Disability and Education Access: Global Statistics and Data Gaps
International evidence shows lower school entry, attendance, progression, completion and foundational learning among children with disabilities. The available figures do not form one continuous global series: age coverage, disability definitions, severity thresholds, survey instruments and the populations included differ across sources.
Executive Snapshot
Values refer to different populations and should be read with their source scope.
Out-of-School Rates Across Education Stages
UNICEF pooled estimates compare children without functional difficulties, children with one or more difficulties and children with difficulties in more than one domain.
School Exclusion Rises Through Secondary Education
Percentage of children not attending primary, secondary or higher education at the corresponding official school age.
The pooled results describe children covered by the UNICEF microdata analysis, not every country or every child worldwide.
Chart analysis
- At primary-school age, the reported rate is 13% for children without functional difficulties and 19% for children with one or more difficulties.
- Children with difficulties in more than one domain record a primary-school-age rate of 24%, an 11-percentage-point difference from children without difficulties.
- At lower-secondary age, the three reported rates are 16%, 21% and 28%.
- At upper-secondary age, the rates rise to 28%, 35% and 43%.
- The upper-secondary difference between children with multiple difficulties and children without functional difficulties is 15 percentage points.
- The chart measures school attendance status. It does not measure classroom participation, available support or learning quality.
A finding such as “49% more likely to have never attended school” is a relative comparison between two groups. It does not mean that 49% of all children with disabilities have never attended school.
Global Disability and Education Evidence Register
The table separates population estimates, absolute attendance rates, relative differences, data-coverage measures and regional administrative findings.
| ID | Indicator | Value | Unit or direction | Population | Coverage | Reference year | Source |
|---|---|---|---|---|---|---|---|
| 1 | Children with disabilities | 236.35 | Million children | Ages 0–17 | Global modeled estimate | 2021 report | UNICEF |
| 2 | Disability prevalence | 10.1% | Population share | Ages 0–17 | Global modeled estimate | 2021 report | UNICEF |
| 3 | Disability prevalence | 4.3% | Population share | Ages 0–4 | Global modeled estimate | 2021 report | UNICEF |
| 4 | Disability prevalence | 12.5% | Population share | Ages 5–17 | Global modeled estimate | 2021 report | UNICEF |
| 5 | Never attended school | 49% | More likely | Children with disabilities versus peers | Comparable pooled microdata | 2021 report | UNICEF |
| 6 | Early childhood education attendance | 25% | Less likely | Children with disabilities versus peers | Comparable pooled microdata | 2021 report | UNICEF |
| 7 | Out of primary school | 47% | More likely | Children with disabilities versus peers | Comparable pooled microdata | 2021 report | UNICEF |
| 8 | Out of lower-secondary school | 33% | More likely | Children with disabilities versus peers | Comparable pooled microdata | 2021 report | UNICEF |
| 9 | Out of upper-secondary school | 27% | More likely | Children with disabilities versus peers | Comparable pooled microdata | 2021 report | UNICEF |
| 10 | Foundational reading and numeracy | 42% | Less likely to have skills | Children with disabilities versus peers | Comparable pooled microdata | 2021 report | UNICEF |
| 11 | Out of primary school | 13% | Absolute rate | No functional difficulties | Pooled survey result | 2021 report | UNICEF |
| 12 | Out of primary school | 19% | Absolute rate | One or more functional difficulties | Pooled survey result | 2021 report | UNICEF |
| 13 | Out of primary school | 24% | Absolute rate | More than one functional difficulty | Pooled survey result | 2021 report | UNICEF |
| 14 | Out of lower-secondary school | 16% | Absolute rate | No functional difficulties | Pooled survey result | 2021 report | UNICEF |
| 15 | Out of lower-secondary school | 21% | Absolute rate | One or more functional difficulties | Pooled survey result | 2021 report | UNICEF |
| 16 | Out of lower-secondary school | 28% | Absolute rate | More than one functional difficulty | Pooled survey result | 2021 report | UNICEF |
