The Employment Quality of Persons with Disabilities: Findings from a National Survey

Purpose Labour market integration is a widely accepted strategy for promoting the social and economic inclusion of persons with disabilities. But what kinds of jobs do persons with disabilities obtain following their integration into the labour market? In this study, we use a novel survey of workers to describe and compare the employment quality of persons with and without disabilities in Canada. Methods We administered an online, cross-sectional survey to a heterogeneous sample of workers in Canada (n = 2,794). We collected data on sixteen different employment conditions (e.g., temporary contract, job security, flexible work schedule, job lock, skill match, training opportunities, and union membership). We used latent class cluster analysis to construct a novel typology of employment quality describing four distinct ‘types’ of employment: standard, portfolio, instrumental, and precarious. We examined associations between disability status, disability type, and employment quality. Results Persons with disabilities reported consistently lower employment quality than their counterparts without disabilities. Persons with disabilities were nearly twice as likely to report low-quality employment in the form of either instrumental (i.e., secure but trapped) or precarious (i.e., insecure and unrewarding) employment. This gap in employment quality was particularly pronounced for those who reported living with both a physical and mental/cognitive condition. Conclusion There are widespread inequalities in the employment quality of persons with and without disabilities in Canada. Policies and programs aiming to improve the labour market situation of persons with disabilities should emphasize the importance of high-quality employment as a key facet of social and economic inclusion.


Introduction
Governments have implemented a wide range of policies and programs with the aim of promoting the social and economic inclusion of persons with disabilities [1,2].Among other important tools, labour market integration is a widely disabilities [14,15].The prevailing emphasis on raising overall employment rates overlooks the problem that not all jobs are created equal, and that some jobs offer few of the assumed benefits of paid employment [5,16,17].Simply put, while labour force participation has the potential to promote social and economic well-being, participation alone does not guarantee inclusion -at least not in the form of high-quality employment [18,19].
Employment quality has deteriorated in Canada and other high-income countries [20].Over the past several decades, there has been a gradual erosion of the so-called 'Standard Employment Relationship' characterized by full-time, permanent, and well-paid employment [20,21].Against a backdrop of decline in the relative share of 'good' jobs, the prevalence of temporary, low-wage, and other forms of precarious employment has increased steadily with time [22,23].Due to the numerous barriers they are known to face in the labour market, persons with disabilities may be particularly vulnerable to the threat of precarious employment [18,24].A lack of overall employment opportunities could, for instance, place undue pressure on persons with disabilities to accept any available job -even a low-quality job.
Despite their potential vulnerability to precarious employment, surprisingly little research has been undertaken to evaluate the employment quality of persons with disabilities [14,15,25,26].Furthermore, existing work in this area has tended to focus on a narrow set of variables -such as hours worked and dollars earned -resulting in a truncated view of employment conditions among persons with disabilities [11].Employment quality, on the other hand, is an inherently multidimensional construct, meant to capture a broad array of employment conditions including those related to employment stability (e.g., job security), material rewards (e.g., adequate pay), working time arrangements (e.g., a predictable work schedule), career prospects (e.g., opportunities for training and skills development), employment rights (e.g., workplace safety), and power relations (e.g., union representation) [27][28][29].There is a need for research that combines these multiple dimensions to provide a more comprehensive picture of employment quality among persons with disabilities.
The goal of this paper is to describe and compare the employment quality of persons with and without disabilities using an explicitly multidimensional approach.Drawing on a novel survey of workers in Canada, we examined how persons with and without disabilities differed across sixteen unique dimensions of employment quality.We also combined the sixteen dimensions into a comprehensive typology of employment quality to assess whether overarching 'types' of employment are more or less prevalent among persons with disabilities.We explored these patterns in terms of both disability status (i.e., yes or no) and type of disability (i.e., physical, mental/cognitive, or both).

Study Design and Sample
Data were collected using an online, cross-sectional survey administered in June 2020, as public health restrictions to limit the spread of the first wave of the COVID-19 pandemic were in the process of being relaxed.Participants were recruited from a broadly representative panel of 100,000 Canadians that is maintained by EKOS Research Associates.Panel members were eligible to participate in the study if they were at least 18 years of age, working 12 or more hours per week, and fluent in either English or French.Purposive sampling ensured that at least one third of recruited participants were living with a physical or mental/cognitive condition that created limitations at their job at least some of the time (i.e., a disability).Informed consent was obtained from all participants.
Invitations were sent to 14,570 individuals, of which 51 were determined to have an invalid email.Two reminder emails were sent.A total of 3,842 surveys were returned, yielding a response rate of 26.5%.We excluded 774 respondents who were on temporary leave, on furlough, or working fewer than 12 h per week.We also excluded 227 respondents who were unemployed, retired, or otherwise unattached to the labour market (e.g., due to school or caring responsibilities).Finally, we removed 47 responses missing data on key study variables.This resulted in a final analytic sample of 2,794 study participants.

