Disability in the workplace — what the 2026 data actually shows, and what it can’t
Almost everywhere it is measured, the finding is the same: people with disabilities are employed at far lower rates than people without them, paid less when they do work, and the gap has barely moved in a generation. This dossier assembles the headline numbers from the bodies that actually compile them — the WHO, the ILO, Eurostat, the US Bureau of Labor Statistics, the OECD, and the Job Accommodation Network. Around 1.3 billion people — about 16% of the world — experience significant disability. In the United States the employment-population ratio for disabled people sits in the low-to-mid 20% range against roughly 65% for everyone else; in the EU the disability employment gap runs at about 24 percentage points. And yet roughly half of all workplace accommodations cost nothing at all. No source publishes a single tidy “2026 global disability employment rate” — and any that claims to should be treated with suspicion.
What the published data shows
- 011.3B
The population the labour market is failing
The WHO’s 2022 Global Report on Health Equity for Persons with Disabilities estimates that about 1.3 billion people — roughly 16% of the world’s population — experience significant disability. That is the denominator against which every employment figure below should be read.
- 02approx. 40 pp
The US employment gap is about forty percentage points
US BLS Current Population Survey data puts the employment-population ratio for people with a disability in the low-to-mid 20% range, against roughly 65% for people without a disability — a gap of around 40 percentage points that has been remarkably stable over time.
- 03approx. 24 pp
The EU disability employment gap runs around twenty-four points
Eurostat measures the gap directly: roughly 51% of disabled people aged 20–64 are in work against about 75% of non-disabled people, a gap of approximately 24 percentage points, with very wide variation between member states. Closing it is an explicit target of the EU Disability Rights Strategy 2021–2030.
- 04approx. 50%
Around half of accommodations cost the employer nothing
The US Job Accommodation Network’s long-running employer survey consistently finds that roughly half of workplace accommodations carry no cost at all, and that most of the rest are one-time expenses of a few hundred dollars. The barrier to hiring disabled workers is rarely money — it is awareness, attitude, and inaccessible recruitment systems.
- 055% / 6%
Europe leans on statutory quotas; the common-law world on rights
Germany obliges employers with 20-plus staff to fill a 5% quota or pay a levy; France sets 6%. The US, UK and others instead use anti-discrimination law backed by a duty of reasonable accommodation. The evidence on which approach works better is genuinely mixed.
- 06record high
Remote work produced the one genuine bright spot
Inside an otherwise bleak picture, the US disability employment ratio rose to record highs in the post-pandemic years — a shift most analysts attribute to the normalisation of remote and hybrid work, which removed the commute and the inaccessible office from the equation for many disabled workers.
- 07widest
The gap is widest for women with disabilities
The ILO reports that the employment gap is consistently largest for women with disabilities, who face the compounding disadvantage of gender and disability together — an intersectional pattern that single-axis statistics tend to obscure.
SourceWHO Global Report on Health Equity for Persons with Disabilities (2022); ILO disability-and-work statistics; Eurostat disability-employment-gap series; US Bureau of Labor Statistics Persons with a Disability: Labor Force Characteristics; OECD Sickness, Disability and Work; US Job Accommodation Network employer-cost surveys. Reference years 2022–2024; figures are the most recent published as this dossier went to press.
01 · How these figures were assembled
This is a synthesis of published official statistics, not an original survey — and that distinction matters. Disability statistics are unusually fragile: definitions vary between surveys, self-identification is sensitive and tends to undercount, and many low- and middle-income countries lack regular labour-force data disaggregated by disability at all. We have drawn every figure here from the authoritative compiler of that figure, and where a number is an estimate or a range, we say so. We have not manufactured fresher or more precise numbers to fill gaps in the record.
Four bodies do most of the reliable counting. The WHO supplies the global prevalence baseline. The US BLS runs a monthly disability question in its Current Population Survey, giving the cleanest single-country time series in the world. Eurostat publishes the EU disability employment gap from its EU-SILC and labour-force instruments. The OECD contributes the comparative work on benefit design. The ILO frames the global and gender picture, and the Job Accommodation Network (JAN) supplies the employer-cost evidence.
02 · The size of the gap
The cleanest national picture comes from the United States. In recent years the BLS employment-population ratio for people with a disability has sat in the low-to-mid 20% range, against roughly 65% for people without a disability — and disabled people who are in the labour force also face an unemployment rate consistently about twice that of non-disabled workers. The European picture, measured differently, tells the same story at a higher level: roughly half of disabled Europeans of working age are employed, against three-quarters of everyone else.
| Measure | With disability | Without disability | Source |
|---|---|---|---|
| US employment-population ratio | approx. 22–25% | approx. 65% | US BLS, recent years |
| EU employment rate (age 20–64) | approx. 51% | approx. 75% | Eurostat, 2022 |
| Global population with significant disability | approx. 1.3 billion (16%) | WHO, 2022 | |
Why “2026 data” is mostly 2022–2024 data
Labour-force surveys are large, expensive instruments, and the disability-disaggregated results are typically published a year or more after collection. The most current authoritative figures available in 2026 are therefore generally from 2022–2024 reference years. We have not invented fresher numbers to fill the gap. On this topic, false precision is its own form of misinformation.
