Skip to main content

Home Education vs State School Outcomes (UK)

Home Education vs State School Outcomes (UK)

Last updated: 2026-08-25

Semantic Summary

The honest headline: the UK does not collect the data needed to say, nationally, whether home-educated children or state-school children "have better outcomes." Home-educated children sit outside the National Pupil Database (which tracks every state-school pupil's attainment at every key stage), UCAS does not report university entry by home-education status, and exam regulators do not publish private-candidate results separately from school-sitting candidates.[5] What evidence does exist is mostly small, volunteer-sampled and skewed towards engaged, confident home-educating families — which pushes results in a positive direction.[3]

What can be said with confidence: home education in England has grown sharply (175,900 children were home-educated at some point in 2024/25, up 15% in a year),[2] the most common reasons are mental health and dissatisfaction with school rather than a positive first choice,[1][3] and the children most likely to be home-educated are disproportionately older, SEND, or previously struggling in school — which makes any like-for-like comparison with the school population unfair from the start.[1][2] The strongest research finding is that home education is a workable alternative when a specific school placement has broken down, and that home-educated children with additional needs show no worse mental-health outcomes than school-registered peers.[3]

Basic Information

  • Type: Education policy / outcomes comparison (England, with UK context)
  • Status: Evidence base thin; policy changing (Children Not in School registers)
  • Key data source: DfE Elective Home Education (EHE) statistics, mandatory collection from autumn 2024
  • Known for: A structural data gap that makes direct comparison impossible

The short answer

There is no national UK dataset comparing the outcomes of home-educated and state-school children.[1][5] Claims online that home-educated children "outperform" or "underperform" national averages are typically drawing on US research, small qualitative UK studies, or overstating what the evidence shows.[5] The Education Committee concluded in 2021 that "without large-scale, objective data, our understanding of the educational attainment and outcomes achieved by EHE children remains largely anecdotal."[4]

Background

How many children are we talking about?

  • Census day autumn 2024: 111,700 children in elective home education (EHE) in England — 1.4% of the compulsory-school-age population.[1]
  • At any point in 2023/24: 153,300 children were EHE at some stage, up from an estimated 126,100 in 2022/23 and 116,300 in 2021/22.[1]
  • At any point in 2024/25: 175,900 — up 15% year on year.[2]
  • Autumn 2025 census day: 126,000 children were being home-educated, out of roughly nine million school-eligible children.[2]

The DfE notes the collection only became mandatory in autumn 2024 (100% of local authorities reporting), so part of the increase reflects better data quality rather than pure growth.[1]

Who home educates, and why

The "why" matters enormously for outcomes, because home education is not a random sample of families:

  • Reasons (autumn 2024): 'mental health' 14%, 'philosophical or preferential' 14%, 'lifestyle choice' 9%; more than 4 in 10 recorded as unknown/other/parent did not provide a reason.[1] By autumn 2025, 16% cited mental health and 12% philosophical or preferential reasons.[2]
  • Age: numbers rise with year group — 3% in year 1 up to 19% in year 11; over a third of EHE children were in years 10 or 11.[1]
  • SEND: 16% of EHE children had SEN support and 6% an EHCP, versus 14% and 5% in the school population.[1] The Children's Commissioner estimated in 2024 that a quarter of children leaving school for home education required SEND support.[2]
  • Movement: in 2023/24, 71,500 children started EHE while 21,500 returned to school.[1] Most children who start EHE previously attended an academy (49%) or LA-maintained school (25%).[1]
  • Pressure rather than choice: Ofsted's 2019 qualitative research found most families moving from mainstream secondary school to home education did so for SEN, medical, behavioural or wellbeing reasons, often as a last resort rather than a positive first choice.[3] Michelle Zaher of Educational Freedom: "Most of the families we support didn't set out to home educate. They were forced into it because their child was struggling, unsupported, or traumatised by their experience at school."[2]

This selection effect is the crux: comparing home-educated children against all state-school children compares a group that disproportionately includes struggling, older, SEND pupils with the general population — not a fair contest.[1][2][3]