| 17 | Out of upper-secondary school | 28% | Absolute rate | No functional difficulties | Pooled survey result | 2021 report | UNICEF |
| 18 | Out of upper-secondary school | 35% | Absolute rate | One or more functional difficulties | Pooled survey result | 2021 report | UNICEF |
| 19 | Out of upper-secondary school | 43% | Absolute rate | More than one functional difficulty | Pooled survey result | 2021 report | UNICEF |
| 20 | Children and adolescents with disabilities | 291.2 | Million people | Children and adolescents | Global estimate in current brief | Not stated | World Bank |
| 21 | Never attended school | 2.5× | Relative likelihood | Children with disabilities versus peers | Global evidence cited by current brief | Not stated | World Bank |
| 22 | Share living in low- and middle-income countries | Approximately 95% | Population distribution | Children and adolescents with disabilities | Global estimate in current brief | Not stated | World Bank |
| 23 | Comparable disability microdata | 42 | Countries and areas | Children covered by available surveys | More than 60 indicators | 2021 | UNICEF |
| 24 | Countries used in global child estimate | 103 | Countries and areas | Children aged 0–17 | 84% of the global child population | 2021 report | UNICEF |
| 25 | Countries without a recent public household survey | 41% | Country share | Countries assessed for disaggregated education monitoring | No survey conducted or made public since 2015 | 2020 report | UNESCO GEM |
| 26 | World population in countries without recent public survey data | 13% | Population share | Global population | Same survey-availability assessment | 2020 report | UNESCO GEM |
| 27 | Disabled children within the out-of-school population | 15% | Population share | Out-of-school children | 14 countries using the Child Functioning Module | 2020 report | UNESCO GEM |
Table analysis
- The UNICEF global estimate identifies 236.35 million children aged 0–17, equivalent to 10.1% of the child population.
- Estimated prevalence is 4.3% at ages 0–4 and 12.5% at ages 5–17.
- The education evidence covers different outcome types: school entry, current attendance, education level, foundational skills and population coverage.
- The highest reported relative education difference in the UNICEF summary is 49% more likely to have never attended school.
- Foundational reading and numeracy show a 42% lower likelihood among children with disabilities.
- The World Bank figure of 291.2 million covers children and adolescents and cannot replace the UNICEF 0–17 estimate without a common definition, age range and reference year.
- The UNESCO survey-availability measure concerns all disaggregated education monitoring, not only disability modules.
- The estimate that disabled children form 15% of the out-of-school population comes from 14 countries and is not a worldwide share.
Regional Distribution of Children with Disabilities
UNICEF regional estimates use harmonized source data and demographic totals from World Population Prospects 2020.
Estimated Prevalence Among Children Aged 0–17
Regional percentage of children identified as having disabilities under the UNICEF estimation method.
Regional differences can reflect population structure, survey instruments, functional domains, response patterns and model adjustments.
Chart analysis
- West and Central Africa has the highest modeled regional prevalence at 14.9%.
- Europe and Central Asia has the lowest modeled regional prevalence at 5.5%.
- The range between these two modeled regional values is 9.4 percentage points.
- South Asia, Eastern and Southern Africa, Latin America and the Caribbean, and North America are clustered between 9.9% and 10.5%.
- Regional prevalence should not be interpreted as an education-access ranking.
| Region | Ages 0–4 | Ages 5–17 | Ages 0–17 | Children aged 0–17 | Count unit |
|---|---|---|---|---|---|
| East Asia and the Pacific | 3.5% | 9.5% | 7.8% | 43.121 | Million |
| Eastern and Southern Africa | 5.2% | 12.8% | 10.4% | 28.865 | Million |
| Europe and Central Asia | 2.7% | 6.5% | 5.5% | 10.814 | Million |
| Latin America and the Caribbean | 3.8% | 12.6% | 10.2% | 19.080 | Million |
| Middle East and North Africa | 4.5% | 16.9% | 13.1% | 20.940 | Million |
| North America | 4.4% | 12.0% | 9.9% | 8.016 | Million |
| South Asia | 3.7% | 13.0% | 10.5% | 64.431 | Million |
| West and Central Africa | 6.8% | 18.9% | 14.9% | 41.083 | Million |
| World | 4.3% | 12.5% | 10.1% | 236.350 | Million |
Table analysis
- South Asia has the largest estimated number of children with disabilities at 64.431 million.