Study Measures
Disability.Disability status was measured using a modified version of the short disability screening questionnaire (DSQ) designed by Statistics Canada [30].Three questions asked whether participants "have any [1) physical; 2) mental or cognitive; 3) other] health problems or long-term conditions expected to last six months or more that can make working difficult at least some of the time?"Examples were provided for physical health problems (e.g., seeing, hearing, walking) and mental or cognitive health problems (e.g., learning, concentrating, psychological conditions like anxiety, depression, substance use).Other health problems were manually coded as physical, mental/cognitive, or both physical and mental/cognitive.Based on these questions, we created a variable for disability status (i.e., yes/no) and type of disability (i.e., physical, mental, or both).
Employment quality.Employment quality was measured using items adapted from the Employment Precariousness Scale (EPRES) [31], the Employment Precarity Index (EPI) [21], and the Copenhagen Psychosocial Questionnaire (COPSOQ) [32].Participants were asked about the following employment conditions.

Analyses
We calculated descriptive statistics to summarize the distribution of variables.Log binomial regression models examined associations between disability status, disability type, and individual employment conditions.We employed log binomial models instead of logistic models due to the Analyses used Stata 16.0 and Mplus 8.1.Regression estimates are presented in the form of adjusted prevalence ratios (PR) and corresponding 95% confidence intervals (CI).

Ethics
Ethics Approval for the study was obtained through the University of Toronto Research Ethics Board [REB#39,085].