Academic attainment

Primary age. The most-cited UK study is Paula Rothermel's 2002 Durham doctoral thesis (419 home-educating families, 196 child assessments): 64% of home-educated Reception-aged children scored above 75% on PIPS Baseline Assessments versus 5.1% nationally, and 80.4% scored in the top 16% band on the National Literacy Project assessment.[3] This is real but volunteer-sampled — families who agreed to formal assessment are the engaged end of the home-ed population — so the fairer read is that home-educated children at primary age are not academically behind, and many are ahead in the samples we have.[3]

Secondary / GCSE and A-level. There is no equivalent UK dataset. Home-educated students sit GCSEs and A-levels as private candidates at registered exam centres, taking the identical exams marked to the same standard.[5] But Ofqual and JCQ do not publish private-candidate results separately — they are absorbed into the centre type where the exam was sat — and JCQ publishes no private-candidate outcome data at all.[5] For national context, in 2025 21.8% of all GCSE grades were 7+, and 67.1% were grade 4+.[5]

What we can't know. No UCAS breakdown by home-education status; the DfE's Longitudinal Education Outcomes dataset does not tag home-educated cohorts; no solid UK data on how home-educated children fare at university or in work.[3]

Social and wellbeing outcomes

  • Socialisation: the "isolated and awkward" stereotype is not supported by the research. de Carvalho and Skipper's 2019 study of UK home-educated adolescents (small — three adolescents and their mothers) found participants describing themselves as happy, confident and socially capable ("We're not just sat at home in our pyjamas!"). International studies (Medlin & Butler 2018; Burton & Slater 2019) find home-educated children report high social engagement and larger cross-age networks. Caveat: self-report and volunteer-sampled.[3]
  • Mental health: Paulauskaite et al. (2022, Frontiers in Psychology) compared 158 UK home-educated children with neurodevelopmental conditions against 1,076 school-registered children. 77.9% of families cited their child's additional needs not being met at school as the primary reason for home educating; 76.5% cited mental health deterioration at school. On outcome measures, there were no significant differences in anxiety, internalising or externalising problems between the two groups.[3] That does not mean home education fixes mental-health problems — but it is not associated with worse outcomes for this cohort either.[3]

University and adult outcomes

Not established in UK data either way.[3] The international evidence is weak and context-specific: the US Cardus Education Survey found long-term homeschoolers had lower bachelor's degree attainment than non-homeschoolers (27% vs 46%) — but the US context does not translate cleanly to the UK.[3]

What the evidence doesn't support

  • "Home-educated children are more independent" — no rigorous UK evidence.[3]
  • "Home-educated children do better at university" — no UK data at all.[3]
  • "Home education is always better for children with SEN" — it can be, but EOTAS under Section 61 of the Children and Families Act 2014 leaves the LA legally responsible for arranging provision, whereas elective home education transfers that responsibility to the parent — a funding trade-off, not a pure benefit.[3]
  • Advocacy sources claim home-educated children consistently outperform schooled peers internationally — treat with caution; the underlying studies are volunteer-sampled and often small.[7]

Oversight and policy

  • The Centre for Social Justice's 2022 report Out of Sight and Out of Mind argued England is an international outlier in having no registration or assessment of home-educated children's progress — oversight that is commonplace across Europe.[5][6]
  • The Commons Education Committee's 2021 report Strengthening Home Education (HC 84) recommended a statutory register and called on the Department to "urgently commission and publish longitudinal research on the life chances and social outcomes of EHE children in England," measuring hard outcomes (literacy, numeracy) and soft outcomes alike.[4]
  • The government's Children's Wellbeing and Schools Bill introduces Children Not in School registers to help local authorities identify children not receiving suitable education.[2] A DfE spokesperson's line — "In the vast majority of cases, children achieve and thrive best in schools" — is a policy position, not an evidence finding.[2]
  • In 2023/24, local authorities issued an estimated 7,000 section 437(1) notices (a pre-attendance-order warning where education is deemed unsuitable) — an increase of almost 80% on the previous year.[1]

Europe: comparable countries regulate, but don't measure either

Short version: the comparison across Europe is about regulation, not results. The one systematic academic survey — Blok & Karsten's 2011 study of 14 European countries (Belgium, Bulgaria, Denmark, England, Estonia, Finland, France, Germany, Ireland, Italy, the Netherlands, Norway, Portugal, Sweden) — found "substantial differences" in how home education is regulated and inspected, and almost no published data on what inspections found or how home-educated children performed.[12] The refrain "no data are available on inspection results" recurs throughout.[12]