- East Asia and the Pacific follows with 43.121 million, while West and Central Africa records 41.083 million.
- Prevalence among children aged 5–17 is higher than prevalence among children aged 0–4 in every region.
- The widest age-group difference appears in the Middle East and North Africa: 4.5% at ages 0–4 and 16.9% at ages 5–17.
- The regional counts sum to 236.35 million, the unrounded value behind the “nearly 240 million” headline.
- Differences in child population size explain why the region with the highest prevalence does not have the largest estimated child count.
SIRIED 2024: Enrollment, Progression and Completion
The findings cover 16 participating Latin American countries and provide a regional administrative-data case study rather than a global average.
Completion Rates by Disability Status
Reported primary and upper-secondary completion rates in participating SIRIED countries.
The calculated completion differences are 23 percentage points at primary level and 21 percentage points at upper-secondary level.
Chart analysis
- Primary completion is 76% among students with disabilities and 99% among students without disabilities.
- The primary completion difference is 23 percentage points.
- Upper-secondary completion is 61% among students with disabilities and 82% among students without disabilities.
- The upper-secondary completion difference is 21 percentage points.
- Completion measures the end of an education stage; it does not show whether accessible materials or learning support were available.
| ID | Indicator | Value | Unit | Population or setting | Interpretation |
|---|---|---|---|---|---|
| 1 | Enrolled students with disabilities | 2.3 | Million students | 16 participating countries | Administrative enrollment count |
| 2 | Enrollment in regular schools | 81% | Student share | Students with disabilities | School placement, not participation quality |
| 3 | Enrollment in special schools | 19% | Student share | Students with disabilities | Separate school placement |
| 4 | Regular schools enrolling disabled students | 47% | School share | Regular schools | Enrollment is concentrated in fewer than half of schools |
| 5 | Regular-school placement in public primary education | 82% | Student share | Disabled students in public primary schools | 18% are in special education |
| 6 | Regular-school placement in private primary education | 56% | Student share | Disabled students in private primary schools | 44% are in special education |
| 7 | Transfer from special to regular school | 4.5% | Student flow | Students changing school type | Movement toward regular education |
| 8 | Transfer from regular to special school | 0.6% | Student flow | Students changing school type | Opposite placement movement |
| 9 | Upper-secondary dropout | 37% | Student rate | Students with disabilities | Leaves before completing upper secondary |
| 10 | Primary grade repetition | 10% | Student rate | Primary students with disabilities | Repeats an academic year |
| 11 | Primary completion | 76% | Completion rate | Students with disabilities | Compared with 99% among students without disabilities |
| 12 | Upper-secondary completion | 61% | Completion rate | Students with disabilities | Compared with 82% among students without disabilities |
| 13 | Enrollment identification at age 12 | 33 | Per 1,000 students | Students identified with disabilities | Highest reported age-specific presence |
| 14 | Enrollment identification at age 17 | 24 | Per 1,000 students | Students identified with disabilities | Lower presence at the end of secondary-school age |
| 15 | Overall dropout among girls with disabilities | 16% | Student rate | Female students with disabilities | Compared with 14% among boys |
| 16 | Upper-secondary dropout among women with disabilities | Approximately 40% | Student rate | Female students with disabilities | Compared with approximately 37% among male students |
Table analysis
- 81% of enrolled students with disabilities attend regular schools, yet only 47% of regular schools enroll any students with disabilities.
