Publisher’s Note: A correction article relating to this paper has been published and can be found at https://sjdr.se/articles/10.16993/sjdr.1619.
Introduction
Persons with disabilities constitute a significant and historically marginalised group globally and within European welfare states.1 According to the World Health Organization (2022), about 16% of the world’s population lives with some form of disability. In Norway, recent estimates suggest that 15% of the population is affected by disability, including physical, sensory, intellectual, and psychosocial impairments (Norges Handikapforbund 2024). While Norway is often viewed as a leader in welfare provision and social inclusion, persistent disparities remain in access to and outcomes of education, employment, income security, and civic engagement for persons with disabilities. These domains represent not only key markers of social integration but are also widely recognised as critical indicators of adult status in contemporary societies.
The barriers to achieving equitable life outcomes for persons with disabilities have been extensively documented in international and national research. Numerous studies emphasise how limited access to education in early life often translates into compounded disadvantage later in adulthood, manifesting in lower employment rates, among other negative life outcomes (WHO 2011; McKinley Yoder and Cantrell 2019; Karmel and Nguyen 2008; Gugushvili et al. 2024). International conventions, such as the United Nations Convention on the Rights of Persons with Disabilities (CRPD), emphasise the right to inclusive education and equal participation in public life. However, the practical realisation of these rights remains uneven. Women with disabilities, in particular, might face intersecting layers of exclusion that shape their access to educational resources and amplify barriers to employment (United Nations Department of Economic and Social Affairs 2019; Hunt et al. 2022).
Conceptual and empirical research in disability studies increasingly shifts away from medicalised understandings of disability as individual pathology and toward social models that emphasise structural barriers and social arrangements (Shakespeare 2013). Within this perspective, educational attainment is a social institution embedded in systems of stratification, capable of both mitigating and reinforcing inequality. In sociological terms, education is widely regarded as a proxy for human capital, often linked to improved life chances and increased participation in economic and civic life. However, whether education operates in the same way for persons with and without disabilities remains an open question.
In recent years, some evidence suggests that the educational attainment of persons with disabilities is improving, particularly in Nordic welfare states such as Norway, Sweden, and Denmark (OECD 2022). Yet higher education alone does not guarantee equal outcomes. Studies in the sociology of disability argue that education may not yield comparable returns for persons with disabilities due to continued labour market discrimination, inaccessible work environments, and entrenched stigma (Tøssebro 2016). From a disability rights perspective, this disconnect between formal equality (educational attainment) and substantive equality (employment) raises critical concerns about the real-world value of education for persons with disabilities.
Despite growing attention to educational disparities and adult status outcomes, there is limited research that examines how these associations unfold over time and differ between people with and without disabilities. Few studies have explicitly investigated whether education exerts similar effects on employment across these groups, or whether disability status moderates the expected benefits of educational attainment (Bliksvær 2018). Likewise, while some international studies address gendered disability inequalities, empirical evidence from the Nordic region on how gender intersects with disability and education to shape adult life outcomes remains relatively scarce.
We address this empirical gap by analysing repeated cross-sectional data from nine waves of the Norwegian European Social Survey to examine whether education serves as an equalising force for employment. Focusing on employment as a key aspect of adult outcomes, we estimate the likelihood of being in paid work across three levels of educational attainment, comparing individuals with and without disabilities and, among the population with disabilities, women and men. Using linear probability models, we assess absolute differences in employment probabilities and whether these associations vary by disability status, gender, and over time. Situating the analysis in Norway, in a context with strong welfare institutions, active labour market policies, and formal commitments to disability inclusion, we contribute to debates on disability-related labour market inequality by asking whether higher education offsets disadvantages linked to disability and gender or whether structural barriers persist even among those with similar qualifications.
The remainder of the article is structured as follows. The next section situates the study within existing research on disability, education, and labour market inclusion in Norway and the Nordic region. We then outline our theoretical framework and hypotheses before describing the data, measures, and analytical strategy. Subsequent sections present the empirical results and discuss their implications for understanding the conditional equalising role of education, followed by conclusions that highlight limitations and directions for future research.
Background of the Study
In contemporary welfare states, the concept of adult status is closely tied to participation in paid employment. Beyond offering financial security, employment signals full societal participation and inclusion and is often seen as a key indicator of adulthood (Wright and Stumm 2025; Walczak 2023; Bernardini 2014). For most people, employment is a way to contribute to society and establish social connections. However, not all individuals have equal access to this foundational dimension of adult life. Among those particularly at risk of exclusion are persons with disabilities who continue to face significant barriers to full participation in the labour market (Gugushvili et al. 2025; Gugushvili, Grue and Finnvold 2025).
Existing Nordic research provides important context for our focus on employment as a key adult outcome for persons with disabilities. Comparative studies show that, despite extensive welfare provisions and anti-discrimination frameworks, persons with disabilities in the Nordic countries remain less likely to hold stable employment and more likely to be located in sheltered or marginal work (Garrels and Høybråten Sigstad 2021; Garrels et al. 2024; Kuznetsova et al. 2017). Existing evidence suggests persistent wage and labour market disadvantages even within a universalistic welfare regime (Ballo 2023; Bjørnshagen and Ugreninov 2021), and a scoping review of Scandinavian reforms finds only limited progress in labour market inclusion for chronically ill individuals (Rydland et al. 2023). Studies also highlight that employers’ assumptions regarding productivity, accommodation needs, and ‘fit’ continue to shape hiring and retention decisions (Strindlund et al. 2019) and that young people with mental health challenges face distinct barriers to secure entry into the labour market (Bjørnshagen and Ugreninov 2021).
Despite Norway’s formal legal protections for equal participation, existing evidence shows that persons with disabilities experience substantial disadvantages in labour market participation. National reports consistently highlight significant gaps in employment between those with and without disabilities. According to Bufdir (2025), the employment rate among persons with disabilities is just 41%, compared to 75% for those without disabilities. These gaps persist even after accounting for age and other demographic factors, indicating structural inequalities that continue to shape the lives of persons with disabilities.
Disparities in education have frequently been cited as one of the root causes of these inequalities. According to the report by NOVA (2014), 64% of persons with disabilities in Norway have only completed primary education, compared to 17% of the general population. Similarly, Finnvold (2018) reports that many individuals with disabilities do not attain a secondary school diploma. These findings point to a strong link between lower levels of educational attainment and diminished employment prospects for persons with disabilities. Since education is one of the strongest predictors of employment in modern economies, these gaps may help explain why persons with disabilities remain under-represented in the labour market.
