Using population survey data for a large number of OECD countries, this chapter presents a set of indicators to measure the social and labour market inclusion of people with disability and compare outcomes across countries and over time. Findings are mixed: while people with disability are more likely today to achieve a higher level of education, disability gaps in both employment and unemployment remain high and largely unchanged. Similarly, the high disability poverty gap has increased further, even though the large majority of people with disability who have no job receive some form of income support. Overall, these outcomes suggest that the current policy approach fails to generate highly needed improvements in the labour market position of people with disability.
Disability, Work and Inclusion
2. Labour market inclusion of people with disability: Where are we now?
Abstract
In Brief
Disability prevalence stands at around 18% but varies hugely across countries. The UN Convention on the Rights of Persons with Disabilities (CRPD) describes disability as a social construct resulting from the interaction between people with impairments and their (attitudinal) environment, and as a concept evolving over time. The measurement of disability is volatile and can be linked to cultural factors and self-perception, reflected in large differences in disability prevalence over time and across countries (Section 2.1).
The education distribution of people with disability is lagging behind by at least 15 years to that of people without disability. Despite the education improvements for people with disability, the education gap with people without disability remains large. Disability also remains a high risk factor for early school leaving (Section 2.2).
The employment rate of people with disability has slightly improved over the past decade on average across OECD countries. Overall, employment rates of people with disability are low, with only 40% of them having a job on average across OECD countries. In most countries, the employment rate of people with disability is higher now than it was in 2008, recovering the pre‑GFC crisis levels. More recent data are needed to assess whether this recovery will be jeopardised in the post-COVID‑19 era and in the current economic context (Section 2.3).
Taken together, the improvements in the employment rate of people with disability were not sufficient to close the disability employment gap. On average across OECD countries, people with disability are about 40% less likely to be in employment than are people without disability, a gap that has remained constant over the past decade (Section 2.3).
Across OECD countries, the large majority of adults with disability that are not working receive social transfers. On average, 70% of people with disability not at work receive a social benefit. Most of these persons are not receiving disability benefits, contrary to what could be expected. On average, only 30% of people with disability not working are receiving disability benefits. The remaining 40% receive a mix of old-age pensions, unemployment benefits, and – very importantly – social assistance (Section 2.4).
Disability remains a major poverty driver in most OECD countries. Despite comprehensive benefit systems and high shares of people with disability receiving benefit, i.e. high coverage rates, poor employment inclusion implies that on average across a large set of OECD countries, one in four people with disability live in a household with income lower than 60% of the median, a share that has increased slightly in the past decade (Section 2.5).
2.1. The size of the population with disability varies hugely across cultures
Around one in six persons across the OECD report having a disability (Figure 2.1). Differences in disability prevalence, however, can be large, both across countries and within countries, across men and women, age groups and educational backgrounds (see Box 2.1 on disability definitions and measurement):
Differences across countries are striking (Panel A). Measurement differences are only a small part of the explanation as most surveys use the same questions. Differences in stigma, self-stigma, perception, culture, attitudes and awareness all play an important part in explaining the variability across countries and also within countries over time, in addition to health itself. The exclusion of mental health conditions in the screening instrument contribute to the exceptionally low disability prevalence of 3% in Korea and the poor recognition of mental health conditions in other countries’ instruments (Chile, Mexico, United States) also lowers the prevalence in those countries.
Further country-specific disaggregation, not shown in the figure, suggests that disability prevalence has increased over the past 15 years in 18 out of the 26 countries for which 2005 data are available. This increase is caused, in part, by a larger share of the 50‑69 year‑olds in the working-age population; they were 32% of the working-age population in 2005 against 38% in 2018.
Another visible distinction relates to gender. Panel B shows on average a difference in disability prevalence between men and women of 3 percentage points, for a number of reasons. Firstly, mental health conditions such as depression and anxiety, a growing cause of disability, impact women more often than men. These conditions have grown in importance to the point of now being one of the main causes of disability. Secondly, and potentially due to perception and (self-)stigma, men are less likely to report having a disability than women.
Older people are more likely to have a disability (Panel C). However, among young people (aged 15‑29), also about 8% on average in the OECD report having a disability. They represent a group of people who may not have had the time to contribute enough to create entitlement social benefits. A group of people whose disability may have prevented them from finishing their studies or entering the labour market with the same ease as their peers, making their disability a double burden.
There is a strong negative correlation between education and disability (Panel D). On average, 25% of low-educated individuals have a disability compared to 11% of high-educated individuals. Those with low level of education also more commonly face instable labour market conditions, a higher risk of job automation, lower wages and riskier or more physically demanding jobs.
There is also considerable heterogeneity within the group of people with disability (PWD), related to the type of disability, the intensity or severity, the cause and the duration or permanence. The high prevalence of disability (on average 18%) only underlines the importance of including PWD in all aspects of our society and especially the labour market, work being a pillar of independence, purpose, self-esteem and social balance. The heterogeneity makes providing the right support for everyone a challenge. It also makes it an opportunity; the opportunity to provide individualised support for everyone, disability or not, while following the strictest definition of mainstreaming. The lack of commonality between two PWD means that support must be provided according to need instead of status and that the same support can be provided to several people all of whom may or may not have a disability and still benefit from it.
Figure 2.2 illustrates that in Europe the prevalence of disability is growing. Certain intangible factors such as cultural differences and changes prevent from drawing hard conclusions from such time trends, similarly to cross-country comparisons. Population ageing explains about half of the overall increase. However, disability prevalence has increased significantly among young adults (aged 15‑29) and, to a lesser degree (and only until 2013), among those with medium and higher levels of education. Thus, disability now seems to affect more people than about 15 years ago. This suggests disability is more likely today to lead to exclusion or disadvantage, or felt as a greater social and labour market barrier despite the considerable efforts made by many countries to follow recommendations from OECD’s previous work, 20 years ago.
Box 2.1. Defining disability and identifying people with disability
People with disability, according to the UN Convention on the Rights of Persons with Disabilities (CRPD), include those who have long-term physical, mental, intellectual or sensory impairments, which in interaction with various barriers may hinder their full and effective participation in society on an equal basis with others. This reflects the understanding of disability as a social construct resulting from the interaction between people with impairments and their (attitudinal) environment. The CRPD further recognises that disability is a concept that is evolving over time.