The European data also exposes a measurement subtlety worth stating plainly: because the disabled population skews older than the general working-age population, and because the survey question relies on self-reported limitation, cross-country comparisons partly reflect who answers “yes” rather than real differences in labour-market exclusion. Eurostat is candid about this, which is why the gap is best read within a country over time rather than as a league table.
03 · The pay gap and the benefits trap
Lower employment is only half the story. Where disabled people are employed, they tend to earn less — the disability pay gap — driven by concentration in lower-paid occupations, higher rates of part-time and insecure work, and outright discrimination. National statistics offices in the UK and elsewhere have measured median pay gaps of the order of 12% to 17% in recent years, though the figure is sensitive to how disability and working hours are defined.
Compounding both gaps is the benefits trap the OECD has documented across its member states: disability and out-of-work benefit systems that withdraw support sharply as earnings rise, so that taking a job — especially a part-time or uncertain one — can leave a disabled person barely better off, or worse off once the costs of working are counted. The OECD’s long-running work argues that the most effective systems pair anti-discrimination law with active labour-market support and benefit designs that taper rather than cliff-edge.
The accommodation-cost myth
The single most persistent employer misconception is that disabled workers are expensive to accommodate. The Job Accommodation Network’s employer survey has found, year after year, that roughly half of accommodations cost nothing, and that most of the remainder are one-time expenses of a few hundred dollars — a screen reader, flexible hours, a modified desk, captioning. The constraint is almost never the budget.
04 · Quotas versus the rights approach
Two broad regulatory traditions try to close the gap. Continental Europe and parts of Asia use statutory employment quotas: a legal obligation on employers above a certain size to ensure a minimum share of their workforce is disabled, usually backed by a levy paid by those who fall short. The common-law world instead relies on anti-discrimination law and a duty of reasonable accommodation.
| Jurisdiction | Mechanism | Threshold / rate |
|---|---|---|
| Germany | Quota + levy (Ausgleichsabgabe) | 5% for employers with 20+ staff |
| France | Quota + AGEFIPH fund | 6% for employers with 20+ staff |
| Italy | Graduated quota (Law 68/1999) | Rises with headcount; “targeted employment” |
| Japan | Statutory employment rate + levy | Raised in steps in recent years |
| United States | Anti-discrimination + accommodation (ADA Title I) | No quota; enforced by the EEOC |
| United Kingdom | Anti-discrimination + reasonable adjustments | No quota; Equality Act 2010 |
Quotas are widely credited with putting a floor under disability employment, but the evidence on their effectiveness is mixed. Many employers treat the levy as a cost of doing business rather than a prompt to hire, and quotas can channel disabled workers into sheltered or segregated settings that Article 27 of the UN CRPD (work and employment) discourages in favour of the open labour market. The rights approach, meanwhile, depends entirely on enforcement: under the Americans with Disabilities Act Title I, a US employer must provide adjustments that let a qualified disabled person do the job unless that would impose undue hardship — a duty mirrored by the UK Equality Act and the EU Employment Equality Directive. The comparative detail by country sits in our national regulations index.
05 · The digital gateway — hiring tech as a barrier
Increasingly, the first barrier a disabled candidate meets is not the building but the application form. Online job portals, applicant tracking systems, and AI-assisted screening tools have become the gatekeepers to the labour market — and many are not operable by keyboard-only or screen-reader users, or use video-interview and game-based assessments that disadvantage disabled applicants. The US EEOC has issued guidance warning that AI hiring tools can violate the ADA, and the EU AI Act classifies employment-related AI as high-risk. An inaccessible apply-flow is a discrimination problem before a candidate is ever judged on merit.
The fix is the same standard as everything else
Accessible recruitment software is governed by the same WCAG 2.2 criteria as any other web service, and public-sector HR procurement across the EU is bound by EN 301 549 and, since 2025, the European Accessibility Act. Employers serious about disability inclusion can start by checking their own careers site and application flow with a free accessibility scan.
06 · What the data does and does not show
Three conclusions hold across every credible dataset. First, the disability employment gap is large, durable, and global — measured in tens of percentage points wherever it is measured at all. Second, the gap is not explained by disabled people’s lower desire or capacity to work; survey after survey finds disabled people who want jobs and cannot get them, blocked by removable barriers. Third, the policy tools that work are the unglamorous ones used in combination: enforceable anti-discrimination law, accessible recruitment, well-designed benefit tapers, and a culture of accommodation — not any single silver bullet.
What the data does not yet tell us well is the granular, current, country-by-country picture for most of the world. The richest figures come from a handful of wealthy economies with strong statistical agencies; for much of the global south, the honest answer to “what is the disability employment rate?” is that nobody is reliably counting. Improving that data baseline is itself a disability-rights priority — because, as with every other indicator, what is not counted is not budgeted for. For the wider context, see our reporting on the history of disability rights activism and the full 2026 record.