- The two measures indicate that regular-school enrollment is concentrated across part of the school system.
- Regular-school placement is 82% in public primary education and 56% in private primary education.
- Transfers from special to regular schools reach 4.5%, compared with 0.6% in the other direction.
- Upper-secondary dropout reaches 37% among students with disabilities.
- The identified presence of students with disabilities falls from 33 per 1,000 at age 12 to 24 per 1,000 at age 17.
- The table includes only students visible in participating education systems. Children who never enrolled are outside its administrative student count.
Why Published Global Estimates Differ
A larger estimate is not automatically newer, more accurate or directly comparable with a smaller estimate.
| Data product | Headline value | Population | Data base | Coverage count | Publication year | Use | Main limitation |
|---|---|---|---|---|---|---|---|
| UNICEF global child estimate | 236.35 million | Children aged 0–17 | Harmonized population sources and modeling | 103 countries and areas | 2021 | Global and regional child prevalence | Not a repeated annual global series |
| UNICEF comparable microdata analysis | 42 countries and areas | Children in available household surveys | Primarily MICS microdata using child functioning questions | 42 | 2021 | Disability gaps across more than 60 indicators | Indicator-specific country coverage and sample sizes vary |
| World Bank inclusive education brief | 291.2 million | Children and adolescents | Global estimate cited in the brief | Not reported | Current brief | Scale of disability inclusion in education | Underlying reference year and exact age span are not stated on the brief |
| SIRIED 2024 | 2.3 million enrolled students | Students identified with disabilities | Education administrative records | 16 countries | 2025 | Placement, progression, repetition, dropout and completion | Does not count children outside participating education systems |
| UNESCO GEM survey-coverage analysis | 41% of countries without recent public survey data | National education data systems | Survey availability review | Global review | 2020 | Assessment of disaggregated education data availability | The measure covers multiple equity dimensions, not disability alone |
Source-year analysis
- The UNICEF and World Bank global headline estimates use different population labels.
- UNICEF provides a stated age range, modeled country coverage and a global prevalence estimate.
- The World Bank brief provides a larger population estimate but does not state the underlying reference year on the brief.
- SIRIED counts enrolled students rather than all children with disabilities in the population.
- Combining these products into a single ranked or historical series would mix population estimates, survey comparisons and administrative enrollment counts.
Measurement and Reporting Timeline
Method development has expanded the types of disability and education information that can be collected, but it has not yet produced a continuous global trend.
| Year | Development | Respondent or data source | Main use | Coverage boundary |
|---|---|---|---|---|
| 2016 | Child Functioning Module released | Mother or primary caregiver in household surveys | Population-level identification of functional difficulties | Surveyed households and included population groups |
| 2020 | UNESCO GEM inclusion and data-coverage assessment | Household survey and administrative data review | Identify missing disaggregated education data | Availability does not guarantee common definitions |
| 2021 | UNICEF global child disability estimate and pooled outcome analysis | Multiple harmonized sources and MICS microdata | Prevalence, education gaps and child well-being indicators | Global estimate and outcome analysis use different source sets |
| 2024 | Inclusive Education Module released | Household survey respondent | School background, environment, non-attendance and participation barriers | Designed for use with the Child Functioning Module |
| 2024 | Child Functioning Module–Teacher Version released | Teacher in school-based systems | Education management data and school surveys | Cannot identify children who never reach school |
| 2025 | Teacher-version comparative evaluation published | Matched caregiver and teacher reports in Kosovo and Malawi | Assess agreement across respondent types | Two study settings; disagreement varies by functional domain |
| 2025 | SIRIED 2024 regional report released | Education administrative systems | Enrollment, school type, transition, repetition, dropout and completion | 16 participating Latin American countries |
Trend analysis
- The sequence shows expansion from population identification toward school-environment and education-pathway measures.
- The 2024 teacher version can add disability information to education management systems at lower collection cost.
- School-based tools improve information on enrolled students but leave never-enrolled children outside the respondent base.
