interview in the language of the respondent while completing the questionnaire in one of the two official languages (English or French). Questionnaires were also translated into Chinese, Punjabi and Inuktitut.25 26
For the purpose of this paper we restricted both samples to those respondents who were aged ≥25 years and who had been employed at some point in the previous 12 months (N=50 557 respondents to the 2003 cycle; N=12 905 respondents to the 2005 cycle). Eighty per cent of our sample comes from the 2003 CCHS since only a random fourth subsample of the 2005 survey was asked information about occupation. Approval for the secondary data analyses was obtained through the University of Toronto, Health Sciences I ethics committee.
Main dependent variables
Respondents were asked if they had been injured, not including repetitive movement injuries (RMIs), in the past 12 months seriously enough to limit their normal activities. Respondents who reported an activity-limiting injury were further asked if the injury occurred at work and if they had received medical attention for their injury within 48 h. Using this information, we classified respondents as having a work-related injury requiring medical attention (yes/no).
Work-related repetitive movement injury
Respondents were also asked if they had had any injuries in the past 12 months due to repetitive strain which were serious enough to limit their normal activities. The examples of carpal tunnel syndrome, tennis elbow and tendinitis were given. Respondents who answered yes to this question were further asked if they were working at a job or business when this repetitive strain injury occurred. Using this information, we classified respondents as having a work-related RMI (yes/no).
Main independent variable: education-to-occupation mismatch
Each respondent was asked their highest level of education completed. This information was used to classify respondents into the following four categories: no secondary education completed; secondary education completed; post-secondary education completed below bachelor's level; bachelor's level education or higher.
The current occupation for each respondent was classified according to the National Occupational Classification (NOC) system.27 Using the second digit of the NOC we classified each respondent's occupational skill level requirements into the following four groups: those requiring no formal education (eg, cashiers); those requiring secondary school education and/or occupation-specific training (eg, sales representatives, sales clerks); those requiring college education, or apprenticeship training (eg, retail trade supervisors) and those requiring university education at the bachelor's level or higher (eg, engineers). The NOC classification has previously been used in the determination of over-education.5
Using the categories for education and occupation, we classified respondents into the following three categories: where educational level was below occupational skill requirements; where educational level was congruent with occupational skill requirements and where educational level was above occupational skill requirements.
Potential confounding variables
Other variables included in our models as confounders between educational and occupational match and work injury included year of survey; hours of work; age (grouped); the length of time the respondent had been in Canada; if the respondent identified themselves as a visible minority; education level; province; if the respondent lived in an urban or rural area and if they were interviewed in English, French or another language. These variables were included in our models since they have previously been linked with work injuries.
Potential mediating variable: occupational physical demands
The physical demands of each respondent's occupation were classified using a system developed by the Occupational Health and Safety Research Institute Robert-Sauvé.28 This classification system groups NOC codes into three categories of physical demands:
Manual—Occupations requiring regular handling of heavy or average loads, or occupations handling lighter loads, but in static postures, or occupations requiring continuous repetitive work;
Mixed—Occupations requiring handling of light loads, but that do not require continuous static postures or occupations that occasionally handle heavy or average loads;
Non-manual—Occupations rarely requiring the handling of loads or physical activity.
Previous research has demonstrated a graded increase in injury risk across mixed and manual occupations (compared with non-manual occupations) using both self-reported and administrative injury and absence outcomes.18
The combined sample of respondents who were over the age of 25 and had been employed in the previous 12 months was 66 075 respondents. Of this sample 194 respondents (0.3%) were missing information on injuries in the previous 12 months and 2163 (3.3%) were missing information on either occupation, work hours in the previous 12 months or educational status, leaving a sample of 63 718. Respondents with missing information were more likely to be female, older and be recruited using the telephone frame of the CCHS, rather than the area frame. Of this sample an additional 256 respondents (0.4%) were missing information on one of the covariates of interest, leaving a final sample of 63 462 respondents (96% of the original sample).