The understanding of disability as a social construct, which is subject to cultural differences as well as changes over time, makes disability measurement and comparisons across time and between countries difficult. However, it also suggests individuals themselves can judge best if they should count towards the group of people with disability at a particular moment. In turn, subjective disability measurement provides a meaningful proxy to assess social and labour market disadvantages arising from disability – even though any comparisons must keep in mind underlying differences in disability prevalence.
Identifying disability in population surveys
Population surveys identify people with disability through a set of disability screening questions. The type of questions used vary across countries and surveys but two concepts are especially widespread. One concept (option 1) uses questions about difficulties people may have when performing activities because of a health problem in different life domains such as vision, hearing, mobility, cognition, self-care and communication, using the WHO’s International Classification of Functioning, Disability, and Health (ICF) as a conceptual basis. These instruments measure both type and severity of disability. A frequently used alternative (option 2) is an instrument consisting of only two questions: the first asking whether people have any permanent or long-standing illness or health problem, and the second asking about the degree to which the long-standing illness or health problem limits activities people usually do. While this instrument allows for a distinction in severity of disability, it does not measure disability type.
Comparability of data from European and non-European countries
EUROSTAT uses option 2 in several surveys, including the European Union Statistics on Income and Living Conditions (EU-SILC) and the European Working Conditions Survey (EWCS) used in this report. Data for other OECD countries are based on national surveys, which all use variants of one of the two disability screening options. For Canada, data come from the Canadian Survey on Disability (CSD), the Canadian Income Survey (CIS), and the General Social Survey (GSS), all using a variant of option 1. Canada is a special case as the same disability-screening instrument is now used in several population surveys. This is a more general trend which the United Kingdom is also currently aiming to follow. For Australia, data come from the Household, Income and Labour Dynamics in Australia Survey (HILDA) using an option‑2 type screening instrument similar to the one used in the European surveys and the Survey of Disability, Aging and Carers (SDAC) using an option‑1 screening instrument. The surveys for Chile (National Socio-Economic Characterization Survey/CASEN), Mexico (National Survey of Household Income and Expenditure/ENIGH) and the United States (American Community Survey/ACS) use the ICF-based option‑1 approach. People with mental health conditions are only partly covered with these instruments. Data for Korea, which come from the Korean Labour and Income Panel Study (KLIPS), also use an option‑1 screening tool but only include people with a persistent physical limitation or disability; mental health conditions are excluded. Finally, income data for the United States are from the Current Population Survey (CPS) which uses its own screening taxonomy (PWD are those who: ever retired or left a job for health reasons; are not in the labour force because of a disability; did not work in the previous year because of illness or disability; are under 65 years old and covered by Medicare or receiving Supplemental Security Income or Veterans’ Allowance in the previous year).
Figure 2.1. Disability prevalence stands at around 18% but varies hugely across countries
Note: OECD is the unweighted average of the 32 countries shown. Based on different sources for European and non-EU countries, data are not fully comparable. European countries define people with disability as those who 1) declared to suffer from any chronic illness or condition and 2) with moderate to severe activity limitation due to health problems. For Australia, persons who declared having a long-term health condition, impairment or disability that restricts them in their everyday activities, lasting for six months or more. For Canada, persons who report a limitation in their day-to-day activities due to difficulty in 1) seeing, hearing, mobility, flexibility, dexterity, pain-related, learning, developmental, mental health-related or memory issues, or 2) because of any other long-term health condition. For Chile, Mexico and the United States persons who reported having difficulty in: 1) Walking, moving around, going up or down stairs; 2) Vision, even when wearing glasses; 3) Talking, communicate or exchange (and difficulty doing errands alone, United States); 4) Hearing, even with a hearing aid; 5) Dressing, bathe or eat; or 6) Concentrating or learn simple things. For Korea, persons who declared having any persistent physical limitations or disability: 1) Visual, auditory problems or speech impediment, 2) Difficulties in physical activities, 3) Difficulties in learning; 4) Difficulties in indoor activities; 5) Difficulties in outdoor activities; 6) Difficulties in working. Levels of education defined according to the International Standard Classification of Education (ISCED). Low refers to below upper-secondary, Medium to upper secondary and High to tertiary education. Data refer to 2016 (Mexico), 2017 (Canada), 2016‑17 (Chile) and 2016‑18 (Belgium, Iceland, Ireland, Italy, Korea, United Kingdom, United States) and 2018 (Australia).
Source: OECD calculations based on the European Union Statistics on Income and Living Conditions (EU-SILC) for European countries; Survey of Disability, Ageing and Carers, Australia: Summary of Findings, 2018 (Table 3.1); the Canadian Survey on Disability (CSD, 2017) provided by Employment and Social Development Canada; Chile’s Encuesta de Caracterizacion Socio-economica Nacional (CASEN, 2016‑17). Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH 2016); the Korean Labour & Income Panel Study (KLIPS, 2016‑18) and the American Community Survey (ACS, 2016‑18).
Figure 2.2. Looking behind the increase in disability prevalence: the role of age and education
Note: Data cover persons aged 15‑69 and show the weighted average of 20 European countries: Austria, Belgium, the Czech Republic, Denmark, France, Greece, Hungary, Iceland, Ireland, Latvia, Lithuania, Luxembourg, the Netherlands, Norway, Poland, Portugal, the Slovak Republic, Slovenia, Spain and Sweden. Panel A: To control by age the prevalence of disability was generated using five‑year age groups in 2006 and applied to the following years to simulate the number of people with disabilities that would have existed if the same prevalence by age group and age proportions were maintained. This same process was used to control for age and education in Panels B and C. Disability is defined as people who 1) declared to suffer from any chronic illness or condition and 2) with moderate to severe activity limitation due to health problems.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC).
2.2. Education improvement is not enough to close the education gap
2.2.1. Continuous improvement in educational attainment
Higher levels of educational attainment increase the likelihood of being employed, and for maintaining and upgrading the skills to maintain and progress in employment (OECD, 2021[1]). It is thus encouraging to see that across all countries, the share of PWD with low level of education has fallen from around 48% to about 30% from 2005 to 2019 (Figure 2.3, Panel A). The share of people without disability (PWOD) with low level of education decreased from about one‑third to one‑fifth over the same period. Accordingly, while the disability education gap has closed slightly, the share of PWD with a low education remains higher than among PWOD by about 10 percentage points.
At the same time, data show a large difference between PWD and PWOD in the growth in the share of the population with a high level of education at tertiary level. Among PWOD, the share with mid-level education – corresponding to completed secondary education – remained rather stable (Figure 2.3, Panel B). Among PWD, however, much of the drop in low education translated into an increase in med-level education; the increase in the share of people with high education was much slower than for PWOD. While there is a high demand for workers with vocational education, this development may indicate that PWD could be hit harder in labour markets characterised by a high degree of job polarisation, i.e. a loss of middle‑skilled jobs.