France — the strictest regulator, still no outcome stats

France went from a declaration regime to an authorization regime under the August 2021 "separatism" law (loi confortant le respect des principes de la République). Home schooling (instruction en famille, IEF) is now allowed on four grounds only: the child's health or disability; intensive sporting or artistic activity; family itinerancy or geographic distance from any public school; or a situation specific to the child justifying an educational project.[9] The numbers fell from 72,369 children in 2021-22 (0.6% of compulsory-school-age) to 30,644 in 2024-25 (0.3%).[9] For context, the pre-law trajectory was 13,000 (2007-08) → 36,000 (2018-19) → roughly 72,000 at the post-Covid peak.[9][11] Of 40,846 applications for 2024-25, 30,644 (about three-quarters) were granted — only 17 refusals were for terrorist or sex-offender register entries, despite the law's stated aim of countering Islamist entryism.[11]

France also inspects: more than 90% of first pedagogical controls were rated "globally satisfactory."[9] Yet the Cour des comptes' June 2025 bilan explicitly flags that indicators on home-educated children's level in national exams are missing, and that children enrolled via the CNED distance-learning provider are not inspected at all.[9] Even the strictest regulator in Europe cannot produce outcome statistics.

Germany — a ban, not a comparison

Homeschooling has been illegal in Germany since 1919; compulsory schooling is enshrined in state constitutions, with exceptions only for severe illness, diplomats' children and child actors. The KMK estimates just 500-1,000 children are home-educated nationwide.[10] The European Court of Human Rights upheld the ban in Wunderlich v. Germany (January 2019), finding no violation of Article 8 and noting the family had not shown their children were properly educated and socialised.[10] Germany's answer to the outcomes question is that there is no measurable cohort.

Netherlands — exemption only

Home education is not a recognised route in the Netherlands: exemption from compulsory schooling (vrijstelling van leerplicht) is available only for specific grounds — illness, objection to the direction (philosophy or religion) of local schools, schooling abroad, or travelling professions.[13] Prevalence was around 0.01% in 2006.[12] The leading Dutch analysis (Blok 2004) is itself a review of mostly American performance data arguing home-schooled children don't fall behind — not Dutch outcome statistics.[12]

Ireland — registered, not measured

Home education is legal and must be registered with Tusla under the Education (Welfare) Act 2000.[14] The register held roughly 2,100-2,500 children during 2024 (Q1 2,455; Q2 2,536; Q3 2,143; Q4 2,325).[15] No attainment comparison with school pupils is published.

Nordic countries — tiny cohorts, school-standard testing

Denmark (around 200 children, 0.02%), Finland (347 in 2002, 0.05%), Norway (0.07%) and Sweden (100-200, 0.01%) all require parents to notify municipalities; Denmark tests home-educated children yearly against the school programme, Finland tests them with local school teachers, and Norway evaluates twice a year.[12] None publishes aggregate outcome data.[12]

The European lesson

The UK's data gap is not an anomaly — it is the European norm. What Europe has that England lacks is input regulation: registers, authorization, home visits, curriculum requirements and testing against school standards. What no European country has — including France, which tries hardest — is published evidence on how home-educated children actually perform at national exams.[9][12]

The developed countries that DO have comparative stats (US and Canada)

United States — the only country with repeated, nationally representative comparative data

The Cardus Education Survey (run in 2011, 2014, 2018 and 2023) interviews a nationally representative sample of adults aged 24-39 about their schooling and later life outcomes. The 2023 wave (2,350 US adults polled by Ipsos; 181 had been homeschooled) splits homeschoolers into short-term (1-2 years), medium-term (3-7 years) and long-term (8+ years):[16][17]

  • Degree attainment: bachelor's degrees were earned by 46% of never-homeschooled adults and 45% of short-term homeschoolers — but only 27% of medium-term and 29% of long-term homeschoolers.[17] Medium- and long-term homeschoolers were also more likely to have ended their education with a high school diploma or an associate's degree.[16]
  • Income and employment: household income above the US median: 52% never-homeschooled, 41% short-term, 33% medium-term, 35% long-term. Unemployment was 19% for never-homeschooled adults versus 33% for medium-term homeschoolers.[17]
  • But wellbeing flips the picture: long-term homeschoolers scored highest on optimism, gratitude and life satisfaction and lowest on depression and anxiety symptoms; 63% of long-term homeschoolers were married (vs 34% of medium-term) and they had the most children on average.[16][17]
  • The authors' caveat is explicit: these are associations, not causal effects — "the findings in this report cannot speak to why homeschooling for different lengths of time might lead to different outcomes."[16][17]