In response to the described challenges, Norway has implemented several legislative and policy measures aimed at promoting equal employment opportunities. The Equality and Anti-Discrimination Act of 2018 is particularly significant in this context. It prohibits discrimination based on disability and requires employers to provide reasonable accommodations for job seekers and employees with disabilities. The Act promotes universal design principles and compels employers to take steps that ensure that persons with disabilities can participate in work life on equal terms (§ 17–23). The Norwegian government has also introduced targeted measures to encourage employers to hire persons with disabilities, including financial incentives and workplace adjustments (Norwegian Ministry of Children and Equality 2019). These efforts reflect a broader commitment to the values enshrined in the CRPD, notably Article 27, which affirms the right of persons with disabilities to work on an equal basis with others.
In addition to employment policy, education policy in Norway emphasises the importance of inclusion and equity. All children are entitled to 10 years of compulsory schooling, and those with special educational needs have the right to adapted education and individual accommodations. Section 21 of the Equality and Anti-Discrimination Act requires educational institutions to ensure accessibility in both physical infrastructure and pedagogical content. These efforts are reinforced at the tertiary level by the work of the Nordic Network for Disability Coordinators, which promotes inclusive practices in higher education across the country. Despite these structural commitments, significant educational disparities remain between persons with and without disabilities, raising questions about how effective inclusive education policies have been in practice (Myhr et al. 2018; Gugushvili et al. 2025).
Previous research has documented the employment challenges faced by persons with disabilities in Norway, but significant gaps remain. Much of the existing literature focuses on specific impairment groups or is geographically limited to urban areas such as the capital city. For instance, Bjørnshagen and Ugreninov (2021) conducted a field experiment in Oslo and found that persons with disabilities were significantly less likely to be called for job interviews. Their study, however, focused narrowly on wheelchair users and excluded other forms of disability and broader regional contexts. Brunes and Heir (2022) also found that higher education was positively associated with employment among individuals with visual impairments, but their findings are similarly limited in scope.
Overall, this body of work demonstrates that Nordic welfare states combine strong formal commitments to equality with enduring structural and cultural barriers to inclusion. Our contribution is to build on this evidence by examining whether higher education mitigates, reproduces, or reconfigures these disadvantages for women and men with disabilities in Norway. Many existing studies emphasise the role of discrimination or physical access barriers, but fewer have directly examined the extent to which education may mitigate these employment disadvantages. Moreover, the extent to which these patterns differ across different demographic subgroups, particularly gender, has also received limited attention. Given Norway’s strong policy emphasis on inclusion and anti-discrimination, as well as the central role of employment in adult life, there is a need to understand how education intersects with disability status in shaping employment opportunities.
Theoretical Framework and Hypothesis
Disability is best understood as a multidimensional phenomenon influenced by the interaction between personal impairments and the social, institutional, and cultural environments in which people live. Historically, the medical model has framed disability as a personal pathology, an inherent deficit or abnormality located within the individual that limits functioning and requires treatment or rehabilitation (Priestley 2005; Retief and Letsos 2018). From this perspective, a disadvantage in employment or civic participation is attributed to the individual’s impairments. In contrast, the social model emphasises how societal arrangements, such as inaccessible infrastructure, exclusionary institutions, and discriminatory norms, create disabling conditions for individuals with impairments (UPIAS 1976). While the social model offers an important shift in emphasis, it has also been criticised for downplaying the lived effects of the impairments themselves. As a result, a hybrid approach has gained prominence, viewing disability as a product of interactions between bodies and environments (Shakespeare 2013).
This hybrid model underpins the present study, which examines whether education serves as an equalising mechanism in the labour market for individuals with and without disabilities. Our point of departure is the human capital theory, which posits that education enhances individuals’ productivity and employability by increasing their cognitive skills, social competencies, and adaptability (Becker 1993; Schultz 1988). In modern economies, education is viewed as the primary mechanism by which individuals improve their chances of securing and maintaining employment. From this perspective, individuals with equivalent educational qualifications should, in principle, have comparable labour market outcomes, regardless of disability status. If education equips all individuals with similar productive capacities, then it should serve as a leveller of opportunity in the employment domain.
Research on heterogeneous returns to education and gendered welfare-state arrangements suggests that educational credentials do not operate in isolation from institutional and cultural contexts (Bliksvær 2018; Bulczak et al. 2025). Related work finds that returns to education among adults with disabilities might vary across social groups and institutional settings (Henderson et al. 2017; Bryan et al. 2025). At the same time, gendered analyses of Nordic welfare states show that egalitarian frameworks coexist with horizontal and vertical gender segregation, part-time work norms, and caregiving expectations that disproportionately shape women’s opportunities (Mustosmäki et al. 2021; Orloff 2009; Risman 2018). For women with disabilities, these dynamics intersect with ableism in hiring and workplace practices (Strindlund et al. 2019), as well as sectoral patterns in Nordic labour markets that channel women into public-sector and care-related occupations. Therefore, we conceptualise education as a potentially equalising but conditionally realised resource whose association with employment may vary by disability status, gender, and institutional context. This framing motivates our focus on whether the employment benefits of higher education accrue equally to women and men with disabilities and whether these relationships shift across time in an inclusive welfare state like Norway.
Hypothesis 1: Education is positively associated with the likelihood of employment for persons with disabilities, just as it is for persons without disabilities.
This hypothesis reflects the foundational assumption of human capital theory. It suggests that higher levels of education should improve employment chances for all individuals, regardless of disability status. However, the actual labour market benefits of education may be unequally distributed due to structural and cultural constraints. Research increasingly suggests that the returns to education in the form of employment may not be equal for persons with disabilities. Structural barriers, such as discriminatory hiring practices, inadequate workplace accommodations, and persistent stigma, can prevent individuals with disabilities from fully benefiting from their educational investments (Banks et al. 2021; Tøssebro 2016). Even highly educated individuals with disabilities may encounter limited opportunities for meaningful or stable employment. These disparities raise concerns about whether education functions as an effective equaliser in societies that formally espouse egalitarianism.