People with severe disability are least likely to achieve the highest level of education, with a share of about 20% with post-secondary education across the six selected countries and slightly lower than this for Norway, compared to about 40% among PWOD and a share slightly higher than this in Canada (Figure 2.3, Panel C). The share of people with a medium level of education is almost the same for PWD and PWOD in all six countries and highest in Austria and Switzerland, at over 50%.
Overall, the level of education has gradually improved between 2005 and 2019, for both PWD and PWOD (data are not available yet on the impact of the pandemic). Nevertheless, PWD remain largely at the same relative education and skills disadvantage. It appears that the education improvements for PWD have been just fast enough to keep the disability employment gap at roughly its current high level.
2.2.2. However, disability remains a risk factor for early school leaving
While levels of educational attainment of PWD have improved, Figure 2.4 shows why closing the education gap has remained and will remain a challenge. One in five young people with disability aged 15‑29 leaves school without completing a secondary degree, here interpreted as an indicator of early school leaving, compared to only one in ten among young people without disability (Panel A). A deeper look into these differences indicates large variation by severity of disability: 15% and over 35% of those with moderate and severe disability, respectively, leave school early on average across OECD countries. The particular disadvantage of young people with severe disability is found in most countries, and in some – including Lithuania, Portugal and Spain – this share can be as high as 60%. In a few countries, like the United Kingdom and the United States, however, early school leaving is rare for all groups.
Figure 2.4 also shows that the transition into the labour market is difficult for PWD. One in three young people with disability and even one in two among those with severe disability belong to the group who are not employed, not in education and not in training, commonly known as NEETs (Panel B). This compares with one in seven young people aged 15‑29 without disability. It is noteworthy that, on this indicator of labour-market opportunity, young people with moderate disability are doing much worse than those without disability (both on average and in most OECD countries).
Figure 2.3. Education levels of people with disability are improving but not enough
Note: Data cover persons aged 15‑69. Levels of education defined according to the International Standard Classification of Education (ISCED). Low refers to below upper-secondary, Medium to upper secondary and High to tertiary education. Panels A and B are weighted averages of 26 European countries: Austria, Belgium, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, the Netherlands, Norway, Poland, Portugal, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, United Kingdom. Panel B: The difference is calculated from 2018 (Belgium, Iceland, Ireland, Italy) and 2016 (United Kingdom). Panel C: Data refer to the average over 2016‑18 for Belgium and to 2017 for Canada; for exceptions and for country definitions of people with disability, see Figure 2.1.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC 2005‑19) for European countries and data provided by Employment and Social Development Canada based on the Canadian Survey on Disability (CSD, 2017).
Figure 2.4. The transition from school to work is more difficult for young people with disability
Note: Data for Canada refer to 2017. No data available in 2019 (Belgium, Iceland, Ireland, Italy, United Kingdom) and in 2018 (Estonia). The purple markers are an unweighted average of the 32 countries shown. Panel A: Early school leavers are defined as persons aged 15‑29 who are not in education and do not have an upper secondary school diploma. Panel B: NEETs are defined as persons aged 15‑29 who are not in education, employment or training. NEETs rates based on other data sources seem to suggest that the data in the figure overestimate the NEET levels in some countries, especially Norway and Italy. Disability gaps are likely less affected by the choice of the data source.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2016‑19) for European countries. Household, Income and Labour Dynamics in Australia Survey (HILDA, from 2016‑17), the Canadian Survey on Disability (CSD, 2017) provided by Employment and Social Development Canada, Chile’s Encuesta de Caracterizacion Socio-economica Nacional (CASEN, 2017), Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2016), the Korean Labor & Income Panel (KLIP) and the American Community Survey (ACS).
2.3. Labour market outcomes have changed little in the past 15 years
2.3.1. Small improvement lately in labour supply
Labour force participation rates measure the labour supply of the population, i.e. the share of the population that has a job or is looking for a job. Labour supply has changed during and after the global financial crisis, in line with the strength of the economy, for both PWD and PWOD and with limited changes in the labour supply gap between PWD and PWOD. Only in the past few years before the COVID‑19 crisis, labour supply has grown faster for PWD than for PWOD, but only slightly so and not in all OECD countries. However, the small reduction in the disability gap in labour force participation has led to an increase in the unemployment gap in many cases, reflecting the challenges in accessing the labour market.
2.3.2. But high levels of unemployment and almost unchanged levels of employment
Employment rates of PWD and the gap in employment rates between PWD and PWOD have changed little on average across the OECD area and also in the majority of countries (Figure 2.5, Panels A and B). The crisis of 2008‑09 has affected employment rates of PWD more negatively than the rates of PWOD but in the past 3‑4 years prior to the COVID‑19 crisis, PWD have been able to catch up a little, just so that the disability employment gap is back to a very similar level overall. Across 32 OECD countries for which comparable data are available, in 2019 about 42% of PWD were employed. The share ranges from less than 30% in Greece, Korea, Spain and Ireland to 54% in Canada and 58% in Switzerland. The disability employment gap, measured as the difference in the employment rate between PWOD and PWD over 2016‑19, was 27 percentage points on average across all 32 countries, ranging from around 15 percentage points in Mexico, Chile and Switzerland to over 35 percentage points in Lithuania, the United States and Ireland. Only a few countries, including Chile, the Czech Republic, Denmark, Lithuania and the United Kingdom, have seen visible improvements in both the employment rate of PWD and the disability employment gap. Improvements in the employment rate of PWD but without narrowing the disability employment gap were also observed in Poland, Hungary, Mexico, Estonia and Latvia. On the other end, some countries, including Germany, Greece, Ireland, Italy, the Slovak Republic, Spain and Sweden, have seen a noticeable increase in the disability employment gap in the past few years.
In 2019, across the same set of countries, 15% of PWD were unemployed, the same unemployment rate overall as a decade ago (Figure 2.6, Panel A). However, unemployment rates of PWD have fallen very considerably in a few countries since 2008 (Belgium, the Czech Republic, Hungary, Poland) and increased in many others (Norway, the Slovak Republic, Sweden, Portugal, Finland, Italy, Greece, Spain). The disability unemployment gap, measured as the difference in unemployment rates between PWD and PWOD, stood at 8.6% on average across the same set of countries in the period 2016‑19, very similar to the level five years earlier (Figure 2.6, Panel B). The disability unemployment gap has increased most noticeably for Ireland and Sweden, and is now highest in Germany. Today, PWD are on average more than twice as likely to be unemployed than PWOD, a higher ratio than observed before and soon after the global financial crisis, suggesting that it has become even more difficult for PWD to find a job.