- The Inclusive Education Module adds information on environmental barriers that enrollment totals cannot provide.
- No listed release creates a retrospective global annual series that can be validly extended back to 2016.
Education Access Is a Sequence
Enrollment is one point in a longer education path. A learner may be registered in school while remaining excluded from teaching, assessment or progression.
Population Identification
The child must first be represented in a census, household survey or disability data collection. Children outside conventional households or without birth registration may be missed.
School Entry
Administrative records begin when a child enrolls. A school database cannot directly count a child who has never entered an education institution.
Regular Attendance
Registration does not show whether inaccessible transport, health needs, cost or unavailable support produce frequent absence.
Classroom Participation
Attendance records rarely show access to sign language, Braille, alternative communication, adapted materials or personal assistance.
Progression
Age-for-grade status, repetition and transition between education levels reveal losses that total enrollment can conceal.
Completion
Completion rates identify whether students reach the end of an education stage. They do not confirm the learning achieved.
Foundational Learning
Learning assessments may omit students who are out of school, exempted from testing or unable to use the assessment format provided.
Transition Beyond School
Comparable disability data become thinner at tertiary education, technical education and the transition from education to employment.
A student may attend a regular school without accessible teaching materials, communication support, adapted assessment, safe transport or meaningful participation in classroom activities.
What Each Data Source Can and Cannot See
No single collection system covers every child group and every part of the education path.
| Data source | Never enrolled | Enrolled students | Attendance | Participation | Learning | School environment | Main blind spot |
|---|---|---|---|---|---|---|---|
| Population census | Possible | Possible | Limited | Often missing | Missing | Missing | Few questions and long intervals between rounds |
| Household survey with Child Functioning Module | Usually captured | Usually captured | Possible | Limited without added module | Possible in selected surveys | Limited without IEM | Institutions and hard-to-reach households may be excluded |
| Education management information system | Missing | Strong coverage | Possible | Often limited | Limited | Possible if fields exist | Children outside registered education institutions |
| Teacher-version functioning module | Missing | Strong coverage | Possible with linked records | Partial | Not a direct learning test | Requires additional questions | Teacher and caregiver reports can differ by domain |
| Inclusive Education Module | Captures reasons | Captured | Captured | Partial | Not a direct learning test | Detailed coverage | Requires survey implementation with functioning data |
| National or international learning assessment | Missing | Selected grades or ages | Test-day presence only | Limited | Strong for included students | Possible through questionnaires | Students may be excluded by sampling or assessment rules |
| Health or disability register | Possible if linked | Possible if linked | Missing | Missing | Missing | Missing | Children without diagnosis or certification |
| SIRIED-type regional administrative system | Generally missing | Strong coverage | Possible | Partial | Limited | Regulation and facility indicators | Coverage depends on participating countries and identification rules |
| Qualitative school and household research | Possible | Possible | Detailed experience | Detailed experience | Contextual evidence | Detailed experience | Cannot produce national prevalence without representative sampling |
Table analysis
- Household surveys are the main source capable of identifying children with disabilities both inside and outside school.
- Education administrative systems provide repeated data on enrolled students but begin after school entry.
- Teacher reporting adds functional information inside schools but cannot resolve the absence of never-enrolled children.
- Learning assessments measure outcomes only for students included in the assessment population and able to use the test format.
- Health and certification registers can undercount children who lack access to formal diagnosis.
- Combining population surveys, administrative records, accessible learning assessments and school-environment measures provides broader coverage than relying on one source.
Seven Data Gaps That Alter Reported Access
Each gap affects the numerator, denominator or population represented in an education indicator.
Identification Gap
Children may be missed when data depend on medical certification, a narrow set of physical domains or an instrument not designed for children.
Enrollment Gap
School records describe registered students. They do not provide a direct count of disabled children who have never entered school.
Attendance Gap
A child can remain registered while missing lessons because of transport barriers, unavailable assistance, health needs or inaccessible facilities.