Hypothesis 1 addresses average patterns across disability groups; however, it is also important to recognise heterogeneity within the population with disabilities, especially regarding gender. Theories of intersectionality emphasise that social categories, such as gender and disability, do not operate independently but interact in ways that produce distinct forms of disadvantage (Crenshaw 1991; Weldon 2008). Women with disabilities often face compounded barriers to employment due to the simultaneous effects of ableism and sexism. Cultural norms around caregiving, unequal household responsibilities, and employer bias can further constrain their labour market participation (Polachek 2012; Charles and Grusky 2005). For instance, women with disabilities may be more likely to take part-time or flexible jobs, exit the labour market due to caregiving duties, or face scepticism about their professional competence. In contrast, men with disabilities may be perceived as more employable, particularly in traditionally male-dominated sectors. These overlapping forms of exclusion suggest that gender mediates the returns to education in the employment domain:
Hypothesis 2: Education is more strongly associated with employment for men with disabilities compared to women with disabilities.
This hypothesis builds on human capital and intersectionality theories, suggesting that gender moderates the link between education and employment. Even with the same level of education, women with disabilities may face greater structural and normative barriers in the labour market than their male counterparts. At the same time, the Norwegian labour market may generate a countervailing pattern through gendered horizontal labour market segregation. Women, including women with disabilities, are disproportionately concentrated in public-sector and care-related occupations such as health, education and social services, where university credentials often function as formal entry tickets and where recruitment is more standardised, union coverage is stronger and accommodation may be more routinised (Nilsen 2018). Men with disabilities are more often located in male-dominated or more physically demanding parts of the private sector, where ableist productivity norms, informal recruitment, and expectations of uninterrupted full-time work may reduce the employment payoff of education. These sectoral differences indicate that women and men with disabilities may convert similar educational credentials into employment under varying institutional conditions.
The extent to which education promotes equitable access to employment may also vary across time. In contemporary welfare states like Norway, egalitarian values, inclusive education policies, and anti-discrimination legislation are designed to reduce inequality in labour market access. Modernisation theory and neo-institutionalist perspectives argue that, as societies develop, meritocratic principles should increasingly override discriminatory practices based on ascriptive traits like gender or disability (Joas & Knöbl 2009; Charles and Grusky 2005). Accordingly, one might expect that over time, education should become a more reliable predictor of employment for all groups.
Nevertheless, the persistence of cultural essentialism, deeply ingrained beliefs about who is fit for work, may weaken these expectations. Stereotypes about productivity, dependability, or physical capacity continue to inform hiring decisions, often to the disadvantage of persons with disabilities. This suggests that even in egalitarian policy environments, cultural beliefs may constrain the equalising potential of education in the labour market.
Hypothesis 3: Across time, education is less strongly associated with employment for individuals with disabilities compared to those without disabilities.
This hypothesis tests whether the labour market benefits of education have become more equitable over time or whether structural and cultural barriers continue to undermine educational returns for persons with disabilities. Finally, we consider whether gender differences in the employment returns to education among persons with disabilities persist across time. Despite institutional commitments to gender equality, cultural norms privileging male workers, particularly in full-time, continuous employment, may reinforce disparities. Ideologies of male primacy can affect both employer expectations and self-perceptions, shaping labour market trajectories from the outset (Charles and Grusky 2005; Moss-Racusin et al. 2012). Women with disabilities may be steered toward lower-paying, precarious jobs and excluded from sectors with high employment stability or upward mobility.
Hypothesis 4: Across time, education is more strongly associated with employment for men with disabilities compared to women with disabilities.
This hypothesis reflects a concern that, despite formal policy commitments to inclusion, gendered and ableist norms continue to shape how educational attainment translates to labour market opportunities. In summary, this study draws on a hybrid model of disability, human capital theory, intersectionality, and modernisation perspectives to examine how educational attainment is linked with employment outcomes among persons with and without disabilities in Norway. It further explores how these patterns are shaped by gender and evolve over time. This theoretical framework and associated hypotheses guide an empirical investigation into whether education promotes equitable labour market integration or whether it reproduces persistent inequalities in a Nordic welfare state context.
In this study, we bring the described perspectives together by treating human capital theory and intersectionality as complementary rather than competing frameworks. Human capital theory specifies why higher education should, in principle, enhance employment chances for all individuals, while intersectional approaches draw attention to how the value of similar educational resources is structured by intersecting power relations, including disability and gender. Hence, we conceptualise education as a potentially equalising resource whose returns are contingent on ableist and gendered labour market contexts. This integrated framework underlies our hypotheses about differential returns to education based on disability status, gender, and over time.
Research Design
The European Social Survey
This study draws on data from the European Social Survey (ESS), a biennial cross-national survey that has been conducted since 2002 to monitor and explain long-term changes in public attitudes, values, and behaviours in various European countries. The ESS employs high methodological standards, including random probability sampling and face-to-face interviews, which together make it a trusted source of data for social scientists. Individuals aged 15 and above who live in private households are eligible for inclusion, and participation is voluntary. One of the key strengths of the ESS lies in its comparability across countries and time. Its harmonised methodology allows researchers to track social change within nations in a reliable manner. The dataset is publicly available at no cost, making it widely used in both academic and policy-related research.
In this study, we use data from Norway, focusing on nine ESS rounds conducted between 2002 and 2018. This period was selected to ensure the stability of the results and to avoid potential distortions linked to the COVID-19 pandemic, which significantly impacted labour markets, health, and social inequality. Including data beyond 2018 might have confounded the analysis with pandemic-related effects that are not central to the study’s core interest. Limiting the scope to pre-pandemic years strengthens the internal validity of our analysis and improves comparability across survey waves. The employed data are fully anonymised, and the authors accessed no personally identifiable information. All ESS waves are conducted in accordance with national and ESS European Research Infrastructure ethical and data protection requirements; respondents are informed about the study and provide informed consent prior to participation. No additional institutional ethics approval was required for this secondary analysis of publicly available data.