In countries like the Czech Republic, France and Switzerland, the employment and unemployment rates of PWD have both increased because of the increase in labour supply. This has not been the case for all countries. In Germany, Austria, Italy and Sweden, for instance, the gap between PWD and PWOD grew larger for both employment and unemployment rates. Only five countries, Canada, Chile, Iceland, Lithuania and the Netherlands, have seen a decrease in both their employment and unemployment gaps for PWD compared to PWOD. For more country-specific details on trends in employment and unemployment rates by disability status, see Annex Figure 2.A.1 and Annex Figure 2.A.1. Lacking improvements in employment rates of PWD and the unchanged disability employment gap are disappointing in view of the improvement over the past 15 years in the level of educational attainment of PWD. This suggests that the necessary policy transformation that many countries have started over the past decade has not gone far enough.
Figure 2.5. Disability employment gaps are large and have changed little in the past decade
Note: Data cover persons aged 15‑69. For country definitions of people with disability, see Figure 2.1. Exceptions Panel A: Year 2019 refers to 2014 (Korea), 2016 (Mexico), 2017 (Australia, Chile) and 2018 (Belgium, Iceland, Ireland, Italy, United Kingdom, United States). Exceptions Panel B: Periods refer to 2013‑16 and 2017‑19 for Canada, 2016‑18 (Belgium, Iceland, Ireland, Italy, United Kingdom) and no data in 2018 (Estonia). The purple bars represent the unweighted average of the 26 countries shown (excluding Korea). For Australia, data presented are based on the Household, Income and Labour Dynamics in Australia Survey (HILDA). When using data from the Australian Bureau of Statistics’ Survey of Disability, Ageing and Carers (SDAC), the employment rate of persons with disabilities in 2018 is slightly higher at 48%, and the gap in employment due to disability is also higher at 32 percentage points.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2005‑19) for European countries; the Household, Income and Labour Dynamics in Australia Survey (HILDA, 2005‑17); Chile’s Encuesta de Caracterizacion Socioeconomica Nacional (CASEN, 2006‑17); Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2010‑16); the Korean Labour & Income Panel Study (KLIPS, 2008‑18) and the American Community Survey (ACS, 2008‑18). Data for Canada provided by Employment and Social Development Canada based on the Canadian Income Survey, 2013‑19.
Figure 2.6. Disability unemployment gaps have remained very high in most OECD countries
Note: Data cover persons aged 15‑69. For country definitions of people with disability, see Figure 2.1. Exceptions Panel A: Year 2019 refers to 2014 (Korea), 2016 (Mexico), 2017 (Australia, Chile) and 2018 (Belgium, Iceland, Ireland, Italy, United Kingdom, United States). Exceptions Panel B: Periods refer to 2013‑16 and 2017‑19 for Canada, 2016‑18 (Belgium, Iceland, Ireland, Italy, United Kingdom) and no data in 2018 (Estonia). The purple bars represent the unweighted average of the 26 countries shown (excluding Korea). For Australia, data presented are based on the Household, Income and Labour Dynamics in Australia Survey (HILDA). When using data from the Australian Bureau of Statistics’ Survey of Disability, Ageing and Carers (SDAC), the unemployment rate of persons with disabilities in 2018 is slightly higher at 10%, and the gap in unemployment due to disability is also higher at 5.7 percentage points.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2005‑19) for European countries; the Household, Income and Labour Dynamics in Australia Survey (HILDA, 2005‑17); Chile’s Encuesta de Caracterizacion Socioeconomica Nacional (CASEN, 2006‑17); Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2010‑16); the Korean Labour & Income Panel Study (KLIPS, 2008‑18) and the American Community Survey (ACS, 2008‑18). Data for Canada provided by Employment and Social Development Canada (Panel A) and OECD calculations (Panel B) based on the Canadian Income Survey (CIS, 2013‑19).
2.3.3. Severity of disability, gender, age and level of education effects
The severity of disability has a significant negative impact on the employment of people with disability (Figure 2.7, Panel A). As shown across OECD countries for which data is available in 2019, about one in four persons with severe disability was employed compared to one in two persons with moderate disability. The negative effect extends to unemployment, as people with severe disability were about 1.6 times more likely to be unemployed than people with moderate disability.
The employment gaps for men with disability, compared to men without disability, are generally larger than for women (Panel B). A larger disability gap for men can also be observed in the unemployment rates between people with and without disability, compared to women. However, gender differences in labour market outcomes are large for both PWD and PWOD, and employment rates for women are generally lower than for men, irrespective of the disability status.
The employment gap also increases with age (Panel C). Prime‑age PWOD (age 30‑49) are about 1.3 times more likely to be employed than PWD of the same age group. This difference is largest for the oldest group of workers (age 50‑69); at this age, PWOD at are about 1.8 times more likely to be employed. This suggests that PWD may find it more difficult to retain employment at that age. In contrast, age does not have a very strong effect on the difference in unemployment between PWD and PWOD; young adults (age 15‑29) have the highest unemployment rates among both groups, PWD and PWOD alike.
Disability has a strong negative impact on the employment rate of all people irrespective of their level of educational attainment but the impact seems strongest for people with medium level of education; across this set of countries, PWD with medium level of education appear to be employed as often as PWOD with low level of education (Panel D). The disability unemployment gap is similar for people with low and medium level of educational attainment, around 10 percentage points on average, and slightly lower than this, around 5 percentage points on average, for people with high level of educational attainment. This finding could also be suggesting that improving education and skills levels of PWD is an important element in any strategy to close the disability employment and unemployment gaps.
2.3.4. The impact of disability prevalence on labour market outcomes
While the messages from the comparative analysis of labour market outcomes seem clear, country-specific differences should be interpreted with some caution. The measurement of disability is volatile and linked to cultural factors and self-perception, reflected in large differences in disability prevalence over time and across countries (Figure 2.1). Disability gaps in employment rates and other social and labour market outcomes could be influenced by underlying differences in disability prevalence, e.g. if the average severity of disability varied. There is reason to assume that people identifying with disability in countries with a low disability prevalence have, on average, a more severe disability, for instance. Similarly, the interpretation of differences across countries or even within countries over time could be hampered if the likelihood to perceive a health condition as disabling varied with age or the level of education, or changed over time.