Participation Gap
Classroom presence does not measure access to communication, adapted teaching, extracurricular activities or peer interaction.
Learning Gap
Assessment results can omit students outside school, exempted students and students unable to use the assessment format.
Transition Gap
Weak linkage across school levels can hide delayed entry, repetition, transfers, dropout and movement between regular and special education.
Support Gap
Many systems count disability status without recording assistive technology, accessible transport, communication support or adapted assessment.
Post-School Gap
Comparable data become less available at technical education, tertiary education, adult learning and the transition into employment.
Interpretation Boundaries
Low Prevalence Can Reflect Under-Identification
A low reported disability rate may result from a narrow definition, limited functional domains, inaccessible diagnosis or missing children rather than lower underlying need.
Improved Identification Can Raise the Reported Rate
Adding learning, communication, psychosocial or behavioral domains can identify students who were absent from earlier administrative categories.
Relative Risk Does Not Give an Absolute Rate
A relative difference requires the comparison-group rate before the percentage can be converted into a group-specific absolute rate.
Regional Results Are Not Global Averages
SIRIED values describe participating Latin American countries and should not be assigned to countries outside that regional source.
Enrollment Does Not Measure Learning
An enrolled student may lack accessible materials, adapted assessment or communication support and may not acquire foundational skills.
Different Respondents Produce Different Evidence
Caregivers observe children across home and community settings, while teachers report behavior and functioning in school. The reports need not match in every domain.
Source and Methodology
UNICEF — Seen, Counted, Included. The global estimate uses data from 103 countries and areas covering 84% of the global population aged 0–17. The report presents 236.35 million as the unrounded estimate and describes it as nearly 240 million.
UNICEF processed child-level microdata from 42 countries and areas to create harmonized functioning, education and background variables. Coverage differs by indicator and available sample size.
World Bank — Inclusive Education. The current brief reports 291.2 million children and adolescents with disabilities, approximately 95% living in low- and middle-income countries, and a 2.5-times likelihood of never attending school.
UNESCO — SIRIED 2024 regional findings. The data cover 2.3 million enrolled students with disabilities across 16 participating Latin American countries. The source was released in 2025.
Washington Group and UNICEF — Inclusive Education Module. Released in June 2024, the module covers child background, school type, school environment, reasons for non-attendance and attendance determinants.
Washington Group and UNICEF — Child Functioning Module–Teacher Version. The instrument contains 20 questions for students aged 5–17 and is designed for education management systems and school surveys.
Child Functioning Module–Teacher Version: A Comparative Evaluation. The 2025 study compares caregiver and teacher reports in Kosovo and Malawi.
UNESCO Global Education Monitoring Report 2020. The report documents gaps in public household surveys, administrative disability records and learning data for children outside school.
Population estimates are shown in millions. Prevalence, attendance, completion and dropout are shown as percentages. Relative likelihood measures are kept separate from absolute rates.
Percentage-point differences were calculated by subtracting the lower reported rate from the higher reported rate. No missing country or regional value was estimated.
UNICEF regional prevalence values retain one decimal place. Regional child counts retain three decimal places in millions. Relative differences retain the whole-number format used by the source.
“Not reported” indicates that the cited source did not state the value needed for direct comparison. Missing source years were not replaced with the publication year.
Disability identification in the UNICEF global estimate was harmonized across multiple instruments. The Child Functioning Module covers more functional domains than short adult-oriented question sets and can therefore identify a broader population of children experiencing moderate-to-severe functional difficulties. UNICEF’s global prevalence estimate, pooled education comparisons, World Bank headline estimate and SIRIED administrative results represent separate statistical products.
No country ranking or causal comparison is produced from these sources. Differences in age range, disability definition, severity threshold, respondent, school inclusion, source year and data type prevent the figures from being combined into one global education-access score or a continuous historical series. School-based records exclude children who never enroll, while household surveys may exclude institutional populations and households that cannot be reached.