Measuring disability
Disability is assessed based on a widely used ESS question: ‘Are you hampered in your daily activities in any way by any longstanding illness, or disability, infirmity or mental health problem?’ Respondents can choose among three responses: ‘yes, a lot’, ‘yes, to some extent’ and ‘No’. Following established practice (e.g., Geiger, van der Wel and Tøge 2017; Gugushvili et al. 2023; Gugushvili and Grue, 2025), we define individuals as having a disability if they answer either of the first two categories. Those who respond with ‘no’ are treated as not having a disability. This operationalisation offers a broad but practical measure of functional limitations and captures individuals whose conditions affect their everyday activities to varying degrees. The study compares individuals with and without disabilities and, within the group reporting disabilities, also examines differences between men and women. This intersectional approach provides a detailed understanding of how disability and gender jointly shape social outcomes. The full dataset includes 9,488 individuals in Norway. Among these, 2,182 report a disability, comprising 1,138 women and 1,044 men. The remainder, 7,306 individuals, do not report having a disability.
Measuring education and employment
Education is the primary independent variable used in the analysis. Educational attainment is measured using the International Standard Classification of Education (ISCED). Although the ESS provides a Norwegian-specific education classification variable, this variable contains a large number of missing observations and was therefore excluded from our analyses to preserve sample size and statistical power. We collapsed the ISCED educational levels into three broader categories to enhance interpretability and ensure adequate statistical precision across subgroups. The first category, primary education, includes ‘less than lower secondary’, ‘lower secondary’ and ‘lower tier upper secondary’ education. The second category, secondary education, includes ‘upper tier upper secondary’ and ‘advanced vocational, sub-degree’. The third category, university education, includes ‘lower tertiary education (BA level)’ and ‘higher tertiary education (MA level or above)’. This classification reflects significant divisions in educational attainment and is commonly used in studies of social stratification and labour market outcomes.
The described grouping also aligns with the understanding of education as a proxy for human capital. Higher levels of education are typically associated with greater skills, knowledge, and access to resources, which in turn enhance opportunities in the labour market. Using categorical rather than continuous education measures helps capture these qualitative distinctions and supports a more straightforward interpretation of differences in employment probabilities across education levels.
Employment is the primary dependent variable examined in this study. It is measured using an ESS question that asks respondents what they were doing during the seven days preceding the survey. Individuals who reported being in paid work (as employees, self-employed, or working in a family business) or temporarily away from such work were coded as employed. All other responses were coded as not employed. This binary classification is consistent with prior studies using ESS data (e.g., Geiger et al. 2017; Gugushvili et al. 2023) and offers a straightforward measure of labour market participation. As discussed above, employment plays a central role in economic and social life, influencing inclusion and autonomy. This is particularly true for persons with disabilities, for whom participation in the labour market remains a persistent challenge due to social and institutional barriers.
Covariates
To ensure robust estimates and to account for potential confounders, we control for a number of socio-demographic variables: age, gender, area of residence (urban or rural), minority status, and survey year. Age is restricted to a range of 25–62 years, based on the assumption that individuals within this range are most likely to have completed formal education and remain active in the labour market. In Norway, higher education typically starts at age 18 and is completed by age 25. Individuals may start drawing pensions from age 62, making this an appropriate upper age limit for analysing employment.
Gender is measured as a binary variable in the ESS, with women coded as 1 and men as 0. Accounting for gender is important, as it is correlated with employment as a result of the traditional gender division of household labour and gender discrimination in the labour market. The survey also includes a self-reported indicator of ethnic minority status, coded as 1 for minority and 0 for majority. We include minority status in our analysis because it captures an additional axis of stratification that is correlated with both educational opportunities and employment chances. Place of residence is used as a proxy for regional labour market conditions and is included to capture potential urban-rural disparities. Urban residents may have better access to job opportunities and support services than those in rural areas. Survey year is modelled as a continuous variable ranging from 1 (corresponding to 2002) to 9 (corresponding to 2018). This allows us to examine time trends and test whether the association between education and employment outcomes has changed over the study period.
Statistical analysis
The primary analytical strategy involves estimating linear probability models (LPMs), which are appropriate for binary dependent variables and yield coefficients that are easy to interpret. In contrast to logistic regression models, where coefficients reflect changes in the natural logarithm of the odds, LPM coefficients represent direct changes in the probability of employment associated with a given independent variable. The reported coefficients in our main regression models are point estimates that can be interpreted as absolute changes in the probability of being employed, expressed in percentage points, relative to the specified reference categories (e.g., a coefficient of 0.10 indicates a 10-percentage-point higher probability of employment).
We estimate separate models for individuals with and without disabilities and additionally split the sample of individuals with disabilities by gender to examine potential intersectional differences. This approach allows us to compare how the education-employment relationship varies both by disability status and within subgroups of the population with disabilities. Both bivariate and multivariate regression models are employed. Bivariate models provide an initial overview of the unadjusted relationship between education and employment, while multivariate models control for the complete set of covariates discussed earlier. Including these controls strengthens the study’s internal validity and helps isolate the specific contribution of educational attainment to employment outcomes. We also test for interactions between education and survey year to explore whether the relationship between education and employment has changed over time. These interaction terms help identify whether trends in educational inequality have remained stable or shifted in either direction. Though interaction models can be complex to interpret, they are helpful in assessing the persistence of social inequalities.
In addition, we conducted robustness checks using alternative coding of the key independent variables (e.g., binary educational attainment and severity-specific disability groupings). The findings from these additional models confirm that the core relationships observed in the main analysis are stable and not sensitive to different operational definitions. In summary, this research design leverages high-quality survey data, a clear set of theoretically informed measures, and well-established statistical techniques to investigate how education is linked to employment outcomes in the Norwegian context, particularly among individuals with disabilities. By accounting for gender and temporal variation, the design provides a foundation for examining both persistent and changing patterns of inequality. Analyses are based on the combined survey weights recommended in the ESS documentation to account for sampling design and differential non-response. All models are estimated using list-wise deletion on the dependent variable, key independent variables, and covariates, with heteroskedasticity-robust standard errors reported. Statistical analyses were performed using R Studio, an open-source software environment widely used for data science and statistical modelling.
Results
Descriptive findings
Tables 1 and 2 present the descriptive characteristics of the full sample and by disability status and gender. We can see that the prevalence of disability is slightly higher among women. Nearly half of the sample (44%) hold a university-level education, 47% have secondary education, and under 10% have only primary education. Educational attainment varies by disability status: only 31% of persons with disabilities report having a university degree versus 48% among those without disabilities. Table 2 further shows pronounced gender differences within these groups: among persons with disabilities, women are more likely than men to hold a university degree (34% vs. 28%), while among non-disabled respondents, women also have a higher university share than men (54% vs. 43%). Respondents with disabilities are older on average (47 vs. 43 years), minority shares are small across groups, and non-disabled respondents are more urban.