A simple way of correcting for prevalence effects, to assess and compare the size of employment gaps, is to weight the resulting disability gap by the country’s disability prevalence. In the case of the employment gap, such a measure can be interpreted roughly as the share of the population deprived of employment because of a disability. Figure 2.8 shows the result of this exercise. Weighted disability employment gaps range from less than 2% in Korea, Mexico and Italy, countries with a very low disability prevalence, to 4.5% on average across the 32 countries, and very similar levels of 5.1‑6.2% in half of the countries. Countries with lower-than-average disability prevalence, such as Norway, Sweden and the United States, appear to have a lower-than-average disability employment gap.
Figure 2.7. Employment and unemployment levels vary across socio‑economic characteristics
Note: PWD: People with disability; PWOD: People without disability. The purple markers represent the unweighted average of 26 European OECD member countries and Canada. Data refer to 2017 (Canada, Panels A and B) and 2018 (Belgium).
Source: OECD calculations based on the European Union Statistics on Income and Living Conditions, 2019. Data provided by Employment and Social Development Canada based on the Canadian Survey on Disability, 2017 (Panels A and B); Canadian Income Survey, 2019 (C and D).
Figure 2.8. Accounting for disability prevalence reduces country differences in the disability employment gap but also blurs the impact of disability on employment outcomes
PWD: People with disability; PWOD: People without disability.
Note: Data cover persons aged 15‑69. For country definitions of people with disability, see Figure 2.1. Exceptions: periods refer to 2013‑16 and 2017‑19 for Canada, 2016‑18 (Belgium, Iceland, Ireland, Italy, United Kingdom) and no data in 2018 (Estonia). The Canadian data is weighted with the disability prevalence from 2017. The purple bars represent the unweighted average of the 26 countries shown (excluding Korea).
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2005‑19) for European countries; the Household, Income and Labour Dynamics in Australia Survey (HILDA, 2005‑17); Chile’s Encuesta de Caracterizacion Socioeconomica Nacional (CASEN, 2006‑17); Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2010‑16); the Korean Labour & Income Panel Study (KLIPS, 2008‑18) and the American Community Survey (ACS, 2008‑18). Data for Canada provided by Employment and Social Development Canada based on the Canadian Income Survey, 2013‑19 for employment data and the Canadian Survey on Disability (CSD, 2017) for disability prevalence.
The interpretation of trends within countries, however, is difficult. In Norway, for example, the weighted gap in 2016‑19 is around the OECD average but has increased quite considerably in the past few years while the unweighted gap, shown in Figure 2.5, suggested a high but unchanged disability employment gap. Hence, both levels and trends hinge on the measurement. Not knowing the reason for the decline in the prevalence rate – in this example, in Norway – limits the interpretation of the weighted gap. The volatility of the measurement of disability prevalence affects cross-country comparisons but is equally relevant across different surveys within a country. Australia is a good example in case. Australia commonly uses the SDAC survey, a dedicated disability survey, to generate evidence on disability prevalence and the situation of PWD. The survey uses a strict definition of disability and, therefore, identifies a relatively low share of people as PWD – about 14.5% of the working-age population, compared with 18% according to the HILDA survey which uses the same disability definition used in EU-SILC and, therefore, offers better comparability with other countries. Moreover, SDAC (in contrast to HILDA) overestimates employment participation considerably; SDAC estimates an employment rate for PWD of 48%, compared with 42% according to HILDA. The employment gap is also much larger according to SDAC (32 percentage points, compared with 25 percentage points with HILDA). There is no right or wrong: both results are valid and none is necessarily better than the other. Using the weighted indicator reduces the difference between the two surveys in the employment gap (around 5% with SDAC and 6% with HILDA), thus reducing the impact of the choice of the survey, but SDAC’s low prevalence rate seems to dominate the result.
2.3.5. Relatively minor discrepancies in employment characteristics and job quality
Once in employment disadvantages related to disability, or differences between people with and without disability, seem to be much smaller. PWD and PWOD do not differ much on most job characteristics. For instance, the share of people who are self-employed or working with a temporary contract hardly differs between PWD and PWOD. This is a confirmation of general labour market findings according to which country differences are larger than individual differences, due to the impact of employment protection legislation. Thus, the use of temporary employment for entry jobs is very common in some countries and not at all in others but there is no particular relationship with disability.
In most OECD countries, however, PWD are significantly more likely to work part-time. The reasons will include a mix of issues, including difficulties in accessing full-time employment and a preference of PWD for part-time employment. A flexible labour market facilitating part-time work can help PWD in accessing the labour market. On average across 32 countries with comparable data, one in five PWD work part-time compared with one in ten among PWOD. In Norway and Sweden, PWD are three times more likely to hold a part-time job, while in many countries with high part-time shares overall this ratio is lower despite a large difference (e.g. in the United Kingdom, the shares are 21% and 32% for PWOD and PWD, respectively). By contrast, in a few countries like the Slovak Republic and Portugal, the difference between PWD and PWOD in part-time employment shares is small.
The average wage of PWD is about 85‑87% of the average wage of PWOD. In some OECD countries, the disability wage penalty is (still) very large: PWD earns only about 65% of the wage of PWOD in Korea and around 75‑79% in Mexico, the Baltic countries, and also Sweden and Norway (Figure 2.9, Panel A). On the other end of the scale, in Austria, Australia and the Netherlands PWD earn 90% or more of the wage of PWOD and Greece and Italy appear not to have any disability wage penalty at all. However, in the latter two countries this is coupled with an exceptionally low employment rate among PWD; hence, selection effects seem to be at stake. Compared to eight years earlier, the disability wage penalty has increased in one‑third of the countries (e.g. Mexico, Hungary, Poland, Sweden, Luxembourg) and decreased in another third (e.g. Chile, Norway, Portugal, Italy), but has remained stable on average. An interpretation of the causes for these different trends is not possible on the basis of such aggregated data.
On the other hand, there are only very small differences between PWD and PWOD in weekly working patterns (Figure 2.9, Panel B). PWD tend to work more often on Sundays with a difference of 4 percentage points between PWD and PWOD. Roughly one in four PWD work in shifts, often including jobs with lower educational attainment, more so than among PWOD. Overall, once employed, however, work patterns are similar for both groups of workers. The shares of workers working on Saturdays, working overnight or working very long hours (10 or more hours per day) are similar among PWD and PWOD.