A preliminary descriptive analysis examined the incidence of work injuries across levels of education and occupation mismatch categories. A series of logistic regression models then examined the probability of injury for respondents with educational mismatch (separately for respondents with education below and above occupational requirements) compared with the reference category of respondents with matching education and occupation skill requirements. Initial models only included the year of the survey and the number of hours worked in the past 12 months, which was included in the logistic analysis as an offset.29 A second model adjusted for all other potential confounding variables. A final model examined changes in probability estimates with the inclusion of occupational physical demands, as a mediator in the relationship between education and occupational mismatch and work injuries. We further examined whether the effects of education/occupation mismatch on work injury differed across immigration periods or occupational demands, by including a multiplicative interaction term in the logistic model. Because work tasks and work injuries differ between men and women, even within similar occupations,30 we stratified all logistic models by sex. To account for the complex sample design of the CCHS, in line with guidelines from Statistics Canada, the CIs around each point estimate have been adjusted using a bootstrap technique.31
Table 1 presents descriptive information from the study sample, stratified by sex. A greater proportion of male respondents (3.6%) reported a work injury that required medical attention than female subjects (1.5%). Conversely, female respondents had a marginally higher incidence of RMIs (8.7%) than men (7.4%). Female respondents were also more likely to have education levels above that which was required for their occupation than men (33.2% vs 26.6%).
Table 2 presents the incidence of injury types across education and occupation match groups for men and women. Among male respondents no statistically significant differences in work-related injuries requiring medical attention or repetitive movement injuries were observed across education/occupation match groups. Differences were more pronounced in the female sample for both types of injuries across education/occupation match categories. In each case the incidence of injury was lowest among respondents whose education level matched their occupational skill requirements. For both work-related injuries and repetitive movement injuries the percentage of injuries was statistically higher among over-educated women than among those whose educational levels were matched to their occupational levels (p<0.05).
Table 3 presents ORs and 95% CIs for different injury types across education/occupation match categories, after adjustment for hours of work and survey year; additional adjustment for potential confounders and final adjustment to assess the mediation of the relationship between education/occupation match and work injury via occupational physical demands (ORs and 95% CIs for other variables included in our final models for work injuries and repetitive movement injuries are presented in online appendix 1). Over-education was associated with both work injury and RMI among men and women after adjustment for other covariates. The relationship between over-education and work injury was stronger among women (OR=1.79, 1.30 to 2.47) compared with men (OR=1.34, 1.09 to 1.64). On the other hand, having an education level that is lower than that of the job requirements was associated with a lower risk of work injury among men and of RMI for both sexes. Adjustment for occupational physical demands attenuated the impact on work injury but did not eliminate the effect of mismatch on RMI among men.
Table 4 presents the interaction between over-education and recent immigration status (≤5 years in Canada) with work injuries requiring medical attention in the male sample. The relationship between over-education and work injury differed among respondents who were recent immigrants compared with those who were not. Among men, recent immigrants who were over-educated were five times more likely than non-recent immigrants who were not over-educated to sustain a work injury requiring medical attention (OR=5.02, 2.12 to 11.86). Adjustment for occupational physical demands attenuated this risk to a large extent; however, even after adjustment male recent immigrants who were over-educated were more than three times more likely to sustain a work-related injury. No interactions were found between immigration period and over-education in the female sample, or between over-education and occupational demands in either men or women.
This study sought to examine the association between education-to-job mismatch and the likelihood of sustaining a work injury. After adjustment for covariates, over-education was associated with both work injury and RMI among men and women. Adjustment for occupational demands attenuated the impact on work injury, suggesting that the physical demands of the job constitute an important pathway from mismatch to injury. However, adjustment for occupational demands did not eliminate the effect of mismatch on RMI among men. In addition, we found that the effect of over-education on work injuries requiring medical attention differed by immigration period among men. The interaction of over-education and recent immigrant status in the male sample resulted in a fourfold increase in the odds of injury requiring medical attention compared with non-recent immigrants who were not over-educated. Even after adjustment for occupational demands, over-educated recent immigrant men still had more than a twofold increase in the odds of sustaining a work injury requiring medical attention. No interactions were found between immigration period and over-education among women.