Table 1
Characteristics of the sample.
| TOTAL SAMPLE N = 9,488 | DISABLED N = 2,182 | NON-DISABLED N = 7,306 | |
|---|---|---|---|
| Education level | |||
| Primary education | 897 (9.5%) | 326 (15%) | 571 (7.8%) |
| Secondary education | 4,440 (47%) | 1,178 (54%) | 3,262 (45%) |
| University education | 4,151 (44%) | 678 (31%) | 3,473 (48%) |
| Employment status | |||
| Employed | 8,151 (86%) | 1,434 (66%) | 6,717 (92%) |
| Unemployed | 1,337 (14%) | 748 (34%) | 589 (8.1%) |
| Gender | |||
| Woman | 4,425 (47%) | 1,138 (52%) | 3,287 (45%) |
| Man | 5,063 (53%) | 1,044 (48%) | 4,019 (55%) |
| Age | 44 (10) | 47 (10) | 43 (10) |
| Ethnic minority status | |||
| No | 9,037 (95%) | 2,059 (94%) | 6,978 (96%) |
| Yes | 451 (4.8%) | 123 (5.6%) | 328 (4.5%) |
| Area of residence | |||
| Rural area | 3,765 (40%) | 992 (45%) | 2,773 (38%) |
| Urban area | 5,723 (60%) | 1,190 (55%) | 4,533 (62%) |
[i] Notes: n (%); Mean, (SD).
Table 2
Characteristics of persons with and without disabilities by gender.
| DISABLED | NON-DISABLED | |||||
|---|---|---|---|---|---|---|
| TOTAL SAMPLE N = 2,182 | WOMEN N = 1,138 | MEN N = 1,044 | TOTAL SAMPLE N = 7,306 | WOMEN N = 3,287 | MEN N = 4,019 | |
| Educational level | ||||||
| Primary | 326 (15%) | 166 (15%) | 160 (15%) | 571 (7.8%) | 242 (7.4%) | 329 (8.2%) |
| Secondary | 1,178 (54%) | 587 (52%) | 591 (57%) | 3,262 (45%) | 1,282 (39%) | 1,980 (49%) |
| University | 678 (31%) | 385 (34%) | 293 (28%) | 3,473 (48%) | 1,763 (54%) | 1,710 (43%) |
| Employment status | ||||||
| Employed | 1,434 (66%) | 707 (62%) | 727 (70%) | 6,717 (92%) | 2,925 (89%) | 3,792 (94%) |
| Unemployed | 748 (34%) | 431 (38%) | 317 (30%) | 589 (8.1%) | 362 (11%) | 227 (5.6%) |
| Age | 47 (10) | 46 (10) | 47 (10) | 43 (10) | 43 (10) | 43 (10) |
| Ethnic minority status | ||||||
| No | 2,059 (94%) | 1,081 (95%) | 978 (94%) | 6,978 (96%) | 3,134 (95%) | 3,844 (96%) |
| Yes | 123 (5.6%) | 57 (5.0%) | 66 (6.3%) | 328 (4.5%) | 153 (4.7%) | 175 (4.4%) |
| Area of residence | ||||||
| Rural area | 992 (45%) | 514 (45%) | 478 (46%) | 2,773 (38%) | 1,217 (37%) | 1,556 (39%) |
| Urban area | 1,190 (55%) | 624 (55%) | 566 (54%) | 4,533 (62%) | 2,070 (63%) | 2,463 (61%) |
[i] Note: n (%); Mean, (SD).
Employment is common across the sample (86%), but a substantial gap exists by disability status: 66% of persons with disabilities are employed versus 92% of non-disabled respondents (a 26-point difference). Table 2 reveals gendered patterns within these groups: among persons with disabilities, 62% of women and 70% of men are employed (an 8-point gap), while among non-disabled respondents, the corresponding rates are 89% and 94% (a 5-point gap). These contrasts, lower education and lower employment among persons with disabilities overall, and gendered differences within both disability groups, motivate the intersectional multivariable analyses that follow.
Links between disability, education, and employment
This section reports the results from LPM models examining the association between education and employment status. Results were estimated for both simple and adjusted models; however, only estimates from the fully adjusted model are discussed here, as they account for potential confounding factors.
Figure 1 and Table S1 in the Supplementary Materials display the regression models testing the relationship between education and employment for respondents with and without disabilities. Results indicate a strong and statistically significant positive association between higher levels of education and the probability of being in paid work for both groups. These findings support the hypothesis linking education to labour market participation. For respondents with disabilities, secondary education is associated with a 20-percentage-point higher probability of employment relative to primary education (β = 0.20; 95% CI: [0.18, 0.23]; p < 0.001). For their non-disabled counterparts, secondary education is associated with a 10-percentage-point increase (β = 0.10; 95% CI: [0.09, 0.11]; p < 0.001). The employment advantage associated with university education is even greater: 36 percentage points for respondents with disabilities (β = 0.36; 95% CI: [0.34, 0.39]) and 14 percentage points for non-disabled individuals (β = 0.14; 95% CI: [0.12, 0.15]). These differences suggest that the positive effect of education on employment is more substantial for persons with disabilities than for those without them, highlighting the potentially transformative role of higher education in mitigating labour market exclusion.
Figure 1
Employment probabilities among disabled and non-disabled persons.
Note: Bars show 95% confidence intervals. Sample size by subgroup: Secondary educated with disabilities: 1,178, secondary educated without disabilities: 3,262, university educated with disabilities: 678, university educated without disabilities: 3,473. Accessibility description Employment probabilities increase with education for both groups, but the gradient is sleeper among persons with disabilities. Respondente with disabilities have lower predicted employment than non-disabled respondente at every educational level, while the largest employment gains are associated with university education, especially among persons with disabites.