Figure 2.9. Wages are lower for people with disability, but working-hour patterns are similar
Note: Panel A: Data refer to annual employee wages employee cash or near cash income for employees and cash profits or losses from self-employment for persons self-employed (European countries); main labour income in cash (Chile); financial year gross wages and salary (Australia); average hourly wage excluding self-employed and the Canadian Armed Forces (Canada); total after-tax yearly earned income (Korea); main wage in main work, piece rates, commissions, payment for extra hours in main work, incentive pay, bonus, holiday pay and cash income second job (Mexico) and total wage and salary earnings (United States). Period 2016‑19 refers to 2012‑15 (Korea). The purple bar is the unweighted average of the countries shown excluding Canada and Germany which do not have data for the earlier period. For country definitions of people with disability, see Figure 2.1; Panel B: Data represent the unweighted average of 21 European countries: Austria, Belgium, the Czech Republic, Denmark, Estonia, Germany, Finland, France, Ireland, Latvia, Lithuania, Luxembourg, the Netherlands, Norway, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland, Türkiye and the United Kingdom). Shift: working in shifts. Night, Saturday, Sunday, More than 10 hours a day relates to normally working at least once a month this type of unsocial working hours.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2008‑19) for European countries. Exceptions: 2016‑18 (Belgium, Iceland, Ireland, Italy, United Kingdom). The Household, Income and Labour Dynamics in Australia Survey (HILDA, 2008‑17); the Canadian Income Survey (CIS, 2016‑19) provided by Employment and Social Development Canada; Chile’s Encuesta de Caracterizacion Socioeconomica Nacional (CASEN, 2010‑17); Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2016).; the Korean Labour & Income Panel Study (KLIPS, 2008‑18) and the United States Current Population Survey (CPS, 2008‑18). Panel B: European Working Conditions Survey (EWCS).
2.3.6. However, labour market dynamics are different for people with disability
On the other hand, data available for European OECD countries show that labour market dynamics differ significantly between PWD and PWOD (Figure 2.10):
Once employed, the likelihood of a job-to-job change during a year is similar (Panel A).
Dropping out of the labour is much more frequent for PWD than for PWOD (Panel B).
Getting into the labour market is much more difficult for PWD than for PWOD (Panel C).
Country-specific differences in labour mobility are large, reflecting differences between countries in the dynamism of the labour market; for instance, both job-to-job transitions and hiring rates can be five times higher in some OECD countries than in others. However, disability gaps in those dynamics are quite similar and systematic across countries. The likelihood to drop out of the labour market is broadly speaking around twice as high for PWD as for PWOD, while at the same time PWD out of work are not even half as likely as PWOD out of work to get into, or re‑enter, the labour market. The disability disadvantage in hiring rates is particularly large in many central European countries, including Slovenia and Austria, as well as Ireland and the United Kingdom.
Further disaggregation of these findings by demographic characteristics (in this case, due to sample size, for all countries taken together) sheds further light on these labour market dynamics (Figure 2.11):
Differences between PWD and PWOD in job-to-job mobility transitions for those in employment are small for any age, gender and education category (Panel A).
Transitions to unemployment are much more frequent for PWD in all demographic groups but those among them with high level of educational attainment have the lowest disability disadvantage (Panel B). Young adults and those with low level of education generally receive notice more often than others, among both PWD and PWOD.
The labour market exit rate, i.e. transitions from employment to inactivity, is significantly larger for young and older workers, irrespective of their disability status. The likelihood of exiting the labour market is much higher for PWD in all socio-demographic groups, suggesting that many workers exit the labour market permanently due to health problems or disability (Panel C).
Hiring rates drop sharply for older workers, which is true for PWOD as well. The disability gap in hiring rates is relatively small for both young workers (age 15‑29) and older workers (age 50‑69), with only a 2.5 percentage points difference on average (Panel D). However, the gap is very large for prime‑age workers (age 30‑49): in this age group, the annual hiring rate is only 10% for PWD but over 25% for PWOD. The disability gap in hiring rates increases with the level of educational attainment and is, on average, 1.5 times larger for men than for women.
Figure 2.10. Job entry and job exit rates are much less favourable for people with disability
Note: For country disability definitions, see Figure 2.1. The purple bars represent the unweighted average of the 25 countries shown.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC).
Figure 2.11. Job hiring and unemployment transition rates vary considerably by age and education
Note: Data show the unweighted average of the 26 countries shown in Figure 2.10. People with disabilities are defined as those who 1) declared to suffer from any chronic illness or condition and 2) with moderate to severe activity limitation due to health problems. Levels of education are defined according to the International Standard Classification of Education (ISCED), where low refers to below upper-secondary, medium to upper secondary and high to tertiary education. Labour market transitions refer to individuals moving from one employment status to another in the year before the survey. Panel A: The job-to-job mobility rate is the share of those who were employed at the time of the survey and the previous year, but changed jobs, out of all the employed. Panel B: The unemployment transition rate is the share of those who were unemployed at the time of the survey, but the previous year were employed, out of all the employed. Panel C: The labour market exit rate is the share of those who were inactive at the time of the survey, but the previous year were employed, out of all the employed. Panel D: The job hiring rates is the share of those who are employed by the time of the survey but did not have a job the previous year, over the non-employed.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2016‑19).
2.4. Benefits play a very critical role in the income of people with disability
2.4.1. People with disability receive all kinds of income‑replacement benefits
Social protection is a key element for people with disability, as it helps breaking the link between disability and poverty, particularly for those with severe disability who cannot work. On average across countries, about 80% of people with severe disability receive at least one income replacement benefit (Figure 2.12). This share is substantially lower for people with a moderate disability (around 60%), with non-invalidating health problems (42%), and without health problems (30%). While differences between countries are large, this finding suggests that overall social benefits are well targeted towards those who need them most.
The number of PWD receiving a health-related benefit is much lower than the number receiving another working-age benefit, across all OECD countries. This is because PWD and in particular people with severe disability, more often receive old-age pensions (via early retirement) and social assistance than PWOD. The large number of people with disability on old-age pensions partly reflects the age pattern of disability. Instead, the large social assistance receipt suggests that people with severe disability may not be covered by social insurance, thus giving a large safety-net role to social assistance.
Figure 2.12. Sickness and disability benefits are the main but not the only benefits received
Note: Data refer to 2017 for Canada. For country definitions of people with disability, see Figure 2.1. The purple markers (AVE) represent the unweighted average of the countries shown in the top panels.