Results
Table 1 presents a description of the study sample.Roughly one third (35.2%) of participants reported living with one or more conditions creating a disability at work.Of the 985 participants with disabilities, 415 (42.1%) reported living with a physical disability, 326 (33.1%) reported living with a mental/cognitive disability, and 244 (24.8%) reported living with both a physical and mental/cognitive disability.Compared to participants without disabilities, participants with disabilities were more likely to be women (52.5% versus 45.3%), more likely to be born in Canada (90.2% versus 83.2%), and less likely to have completed a postsecondary degree (70.0%versus 78.3%).Persons who reported a physical disability tended to be older and were more likely to be men, relative to persons who reported either a mental/ cognitive disability or both a physical and mental/cognitive disability.
Table 2 shows the weighted prevalence of employment conditions across study groups, as well as fully adjusted poor interpretability of odds ratios with common outcomes -i.e., outcomes with point prevalence rates > 10%, such as those in the present study [33,34].Models were adjusted for age, gender, immigration status, and education.Persons without disabilities were chosen as the reference group.We used latent class analysis (LCA) to construct an empirical typology of overall employment quality, based on the individual dimensions included in the preceding step.LCA can reduce a heterogeneous sample of units into a smaller number of groups that are relatively homogeneous with respect to a given set of characteristics -in this case, employment conditions [35].We used LCA to place workers into distinct 'types' of employment, each characterized by a unique combination of employment conditions [26][27][28] To determine the most appropriate typology, we estimated multiple LCA models enumerating a progressively larger number of latent groups.To assess the adequacy of model solutions, we used formal fit indices -the Bayesian Information Criterion (BIC), the Akaike Information Criterion (AIC), and the Vuong-Lo-Mendell-Rubin (VLMR) likelihood ratio test -in combination with a substantive interpretation of the resulting outputs [35].Upon selection of the final typology, descriptive statistics summarized the distribution of individual employment conditions across employment quality types.We also summarized the prevalence of each employment quality type by disability status and disability type.Finally, we used multinomial regression models to examine associations between disability status, disability type, and employment quality.The models were adjusted for age, gender, education, immigration status, and region of residence.LCA models enumerating a progressively larger number of latent groups suggested that the optimal solution lay between three and six classes.Model solutions with seven or more classes extracted groups that were negligible in size (i.e., less than 2% of the sample).These solutions were therefore not considered.The VLMR likelihood ratio tests were consistently significant (p < 0.01) across sequential models, suggesting that model fit improved as the number of enumerated classes increased.However, AIC and BIC values plateaued beyond the four-class solution, indicating little improvement in in statistical fit beyond this solution.
negative binomial regression models estimating associations between study variables.Persons with disabilities reported lower quality employment conditions across a large majority of dimensions, including significantly higher rates of temporary employment, part-time employment, job insecurity, income insecurity, gig work, wage theft, job lock, and skill mismatch.They were also significantly less likely to report flexible work schedules, training opportunities, and positive workplace safety climates.Conversely, persons with disabilities were more likely to report membership in a union or professional association.We observed no differences between persons with and without disabilities in the prevalence of long working hours, irregular or unpredictable work schedules, and pension benefits.These associations generally held across different types of Portfolio employment was the second largest group, covering 26.5% of the sample.It was also characterized by largely favourable employment conditions, including high rates of job security, income security, flexibility, and opportunities for training.However, relative to those in standard employment, workers in this group had a higher probability of temporary employment and were more likely to report long, irregular, or unpredictable work schedules.Compared to other groups, respondents reported the lowest rates of union or association membership and employer-sponsored pension benefits.Portfolio employment could be summarized as 'secure but demanding'.
The third group, comprising 15.5% of the sample, was labeled instrumental employment.It described employment situations affording high levels of job and income security as well as regular and predictable work schedules, but little else in the way of opportunities and rewards.A vast majority of workers in this group reported being employed on a permanent basis.They also reported the highest rates of membership in a union or professional association.However, compared to other groups, workers were more likely to feel trapped in their job and least likely to perceive opportunities for training and growth.They also had a high probability of perceiving mismatch between the skills they have and those that are required for the job.Instrumental employment could be summarized as 'secure but trapped'.
The fourth and final group, accounting for 20% of the sample, was labeled precarious employment.It comprised workers who reported consistently unfavourable employment conditions across all sixteen indicators.Workers reported the highest rates of temporary employment, parttime employment, irregular schedules, unpredictable schedules, gig work, and skill mismatch.They were more likely than other groups to report job insecurity, income insecurity, and wage theft.A small minority of these workers belonged to union or professional association, and most did not consider their workplace to be an environment where they could raise concerns about their health, safety, and rights as workers.Precarious employment could be summarized as 'insecure and unrewarding'.
Table 4 shows the weighted prevalence of employment quality types across study groups.Compared to persons without disabilities, persons with disabilities were less likely to fall in the Standard Employment and Portfolio Employment groups, and more likely to fall in the Instrumental Employment and Precarious Employment groups.This was especially true for persons reporting both a physical and mental/cognitive disability.
Table 5 presents results from fully adjusted multinomial regression models estimating associations between disability status, disability type, and the employment quality typology.Standard employment was selected as the base We therefore selected the four-class solution as our typology of choice.
Table 3 shows the weighted prevalence of individual employment conditions across employment quality types.We labeled the four types: standard employment, portfolio employment, instrumental employment, and precarious employment [28,29].Standard employment was the largest group, representing 38% of the sample.It was characterized by generally favourable employment conditions.Workers described employment providing high levels of job and income security as well as regular and predictable work schedules.They also reported high degrees of flexibility, opportunities for training, and positive environments in which to raise concerns about their health, safety, and rights as workers.A majority of this group belonged to a union or professional association and all respondents reported having an employer-sponsored pension benefit.Standard employment could be summarized as 'secure and rewarding'.