In further analyses depicted in Figure 2 and Table S2 in the Supplementary Materials, we distinguish the association between education and employment by gender among persons with disabilities. For women with disabilities, secondary education is associated with a 24 percentage point (95% CI: [0.20, 0.27]) increase in the probability of employment compared to primary education, while university education corresponds to a 40 percentage point (95% CI: [0.37, 0.44]) increase. Among men with disabilities, the respective increases are 16 (95% CI: [0.13, 0.20]) and 32 (95% CI: [0.28, 0.36]) percentage points. These findings may suggest that while higher education benefits both genders, its magnitude may be greater for women. This gender difference invites further investigation, particularly in relation to the intersecting barriers faced by women with disabilities in the Norwegian labour market over time.
Figure 2
Employment probabilities among men and women with disabilities.
Note: Bars show 95% confidence intervals. Sample size by subgroup: Secondary educated men: 591, secondary educated women: 587, university educated men: 293, university educated women: 385. Accessibility description: Employment probabilities increase with education for both disabled women and disabled men, but the gradient is steeper for women. Among persons with disabilities, women have lower employment at lower levels of education, but they appear to gain more from secondary and especially university education than men, indicating a stronger education-employment association among disabled women.
Over time trends
To assess whether the association between education and employment changed across ESS waves, we estimated interactions between education and survey year. Table 3 shows clear, positive education × year interactions for both persons with and without disabilities. Among respondents with disabilities, the association between secondary education and employment increases by about 0.03 points per wave in Model 2, and the association for university education increases by about 0.04 points per wave (all statistically significant). For non-disabled respondents, the corresponding increases are smaller, about 0.01 points per wave for both secondary and university education (statistically significant). Together with the negative main effect of year, these patterns indicate a modest widening of the educational gradient in employment over time: overall employment probabilities drift slightly downward, but higher education increasingly offsets that decline, especially among persons with disabilities.
Table 3
Changes in employment across time among persons with and without disabilities, point estimates from LPMs.
| DISABLED | NON-DISABLED | |||
|---|---|---|---|---|
| MODEL 1 | MODEL 2 | MODEL 1 | MODEL 2 | |
| Secondary education (Ref. primary education) | 0.10*** | 0.08*** | 0.02* | 0.04*** |
| (0.06, 0.15) | (0.04, 0.13) | (0.003, 0.04) | (0.02, 0.06) | |
| University education | 0.22*** | 0.19*** | 0.04*** | 0.07*** |
| (0.17, 0.26) | (0.15, 0.23) | (0.03, 0.06) | (0.06, 0.09) | |
| Year | –0.02*** | –0.03*** | –0.01*** | –0.01*** |
| (–0.03, –0.02) | (–0.03, –0.02) | (–0.02, –0.01) | (–0.01, –0.004) | |
| Women (Ref. men) | – | –0.10*** | – | –0.07*** |
| – | (–0.11, –0.09) | – | (–0.07, –0.06) | |
| Age | – | –0.002*** | – | 0.003*** |
| – | (–0.003, –0.002) | – | (0.003, 0.003) | |
| Ethnic minority (Ref. ethnic majority) | – | –0.10*** | – | –0.11*** |
| – | (–0.13, –0.08) | – | (–0.12, –0.10) | |
| Urban area (Ref. rural area) | – | 0.01 | – | –0.02*** |
| – | (–0.004, 0.02) | – | (–0.02, –0.02) | |
| Secondary education*Year | 0.02*** | 0.03*** | 0.02*** | 0.01*** |
| (0.01, 0.03) | (0.02, 0.03) | (0.01, 0.02) | (0.01, 0.02) | |
| University education*Year | 0.03*** | 0.04*** | 0.02*** | 0.01*** |
| (0.02, 0.04) | (0.03, 0.04) | (0.01, 0.02) | (0.01, 0.02) | |
| Constant | 0.53*** | 0.71*** | 0.87*** | 0.78*** |
| (0.49, 0.58) | (0.66, 0.76) | (0.85, 0.89) | (0.76, 0.80) | |
| Observations | 2,182 | 2,182 | 7,306 | 7,306 |
[i] Note *p < 0.05; **p < 0.01; ***p < 0.001; 95% confidence intervals in brackets. Coefficients represent percentage-point differences in the probability of being employed relative to the reference category.
Table 4 disaggregates the disability sample by gender. For women with disabilities, both secondary year (0.04) and university year (0.05) interactions are positive and statistically significant in Model 2, indicating a comparatively stronger strengthening of education-employment links across waves. For men with disabilities, university × year is positive and significant (0.03), and secondary × year becomes positive and significant in the adjusted model (0.02). Thus, while higher education remains beneficial for both genders, the increase over time appears larger among women with disabilities, consistent with the gendered sectoral and institutional mechanisms discussed in the theoretical framework section.
Table 4
Changes in employment across time among women and men with disabilities, point estimates from LPMs.
| WOMEN | MEN | |||
|---|---|---|---|---|
| MODEL 1 | MODEL 2 | MODEL 1 | MODEL 2 | |
| Secondary education (Ref. primary education) | 0.08** | 0.07* | 0.12*** | 0.07* |
| (0.02, 0.14) | (0.02, 0.13) | (0.06, 0.18) | (0.01, 0.13) | |
| University education | 0.19*** | 0.18*** | 0.23*** | 0.19*** |
| (0.13, 0.24) | (0.12, 0.24) | (0.18, 0.29) | (0.13, 0.25) | |
| Year | –0.04*** | –0.04*** | –0.01 | –0.01 |
| (–0.05, –0.03) | (–0.05, –0.03) | (–0.02, 0.01) | (–0.03, 0.0001) | |
| Age | – | 0.001** | – | –0.01*** |
| – | (0.0003, 0.002) | – | (–0.01, –0.01) | |
| Ethnic minority (Ref. ethnic majority) | – | –0.11*** | – | –0.08*** |
| – | (–0.15, –0.08) | – | (–0.11, –0.05) | |
| Urban area (Ref. rural area) | – | 0.02* | – | –0.01 |
| – | (0.002, 0.04) | – | (–0.03, 0.002) | |
| Secondary education*Year | 0.03*** | 0.04*** | 0.01 | 0.02** |
| (0.03, 0.04) | (0.03, 0.05) | (–0.004, 0.02) | (0.01, 0.03) | |
| University education*Year | 0.05*** | 0.05*** | 0.02** | 0.03*** |
| (0.04, 0.06) | (0.04, 0.06) | (0.01, 0.03) | (0.01, 0.04) | |
| Constant | 0.54*** | 0.49*** | 0.54*** | 0.87*** |
| (0.49, 0.60) | (0.42, 0.56) | (0.48, 0.60) | (0.80, 0.94) | |
| Observations | 1,138 | 1,138 | 1,044 | 1,044 |
[i] Note: *p < 0.05; **p < 0.01; ***p < 0.001; 95% confidence intervals in brackets. Coefficients represent percentage-point differences in the probability of being employed relative to the reference category.