Source: The Canadian Survey on Disability, (CSD, 2017) provided by Employment and Social Development Canada and OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2016‑19).
2.4.2. Type‑of-benefit distributions vary by country but also with age
Figure 2.13 shows the distribution of social assistance and health-related benefits in more detail, country by country. On average across OECD countries, health-related benefits represent almost half of the total of benefits received by PWD, but with large variation across countries, ranging from 19% of all benefits in France to 89% in Norway. Social assistance represents around 15% of the benefits received by PWD, a slightly higher share than for PWOD. Again, there is large variation across countries: in Switzerland, social assistance amounts to 40% of the benefits PWD receive (60% for PWOD), while the share is below 5% in countries like Germany and Italy. There is also substantial variation in the benefit composition across age groups. Social assistance represents a higher share of benefits received by young people compared to older age groups, regardless of disability status. This is in line with the exclusion of younger age groups from social insurance, due to limited employment histories and contribution payments. In some countries, like Switzerland and to a lesser degree the Netherlands, social assistance is critically important as source of income for PWD, unlike the average OECD country, where disability insurance is still the principal source of social support. This reflects an effort of mainstreaming of social protection of PWD, effectively reducing any compositional differences between PWD and PWOD in the type of benefits received.
2.4.3. Most people with disability who do not work receive social benefits
The coverage of social protection programmes is best evaluated when observing the benefit receipt of people with disability who are not working, particularly for those who are most vulnerable. Figure 2.14 shows two alternative estimates of benefit coverage:
People not working and receiving any type of income support (broad coverage).
People not working and receiving sickness or disability benefits (narrow coverage).
Most PWD not working are supported through at least one type of income‑replacement benefit: on average across the OECD, 82% of PWD receive any benefit (Panel A). The share for those with severe disability is even higher, 87%, and in most countries, the difference between those with moderate and severe disability is about 5‑10 percentage points; only in Italy and Iceland, this difference is much larger. Among those without a disability who are not working, about one in two receive income support.
Country differences appear much larger when looking at more narrowly defined health-related benefits only, i.e. sickness and disability benefits. Across all OECD countries, one in two non-working persons with severe disability and one in four with moderate disability receive a health-related benefit (Panel B). Country differences range from only 20% for people with severe disability in France to almost 70% for persons with moderate disability in Norway – again reflecting how the role of different working-age benefits varies across countries. The share of non-working PWOD receiving such benefits is generally lower than 10% but amounts to 20% in Ireland and Luxembourg, and reaches 35% in Norway.
Further disaggregated data for the six study countries suggest that benefit coverage is very high for older workers and still rather high for prime‑age workers (Panel C). On the contrary, most of these countries are more careful in granting benefits to young people with disability, who typically face benefit coverage rates of around 50%. Narrow coverage by disability benefits for those with severe disability varies less with age.
Figure 2.13. Distributions across types of benefits depend on benefit system design and operation
Note: The purple bars (AVE) represent the unweighted averages of the countries shown in the top panels. Data for Canada refer to 2017 and are not available for PWOD. For country definitions of PWD, see Figure 2.1. Due to an overlap in benefits – person simultaneously receiving disability and unemployment benefits – a benefit hierarchy was generated. If a person receives any benefit related to health (disability or illness), they are classified as a beneficiary of health benefits. If a person receives unemployment benefits and some other income replacement benefit other than health, they are classified as a beneficiary of unemployment benefits. If a person receives a social assistance benefit while receiving old age benefits, they are classified as a beneficiary of social assistance. In the case of Canada, for methodological reasons the benefit hierarchy of the last two categories is swapped: social assistance comes last and old-age benefits second to last. Data not available in 2019 (Belgium, Ireland, Iceland, Italy, United Kingdom) and 2018 (Estonia). PWD ages 30‑49 refer to ages 15‑49 for Canada.
Source: Data provided by Employment and Social Development Canada based on the Canadian Survey on Disability (CSD, 2017) and OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2016‑19).
Figure 2.14. Most people with severe disability who are unable to work receive benefits
Note: Any income replacement benefits include disability, sickness, unemployment, old age and social assistance benefits. For Canada, any income replacement benefits includes health-related or unemployment or old age or social assistance and refer to 2017. The purple markers (AVE) represent the unweighted average of 26 the countries shown. For country definitions of people with disability, see Figure 2.1. Data are not available in 2019 (Belgium, Ireland, Iceland, Italy, United Kingdom) and 2018 (Estonia).
Source: Data provided by Employment and Social Development Canada based on the Canadian Survey on Disability, 2017 and European Union Statistics on Income and Living Conditions (EU-SILC, 2016‑19).
2.5. Inequality and poverty remain high for people with disability
Ultimately, the objective of effective employment and social policies is to ensure equality across population groups and a decent standard of living for those who are unable to work or restricted in their capacity to earn a living. Household-based, household-size adjusted measures of income levels, income distributions and poverty risks best capture the success of countries’ policies in achieving this objective.
2.5.1. Income levels differ between households with and without people with disability
Figure 2.15. shows that on average across all OECD countries households including PWD live with 84% of the disposable income of households not including PWD (Panel A). This difference reflects on the one hand the higher costs often linked to disability, such as medical costs, and on the other hand, the lower income they receive. Benefits received seem insufficient to compensate for the lower rate of employment. Data also show that differences between households with and without PWD have increased slightly in the past eight years. Differences between OECD countries, however, are large. In a number of European OECD countries, including Austria and Switzerland, the difference in incomes of households with and without PWD is less than 10%. On the other end, the difference is larger than 25% in Ireland, Lithuania, Mexico and Korea and as large as 35% in the United States. The largest increase in the income difference was observed in Mexico and Sweden, while Chile has seen the largest drop in that difference.
2.5.2. Type‑of-benefit distributions vary by country but also with age
Figure 2.15 (Panel B) shows how income sources – benefit income, work income, and other income – vary by disability status but also by socio‑economic characteristics (for country-specific details see Annex Figure 2.A.4). Benefits make up about 50% of the income of households with people with severe disability, and thus more than income from work. The share is even higher for those among them with low education or over age 50. Overall, even for those with moderate disability benefits represent a twice‑larger share of income than in household without PWD. Gender differences in income source distributions are negligible but differences by age are large: the older, the more important benefit income becomes – and for those aged 50‑69, even among those with health problems but no disability, benefit income is central.