The other well-known US figures — homeschooled children scoring well above national norms on standardised tests (Rudner 1999, 20,000+ children; Ray 2010, 11,729) — are the ones advocates cite, but they are methodologically weak: volunteer families who used private testing services, parents often proctoring the tests, and comparisons against test norms rather than a real control group of school pupils.[18] When a trained examiner administered tests to both groups, the homeschool advantage disappeared (Delahooke 1986).[18]

Canada — the rare controlled test-score comparison

Martin-Chang et al. (2011, Canadian Journal of Behavioural Science) tested 74 children aged 5-10. Structured homeschoolers (taught from organised lesson plans) scored higher than public-school children on standardised tests; unstructured homeschoolers scored lowest of all three groups.[18] It is a small sample, but it is the closest thing in the literature to a controlled comparison — and it points to how a child is homeschooled mattering more than whether they are.

Australia and New Zealand — registers but no outcome stats

Australia's NAPLAN national tests are administered through schools to "all Australian students in Years 3, 5, 7 and 9"; published national results break down by gender, Indigeneity, remoteness, language background and parental education — but not by home-education status, and home-educated children are not in the tested cohort.[19] New Zealand registers home-educated children with the Ministry of Education and allows them to sit NCEA externally, but published NCEA statistics are not disaggregated by home-education status.

Why this doesn't settle the UK question

All of this is survey- or volunteer-based. No developed country publishes administrative census data on home-educated attainment equivalent to England's National Pupil Database. The US homeschool population is also demographically and ideologically different from the UK's (religious motivation and a school-choice market loom far larger), so even the strongest US findings do not translate cleanly. What the developed-country evidence does support is the pattern already visible in the UK:[3] home-educated children are not academically behind on average, structure and family resources matter more than the setting itself, and long-term outcomes vary enormously with why and how long a child was home-educated.[16][17][18]

Crime and offending: the UK has no stats — and the reason matters

There is no UK dataset comparing criminal outcomes for home-educated and state-educated people. The Youth Justice Board's annual Youth Justice Statistics (England and Wales, 2024-25, published January 2026) break children in the justice system down by age, sex, ethnicity, offence and outcome — but record no education-setting variable, and the publication contains no mention of home education at all.[20] The structural reason is the same one that blocks attainment comparisons: a home-educated child is not on a school roll, the justice system does not record education history, and without a register of home-educated children there is nothing to join the dots.[1][4][5] The Education Committee's 2021 call for longitudinal research covered life chances generally — no one has ever been able to measure this specific outcome.[4]

The adjacent data that does exist:

    Safeguarding status, not crime: the DfE EHE census records LA-known safeguarding involvement. In autumn 2024, 1% of EHE children were recorded as children in need, 0.5% had a child protection plan and under 0.5% were looked after — versus 3%, 0.4% and 0.7% respectively in the general child population.[1] Home-educated children therefore show lower recorded rates of child-in-need status and care entry — but the DfE caveat is decisive: status is only "as known to the local authority at census day", so the figure reflects what LAs can see, not what is happening.[1] It is a visibility measure, not a crime comparison. Criminal exploitation: approximately 14,500 children in England and Wales were identified as at risk of or involved in child criminal exploitation in 2023-24.[21] No breakdown by education setting is published, and practitioners' guidance treats children missing education (CME) as a risk factor — a different legal category from elective home education, which the DfE defines separately.[1][21] The "invisible children" argument (US): the most-cited evidence on home education and harm concerns abuse of home-educated children, not crime by them. The US Coalition for Responsible Home Education's "Homeschooling's Invisible Children" project has analysed 520+ cases involving 240 deaths of homeschooled children, finding 62% involved imprisonment or food deprivation, 38% involved known withdrawal from school to homeschool, and that in 61% of cases harm only came to light when the child was dead or dying.[22] The counterpoint (from homeschool advocacy research) is that no conclusive comparative evidence shows school children are maltreated at lower rates than home-educated children.[23] Either way, this is about detection and oversight of abuse — it is not a statistic on offending by home-educated people, and no comparable UK case-level dataset exists.