In Table S3 of the Supplementary Materials, we included the survey year as a categorical variable (instead of the continuous measure) to examine changes in the education-employment relationship across different rounds of the ESS between 2002 and 2018. No other variables were modified from the main models. These results closely matched the main findings. While some weakly significant effects appeared in the 2008–2012 waves, the overall pattern showed that the relationship between education and employment strengthened after the 2014 wave, especially for persons with disabilities.
Robustness checks of the main findings
To test the reliability of the main results, we conducted a series of additional tests using alternative model specifications of the key variables in Tables S4–S7 in the Supplementary Materials. First, we considered the severity of disability by separating respondents who reported being limited ‘a lot’ from those limited ‘to some extent’. This was done to examine whether the association between education and employment differs between individuals with more severe and those with milder disabilities. Second, we recoded the education variable into a binary form. Individuals with lower or higher tertiary education (categories V1 and V2 in the ESS education variable) were combined into one group and coded as 1, representing tertiary education. The remaining categories (I to IV) were coded as 0, representing less than tertiary education. The results of these checks were consistent with the main findings. The positive association between education and employment remained statistically significant for both persons with severe and mild disabilities. Similarly, when using the binary education variable, the main pattern persisted. As in the original models, we found significant changes over time in the link between education and employment, regardless of gender or disability status. In Tables S8–S11 in the Supplementary Materials, we also fit the logit and probit models for the main results and demonstrate that the substantive significance of the findings remains the same as in the LPMs.
Discussion
This study examined the relationship between educational attainment and employment outcomes among individuals with disabilities in Norway, with a focus on gender differences and temporal trends. Existing scholarship has demonstrated that education generally enhances employment prospects; however, few studies have examined how this relationship may differ for individuals with disabilities. Even fewer studies have examined whether gender influences these associations. Using data from the ESS collected between 2002 and 2018, this study helps fill this gap by examining whether education plays a similar or differential role for people with and without disabilities and whether the effects differ for men and women with disabilities or change over time.
Our results provide strong evidence that education is positively associated with employment among individuals with disabilities, consistent with human capital theory. Notably, this effect appears even more pronounced for individuals with disabilities compared to those without. The Norwegian context, with its relatively inclusive educational system, egalitarian labour market, and strong anti-discrimination legislation, may partially explain this finding. The data suggest that education enables labour market participation among persons with disabilities, particularly among those who attain tertiary education. This finding provides empirical support for Hypothesis 1, which posits that education is positively associated with employment among individuals with disabilities and that this association is at least as strong as among individuals without disabilities.
However, the gendered patterns of this association reveal additional complexity. In contrast to Hypothesis 2, which anticipated that education would benefit men with disabilities more strongly, we find that higher education is more strongly linked to employment among women with disabilities than among men. Several potential mechanisms may help explain this finding. One important explanation is gendered horizontal labour market segregation. In Norway, university credentials often function as formal entry tickets to public-sector and other credential-intensive occupations, especially in health, education, and care, where women are disproportionately employed. These sectors tend to have more standardised hiring rules, stronger union protection and greater institutional experience with accommodations, which can make it easier for women with disabilities to translate educational qualifications into employment (Kuznetsova et al. 2017; Mustosmäki et al. 2021). By contrast, men with disabilities are more likely to be oriented toward male-dominated private-sector or physically demanding jobs, where expectations of bodily strength, uninterrupted full-time availability, and informal recruitment practices may intensify ableist exclusion and reduce the marginal employment payoff of education. The greater absolute employment gains among women with disabilities should therefore not be read as evidence that they face fewer barriers overall; rather, tertiary education may provide a more effective route into comparatively more accommodation-friendly segments of the labour market. Although the ESS data do not allow us to test this mechanism directly with detailed sectoral or occupational indicators, the interpretation is consistent with the gendered institutional structure of the Norwegian labour market and existing Nordic research.
At the same time, our findings should be interpreted within a broader context of persistent gender inequalities. Although women with disabilities may gain more from education in terms of employment probability, this does not necessarily translate into equal experiences at work or equal pay. As suggested by earlier research, gendered expectations and structural discrimination may still result in horizontal segregation and wage disparities. Therefore, the strong education-employment link may coexist with ongoing disadvantages in other dimensions of labour market outcomes not explored in this study. In line with research on gendered welfare states and gendered labour markets, these patterns suggest that the Norwegian context may simultaneously enable relatively strong employment returns to education for women with disabilities and reproduce horizontal segregation, pay gaps, and constrained advancement prospects (Orloff 2009; Mustosmäki et al. 2021). Our interpretation is that higher education opens doors to employment for women with disabilities, but primarily into positions structured by enduring gendered and ableist hierarchies.
Hypothesis 3 anticipated a temporal change in the education-employment association for persons with disabilities. Our estimates indicate a consistent strengthening over time, not a weakening. In the pooled disability/non-disability models, the education × year interactions are positive for both secondary and university education, with larger increases among respondents with disabilities, together with the negative main effect of year. This implies that while overall employment probabilities edge downward, higher education increasingly offsets that decline, especially for groups with disabilities. Gender-stratified models show that this strengthening is more pronounced for women with disabilities than for men, reinforcing the intersectional pattern highlighted above. In short, the educational gradient in employment has widened across waves, particularly within the population with disabilities, rather than converged. These findings align with the interpretation that, in credential-intensive and comparatively inclusion-oriented sectors, tertiary education has become an ever more important entry ticket for persons with disabilities, even as structural barriers persist. Wave-indicator models tell the same story, with later rounds showing clearer strengthening of educational differentials, especially post-2014. Hence, we reject the expectation of declining returns over time and instead read the evidence as pointing to increasingly conditional equalisation: education matters more for securing employment among persons with disabilities, but it does not, by itself, remove the constraints that shape access and progression.