Discussion
Previous research has noted a large and persisting employment gap between persons with and without disabilities [9][10][11].Starting from the premise that not all jobs are created equal -and that having any job is not the same as having a good quality job -our study moves this literature forward by examining whether analogous gaps exist in the quality of employment obtained by persons with disabilities following their integration into the labour market.We are among the first to address this question using an explicitly multidimensional approach.By taking this approach, we have provided the most comprehensive picture yet of the employment quality of persons with disabilities in Canada.Our findings point to the presence of widespread inequalities in the employment quality of Canadians with and without disabilities, suggesting that there is still a long way to go for 'full participation' in the labour market.These findings have implications for research, workplace practices, and policy in the areas of disability and employment.Consistent with our multidimensional approach, we observed inequalities not only with respect to conventional indicators of employment quality, such as hours worked and dollars earned, but across an extensive set of employment conditions ranging from job security and flexible work arrangements to skill match and opportunities for training.Taking all sixteen of our dimensions into account, persons with disabilities were nearly twice as likely as their non-disabled counterparts to report low-quality employment in the form of either instrumental (i.e., secure but trapped) or precarious (i.e., insecure and unrewarding) employment.Notably, we observed the largest gaps among those who reported living with both a physical and mental/cognitive disability.These findings highlight the importance of adopting a broad and comprehensive definition of labour market inclusion, and reinforce the need to better distinguish between employment status (i.e., having any job) and employment quality (i.e., having a good quality job) as related but still distinct aspects of inclusion [18,19,36].
Previous studies have documented how persons with disabilities experience barriers not only with respect to obtaining a job, but also with respect to the terms and conditions of their employment [11].Literature from comparable jurisdictions, such as the United States and Australia, suggests that persons with disabilities experience consistently lower quality employment than their counterparts without outcome category.The probability of portfolio (versus standard) employment was similar between persons with and without disabilities (PR: 1.09, CI: 0.88-1.35).This finding held across all three disability types.By contrast, the probability of instrumental (versus standard) employment was twice as high among persons with disabilities (PR: 2.14, CI: 1.69-2.70).Similarly, the probability of precarious (versus standard) employment was two and a half times higher among persons with disabilities (PR: 2.49, CI: 1.99-3.10).Consistent with our earlier findings, these associations were strongest among persons who reported living with both physical and mental/cognitive disabilities.Thus, in the latter group, the probability of instrumental (versus standard) employment was more than three times higher (PR: 3.35, the employment experiences of persons with disabilities by drawing on and even extending the multidimensional framework adopted here.Follow up studies should also be conducted to determine whether our findings are generalizable to other high-income jurisdictions (e.g., the United States, Europe, and Australia) and whether they hold in low-and middle-income countries, which have markedly different legislative and labour market contexts with important implications for the employment of persons with disabilities.
Our study did not gather information on employment preferences.Consequently, we had no means of ascertaining whether persons with disabilities prefer certain employment conditions over others.Nevertheless, our findings should be interpreted in light of a growing body of evidence indicating that persons with disabilities desire high-quality employment and share the same perspectives as persons without disabilities on the value and importance of key employment characteristics, such as job and income security [44].Notwithstanding a lack of information on worker preferences, the results of the present study point to a pattern of responses indicating that many persons with disabilities experience employment conditions that are not of their choosing.For example, it is hard to imagine a 'preference' for job insecurity, income insecurity, wage theft, job lock, skill mismatch and a lack of training opportunities -all of which were more commonly reported by persons with disabilities.Researchers have noted that persons with disabilities might actively select into certain forms of non-standard employment (e.g., part-time, temporary, and gig work) given that fewer hours and more flexible contracts could be of special benefit to individuals balancing the obligations of employment with personal health needs [45,46].While persons with disabilities might prefer non-standard employment arrangements because they afford greater flexibility, it is worth nothing that such arrangements cluster together with a number of indisputably adverse employment conditions.Thus, in our typology, the group of workers that reported the highest rates of temporary, part-time, and gig work (i.e., the 'precarious employment' group) also reported the greatest levels of job insecurity, income insecurity, and unpredictability.This finding suggests that meeting and accommodating the employment needs of persons with disabilities will require striking a better balance between flexibility and stability -to ensure that the former does not come at the expense of the latter [45].Future research should explore these issues by incorporating measures that explicitly assess worker preferences [44].Preferences aside, it could very well be that some study participants obtained poor quality jobs before experiencing their disability, and that the precarious nature of their jobs in turn led to a disabling injury or illness [47].Due to the cross-sectional study design, however, we were unable to explore this complexity.Longitudinal research disabilities, including fewer hours, lower pay, less job security, and fewer benefits [14,15,25,26,[37][38][39]. Whereas previous studies focused on one or two dimensions of employment, resulting in a truncated view of the problem, our analysis incorporated a wide range of variables to present a more complete picture of the employment conditions experienced by persons with disabilities.A key advantage of this approach is that we are better able to identify not only where gaps exist between persons with and without disabilities, but also where along the spectrum of employment conditions these gaps are largest and smallest.For instance, we observed larger gaps in subjective employment conditions (e.g., job security, income security, job lock, and skill match) than objective employment conditions (e.g., temporary employment, part-time employment, and regular work schedule).Further research is needed to explain this divergence across the subjective and objective dimensions of employment, as well as its implications for the measurement of employment quality among workers in general and workers with disabilities in particular [40][41][42].By incorporating a diverse set of employment variables, our analysis can also inform strategies to improve the working lives of persons with disabilities.Beyond encouraging greater attachment to the labour market, our findings emphasize the importance of job security, wages, flexible work arrangements, job fit, and training opportunities as particularly meaningful intervention targets in need of improvement.Addressing these gaps will require a combination of societal (e.g., employment standards legislation) and organizational (e.g., inclusive workplace practices) interventions.
Using a typological approach, our study also serves to demonstrate how multiple interacting dimensions of work produce distinct 'types' of employment that we labeled standard, portfolio, instrumental, and precarious employment.The typology helped us to organize a complex array of factors and highlighted that the various employment 'types' are unevenly distributed across labour market groups, including persons with and without disabilities [28,29,43,44].It was not just that persons with disabilities experienced disadvantage with respect to one or another dimension of employment, but they were also more likely to experience negative overall employment arrangements, including seemingly dead-end jobs providing little hope of career advancement and precarious jobs that offered minimal security, stability, and pay.In fact, our findings indicate that inequalities in overall employment quality were substantially larger than those observed along individual dimensions of work, suggesting that previous literature in this area may have underestimated both the nature and the extent of the employment gap between persons with and without disabilities.Future research should test the ecological validity of the typology and further strive to capture the full breadth and depth of