The results of our study have important implications for high-income countries where increasing levels of education at the population level have not been matched with concomitant increases in the number of skilled occupations across all fields. Although previous studies have shown formal education to have a protective effect on the risk of work injury, based on our results we suggest that working in a job for which one is over-educated poses a risk for injury, particularly for men who are recent immigrants. Over-education may be associated with skills obsolescence by a significant proportion of workers. As a result, workers may not fully understand the risks connected to their job and may underestimate the knowledge and experience needed to perform their job safely.20 Knowledge and experience can be acquired on the job; however, over-educated workers may be more likely to experience high turnover rates because of lower job satisfaction.23 At the same time, they may be less likely to receive employer-sponsored training either because their jobs provide them with fewer such opportunities or because over-educated workers substitute their formal education for training.21–23 Having a higher level of education than required may also lead to lack of solidarity among colleagues or to conflicts with supervisors, which may in turn reduce the ability of over-educated workers to establish alliances that can help with the regulation of their workload or to obtain other forms of assistance.24
Our results among over-educated recent immigrant men are in line with previous research that has shown the risk of work-related health problems for immigrants within industries and occupations to be especially high in the period following immigration,32–34 and identify a potential mechanism for the higher injury rate. The raised impact of over-education among recent immigrants may be due to a lack of familiarity with the host country and therefore possession of less knowledge about how to raise concerns about workplace safety and with whom.20 Recent immigrants also have greater language barriers than their more established counterparts35 and may therefore have difficulty in understanding and communicating information relative to health and safety.36 While our analyses controlled for language of interview, it is possible that this measure represents an imperfect proxy for proficiency in the dominant language. In addition, recent immigrants carry a large financial burden because of the costs of resettlement and the common practice of sending remittances. This may influence their decision to accept or take risks or to participate in activities that can increase duration of exposure and reduce time for recuperation, such as working overtime or taking classes to improve their labour market opportunities.37 The lack of interaction between over-education and immigration period among women might be due to the fact that, unlike men, non-recent immigrant women who are not over-qualified may still face a high risk of injury because of the jobs in which they work.38
The results of our study should be interpreted bearing in mind the following limitations. The current occupation for each respondent—and not the occupation held when the injury occurred or was perceived—was used to establish our education-to-occupation mismatch categories. There may therefore be inconsistencies between time of exposure and outcome. In addition, our adjustment for working conditions was done using broad categories of physical demands which may not fully characterise workplace risks, particularly for RMIs. This categorisation of occupations captures risk factors for acute injuries28 and fails to consider psychosocial factors which can play a role in the genesis of RMI.39 Also, we lacked information on translation validity procedures for the CCHS and were therefore unable to assess the comparability of the different versions of the survey. Lastly, we were unable to examine the effect of working in a job which is not related to one's field of education (horizontal mismatch) and we did not have information on job tenure or work experience, workplace training, availability of personal protective equipment, social support at work, or workplace size or industry, all of which are important determinants of work injury risk.40–42
However, our study also has a number of strengths: it is based on a large representative sample of the population; it includes data collected over two cycles; it uses an objective measure of over-education and it adjusts for many possible confounders. Our study lends support to research which has shown that over-education has a negative impact on health and has highlighted the need to address barriers to suitable employment, particularly among recent immigrants. Our findings of a higher risk of injury among over-educated workers are significant given that a work injury can perpetuate over-education by limiting occupational mobility.37 Future research should explore the relationship between various forms of education-to-job mismatch and work injuries and explore the mechanisms which lead to this increased risk of injury.
What this study adds
We found that having a level of education that exceeds one's job requirements is associated with an increased risk of work injury and repetitive motion injuries for both sexes. In addition, the interaction of over-education and recent immigrant status resulted, among men, in a fourfold increase in the odds of injury compared with non-recent immigrants who were not over-educated.
Our findings highlight the need to deal with barriers to suitable employment, particularly among recent immigrants.
What is already known on the subject
An increasing number of studies have shown that education-to-job mismatch has a negative impact on workers' health, but to our knowledge the effect on the risk of work injury has not been examined. Previous studies have found formal education to have a protective effect on the risk of work injury.
Competing interests None.
Provenance and peer review Not commissioned; externally peer reviewed.
Data sharing statement No unpublished data from the study are available for sharing. The public use microdata file of the Canadian Community Health Surveys is available from Statistics Canada.
↵i Over-education is an explicit form of over-qualification. Over-qualification may make reference to education, experience and/ or expectations.
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.