Furthermore, the widening educational gradient, particularly for persons with disabilities and, most clearly, among women with disabilities, suggests that tertiary and upper-secondary credentials have become increasingly important ‘entry tickets’ to employment in Norway’s credential-intensive labour market. On the other hand, the fact that stronger educational returns do not translate into full convergence implies that inclusion policies have not fundamentally altered baseline disadvantages or the institutional and cultural filters through which education is converted into jobs. Persistent employer stereotypes about productivity and reliability, the salience of uninterrupted full-time norms, and the concentration of many workers with disabilities in particular sectors or contractual forms likely continue to limit the transformative potential of formal qualifications. In this sense, equalising access to education remains necessary but insufficient; reducing the gap requires parallel efforts to change workplace practices and opportunity structures so that rising credentials yield equitable employment across groups.
In turn, Hypothesis 4 predicted that any increase in the education-employment association over time would be more pronounced for men with disabilities than for women. Our results indicate the opposite. In the gender-stratified models and the wave-indicator specification, the strengthening of education’s association with employment is larger for women with disabilities than for men. A plausible explanation is the alignment of tertiary credentials with credential-intensive and comparatively inclusion-oriented segments of the Norwegian labour market (public sector; health, care, and education), where formal qualifications function as entry tickets and accommodation is more routinised. Family policies that support continuous attachment may further help highly educated women with disabilities convert schooling into employment. By contrast, men with disabilities are more often positioned in male-dominated or physically demanding private-sector jobs where continuous full-time availability and informal recruitment practices can mute the marginal payoff of additional schooling. While sample size limits precision in some subgroups, the direction and pattern are consistent, leading us to reject Hypothesis 4.
Overall, our findings reinforce the idea that while education can be a powerful equaliser in employment prospects for persons with disabilities, its effectiveness is conditioned by both gender and broader social structures. Policies aimed at reducing employment disparities for this population should continue to focus on improving access to higher education, ensuring that educational credentials are recognised and rewarded in the labour market, and promoting inclusive workplace environments. In addition, greater efforts are needed to address intersectional forms of disadvantage, particularly the compounded effects of disability and gender, which can shape not only employment chances but also the quality and sustainability of jobs attained.
Limitations
Several limitations of this study should be acknowledged. First, disability was measured using a self-reported question on activity limitation, which may not capture the full range of impairments or reflect consistent interpretations across individuals. Second, the analyses rely on repeated cross-sectional ESS data rather than longitudinal follow-up. As a result, we cannot trace individual trajectories or make strong causal claims about the effects of education on employment; unobserved characteristics, such as early-life health, cognitive resources, family background, or employer selection, may influence both educational attainment and labour market outcomes. In addition, our models do not include detailed occupational or sectoral indicators, union coverage, household composition, marital status, the presence of children, or direct measures of caregiving responsibilities. This means that our interpretation of gendered horizontal labour-market segregation, public-sector pathways, and gendered caregiving expectations remains theoretically grounded rather than directly testable with the present data. In particular, because family structure and actual caregiving burdens may shape both educational attainment and labour market attachment, they represent potential sources of unobserved confounding in our gender-stratified results. Consequently, our findings should be interpreted as documenting systematic associations in employment probabilities across educational levels and social groups. Third, the binary nature of the employment outcome does not account for variations in job quality, security, hours worked, or occupational status, which are important dimensions of labour market inequality. Lastly, the relatively small number of respondents with disabilities, especially when stratified by gender and time, reduces statistical power and may obscure smaller or subgroup-specific effects.
Conclusion and Future Research
This study investigated the relationship between educational attainment and employment outcomes among individuals with disabilities in Norway, with a focus on gender and temporal variation. The findings confirm that education remains a strong predictor of employment and that this association is at least as strong, and often stronger, among persons with disabilities than among those without. In particular, women with disabilities appear to gain more from higher education than men, consistent with sectoral dynamics and institutional support structures discussed above. Across multiple specifications, including pooled disability/non-disability models, gender-stratified models, and a wave-indicator specification, we observe a consistent strengthening of the education-employment association between 2002 and 2018, implying a widening educational gradient, most pronounced among persons with disabilities and especially among women with disabilities. While sample sizes limit precision for some subgroups, the overall direction is clear.
These findings highlight the conditional nature of education’s equalising role. Even in an inclusive welfare state like Norway, the conversion of credentials into employment remains shaped by gendered labour market structures, employer expectations, and enduring forms of discrimination. As the returns to education become more important for securing work, particularly for women with disabilities, reducing disparities requires more than expanding access to higher education. Policy should also target workplace practices and opportunity structures (e.g., accommodation capacity, hiring and retention norms, and the organisation of work) so that rising credentials translate into equitable employment across groups.
Future research should explore these mechanisms in greater depth. Longitudinal administrative data could provide more robust insight into employment trajectories, job quality, and cumulative disadvantage over time. In addition, qualitative studies are needed to understand how persons with disabilities navigate the school-to-work transition and interpret their labour market experiences. Comparative work across welfare state regimes would also help to identify which institutional conditions best support the equitable translation of education into employment. Finally, future studies should consider a broader set of intersecting inequalities, such as migration backgrounds and regional belongings, to further unpack the complexity of disability-based disadvantage in the labour market.
Additional File
The additional file for this article can be found as follows:
Notes
[5] Throughout the article, we use person-first language as our default formulation and therefore refer primarily to persons with disabilities, consistent with the rights-based language of the CRPD and the journal’s general guidance. We retain shorter formulations, such as disabled women and disabled men, only as concise subgroup labels in the analysis and when referring to scholarship informed by the social model of disability.
Data Accessibility Statements
The data used in this study are openly available from the European Social Survey data portal: https://www.europeansocialsurvey.org/data-portal.
Ethics and Consent
This study is based exclusively on secondary analysis of publicly available, fully anonymised data from nine waves of ESS. The authors did not interact directly with human participants and did not access personally identifiable information. Accordingly, under local laws and institutional guidelines, no additional formal institutional ethical approval was required for this study. The original ESS data collection was conducted in accordance with applicable national and ESS European Research Infrastructure ethical and data-protection requirements, and all respondents were informed about the study and provided informed consent prior to participation.
Author Contributions
Both authors contributed to the study conception and design. Data access, data management, and analysis were performed by Edith Karlsen. The first draft of the manuscript was written by Edith Karlsen and Alexi Gugushvili, and both authors commented on previous versions of the manuscript. Both authors read and approved the final manuscript.