Conclusion
The present study demonstrates that persons with disabilities experience inequalities not only with respect to obtaining paid employment, but also with respect to the quality of employment opportunities available to them following their integration into the labour market.Given the declining share of 'good' jobs and the increasing prevalence of precarious employment, a narrow focus on employment participation (as an end in itself) has the potential to exacerbate rather than mitigate the problem of social and economic exclusion.

Data Availability
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made.The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material.If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder.To view a copy of this licence, visit http://creativecommons. org/licenses/by/4.0/.could be beneficial in this regard, allowing a view of change over time in both disability status and employment quality.
Strengths of the present study include the following.First, our observations are drawn from a heterogeneous and broadly representative panel of Canadian workers.Second, through purposive sampling methods, we recruited persons with diverse experiences of disability, enabling a disaggregated approach accounting for both disability status and type of disability.Finally, compared to previous work in this area, we adopted a more comprehensive and multidimensional approach to the measurement of employment quality.At the same time, our findings and the conclusions we have drawn should be situated in the context of some key limitations.Despite the broadly representative panel from which the study participants were drawn, our survey only returned a response rate of 26.5%.This response rate is comparable to other surveys of this kind.Nevertheless, it introduces the potential for selection bias in our analytic sample.For instance, it is probable that certain groups of workers, such as those living with severe disabilities, were less likely than others to participate in the panel or respond to our invitation, leading to further problems of generalizability.These limitations of the sample should be weighed against the fact that population-based employment surveys in Canada -such as the flagship Labour Force Survey and recent Survey on Quality of Employment -lack measures of disability status and were therefore not suitable for our purposes.Second, as noted above, the cross-sectional nature of the data prohibited us from determining whether participants experienced their disability before or after entering their most recent employment situation.As a result, we were unable to test the directionality of observed associations between disability and employment quality.Finally, data were collected a few months after the onset of the global COVID-19 pandemic, raising concerns that our findings may reflect employment and other surrounding conditions specific to this exceptional period.Indeed, for many of the same reasons discussed in the text above, persons with disabilities were particularly vulnerable to the labour market impacts of the COVID-19 pandemic and associated policy responses [48].It is worth noting, however, that our survey was administered during a period in which public health restrictions were being lifted and a large majority of Canadians were attending their usual workplaces [49].At the same time, the growing availability of telework arrangements in the aftermath of the pandemic raises important questions about the labour market integration and workplace accommodation of persons with disabilities [50].It will be important for future research to examine how the increasing prevalence of home-based work is affecting the labour market and employment outcomes of persons with disabilities.
Policies and programs aiming to improve the labour market situation of persons with disabilities should emphasize the importance of high-quality employment as a key ingredient of inclusion.from the Social Sciences and Humanities Research Council of Canada and the Canadian Institutes of Health Research #890-2016-3002.The Institute for Work & Health operates with the support of the Province of Ontario.The views expressed herein are those of the authors and do not necessarily reflect those of the province of Ontario.

Table 1
Descriptive characteristics of the study sample

Table 2
Prevalence and adjusted prevalence ratios of employment conditions, by disability status and disability type Prevalence Ratio; CI = Confidence Interval; Estimates are adjusted for age, gender, education, immigration status, and region of residence.

Table 3
Prevalence of employment conditions, by employment qual-

Table 4
Prevalence of employment quality types, by disability status No Disability