2.5.3. Disability poverty gaps are large in a majority of OECD countries
Putting all pieces of information together, disability remains a major poverty driver in most OECD countries. Despite comprehensive benefit systems and high coverage rates, poor employment inclusion implies that on average across a large set of OECD countries, one in four PWD live in a household with income lower than 60% of the median (Figure 2.16, Panel A). This share has increased slightly in the past decade. Differences across countries in the share living in income‑poor households are large, ranging from 10‑15% of PWD in Iceland, the Slovak Republic and Denmark, to 33‑50% in Ireland, Korea and the United States.
Figure 2.16 (Panel B) compares disability poverty gaps across countries and over time. The poverty gap is just below 10 percentage points on average and has increased slightly in the past decade. This increase is the result of a deterioration in the relative income position of PWD in two‑thirds of the countries. Some of the largest increases in the disability poverty gap over time took place in countries with large initial gaps already (Lithuania, Ireland), while other European countries (Czech Republic, Hungary, Sweden) have seen fast increases from a low level. The United States continues to be the country with the largest disability poverty gap (27%), followed by Korea (22%). Several European OECD countries have disability poverty gaps of less than 5% in 2016‑19. Country-specific trends shown in Annex Figure 2.A.3 show that deteriorations in the disability poverty gap over time are in most cases the result of a worsening in poverty risks for PWD together with no change in the poverty risk of PWOD.
Figure 2.15. Income levels and income sources differ for people with and without disability
SD: People with severe disability, MD: People with moderate disability, HP: People with a long-standing health problem but without disability, ND: People with no health problems.
Note: Data cover persons aged 15‑69. Panel A shows the average equivalised disposable household income ratio between households with and without people with disability. Household income was equivalised dividing by the square root of the size of the household. The purple bar represents the unweighted average of the countries shown for both time periods (excludes Canada and Germany). Panel B: Data are an unweighted average of 26 countries: Austria, Belgium, the Czech Republic, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Lithuania, Luxembourg, the Netherlands, Norway, Poland, Portugal, the Slovak Republic, Slovenia, Spain, Sweden, Switzerland and the United Kingdom. For country definitions of people with disability, see Figure 2.1.
For the distribution of income sources by country, see Annex Figure 2.A.4.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2008‑19) for European countries. Data not available in 2019 (Belgium, Iceland, Ireland, Italy, United Kingdom) and 2018 (Estonia). The Household, Income and Labour Dynamics in Australia Survey (HILDA, 2008‑17); the Canadian Income Survey (CIS, 2016‑19) provided by Employment and Social Development Canada; Chile’s: Encuesta de Caracterizacion Socioeconomica Nacional (CASEN, 2016‑17); Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2016); the Korean Labour & Income Panel Study (KLIPS, 2008‑18) and the United States Current Population Survey (CPS, 2008‑18).
Figure 2.16. Despite high benefit coverage, poverty rates and gaps remain unacceptably high
PWD: People with disability, PWOD: People without disability.
Note: The data in Panel A show relative income poverty, i.e. the share of people living in a household with an income below 60% of the median income. Household income is equivalised for household composition by dividing by the square root of the size of the household. The data in Panel B show the percentage point difference between households with and without people with disability in the risk to live in a low-income household. The data for 2008‑11 refers to: 2013‑16 (Canada); 2012‑15 (Germany) and 2010 (Mexico). Data for 2016‑19 refers to 2016‑17 (Australia); 2017‑19 (Canada); 2017 (Chile); 2016‑18 (Belgium, Iceland, Ireland, Italy, United Kingdom, United States); 2016‑17, 2019 (Estonia); 2018‑19 (Korea); 2016 (Mexico). The purple bars represent the unweighted average of the countries shown in each panel.
Source: OECD calculations based on the European Union Statistics on Income and Living Conditions (EU-SILC, 2008‑19) for European countries. The Household, Income and Labour Dynamics in Australia Survey (HILDA, 2008‑17); the Canadian Income Survey (CIS, 2013‑19) provided by Employment and Social Development Canada; Chile’s: Encuesta de Caracterizacion Socioeconomica Nacional (CASEN, 2016‑17); Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2016); the Korean Labour & Income Panel Study (KLIPS, 2008‑18) and the United States Current Population Survey (CPS, 2008‑18).
Annex 2.A. Additional figures
Annex Figure 2.A.1. Employment rate by country and disability status, 2005‑19 (percentage)
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2005‑19) for European countries. The Household, Income and Labour Dynamics in Australia Survey (HILDA, 2005‑17), Canadian Income Survey (CIS, 2013‑19) provided by Employment and Social Development Canada, Chile’s Encuesta de Caracterizacion Socioeconomica Nacional (CASEN, 2006‑17), Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2010‑16), the Korean Labour & Income Panel Study (KLIPS, 2008‑18) and the American Community Survey (ACS, 2008‑18).
Annex Figure 2.A.2. Unemployment rate by country and disability status 2005‑19 (percentage)
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2005‑19) for European countries. The Household, Income and Labour Dynamics in Australia Survey (HILDA, 2005‑17), the Canadian Income Survey (CIS, 2013‑19) provided by Employment and Social Development Canada, Chile’s Encuesta de Caracterizacion Socioeconomica Nacional (CASEN, 2006‑17), Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (ENIGH, 2010‑16), the Korean Labour & Income Panel Study (KLIPS, 2008‑18) and the American Community Survey (ACS, 2008‑18).
Annex Figure 2.A.3. Poverty rate by country and disability status, 2005‑19 (percentage)
Note: Poverty line is 60% of the median equivalised household income. For Canada, based on after-tax economic family income below 60% of the median after-tax family income adjusted by family size.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (2005‑19) for European countries. The Household, Income & Labour Dynamics in Australia Survey (2005‑17), Canadian Income Survey (2013‑19) provided by Employment and Social Development Canada, Chile’s Encuesta de Caracterizacion Socioeconomica Nacional (2006‑17), Mexico’s Encuesta Nacional de Ingresos y Gastos de los Hogares (2010‑16), the Korean Labour & Income Panel Study (2008‑18) and the US Current Population Survey (2007‑18).
Annex Figure 2.A.4. Distribution of income sources across various characteristics, 2016‑19
Note: “Other” refers mostly to capital and personal transfers. AVE (pink bars) are the unweighted averages of the countries shown. Data are the annual average over the period 2016‑19.
Source: OECD calculations based on European Union Statistics on Income and Living Conditions (EU-SILC, 2016‑19), data not available in 2019 for Belgium, Ireland, Iceland, Italy and the United Kingdom and in 2018 for Estonia.