    Bottom line on crime: as with attainment, nobody measures it. UK crime statistics do not record whether an offender was home-educated, there is no register to link criminal records to education status, and no research has ever attempted the comparison.[4][20]

    Bottom line

    • If the question is "which group scores better on national measures?" — unanswerable. State-school attainment is measured (National Pupil Database, Progress 8, KS2/KS4 statistics); home-educated attainment is largely unmeasured, and where it is measured (private-candidate GCSEs) results are not published separately.[5]
    • If the question is "does the evidence show home education harms children?" — no. The published UK evidence does not support the assumption that home-educated children fall behind academically, and it shows no worse mental-health outcomes for children with additional needs.[3]
    • If the question is "does the evidence show home education is better?" — no, not as a blanket claim. The consistent, well-evidenced finding is narrower: home education can be better than the particular school placement that stopped working (SEN, bullying, mental health).[3]
    • "Is there a European country doing measurement better?" — no. France, Ireland, Denmark and Finland all regulate home education far more tightly than England (authorization, registers, inspection, curriculum testing), but none publishes outcome comparisons; even France's audit body admits national-exam indicators for home-educated children don't exist.[9][12]
    • "Has any developed country actually got the stats?" — only the US, via surveys (Cardus 2023: medium- and long-term homeschoolers show degree and income gaps but better wellbeing), plus one small Canadian controlled study (structured homeschool > school > unstructured). No country publishes administrative attainment data for home-educated children.[16][17][18]
    "Do home-educated people commit more or less crime?" — unanswerable with UK data. Youth justice statistics record no education-setting variable and never mention home education; no register exists to link criminal records to home-education status. The only adjacent measure (LA-known safeguarding status) shows home-educated children with lower recorded child-in-need and care rates, but that reflects LA visibility, not outcomes.[1][20] The fairest summary: outcomes depend on why the child left school, the family's resources and structure, and how deliberately the child's social and academic needs are met — and the UK currently has no data infrastructure to measure any of this at national level.[1][3][4]

    Tags

    • home-education
    • education
    • UK-politics
    • outcomes
    • SEND
    • mental-health
    • data-gap
    • policy
    • europe

    Sources

    [1] https://explore-education-statistics.service.gov.uk/find-statistics/elective-home-education/2024-25-autumn-term — DfE: Elective home education statistics (autumn 2024/25) [2] https://news.sky.com/story/number-of-children-being-taught-at-home-increases-by-15-in-a-year-report-finds-13494608 — Sky News: Home-educated numbers up 15% (Jan 2026) [3] https://cognito.org/blog/benefits-of-homeschooling-research — Cognito: Benefits of homeschooling — what the research shows [4] https://publications.parliament.uk/pa/cm5802/cmselect/cmeduc/84/8403.htm — Education Committee: Strengthening Home Education summary (HC 84, 2021) [5] https://myedspace.co.uk/blog/post/home-educated-students-exam-results — MyEdSpace: How do home educated students' exam results compare? [6] https://www.centreforsocialjustice.org.uk/wp-content/uploads/2022/11/CSJ-Home-education-report.pdf — CSJ: Out of Sight and Out of Mind (2022) [7] https://www.home-education.org.uk/articles/wc/wc-he-outcomes.pdf — Home-education.org.uk: Outcomes for EHE children (advocacy) [9] https://www.vie-publique.fr/en-bref/299312-instruction-en-famille-quel-bilan-quatre-ans-apres-la-reforme — Vie publique: Instruction en famille — bilan 4 ans après la réforme (Cour des comptes) [10] https://www.dw.com/en/european-court-rules-against-german-homeschooling-family/a-47021333 — DW: ECtHR rules against German homeschooling family (2019) [11] https://www.cpdh.org/articles/instruction-a-la-maison-la-cour-des-comptes-dresse-un-premier-bilan — CPDH: Cour des comptes bilan on instruction en famille (June 2025) [12] https://www.researchgate.net/profile/Henk-Blok/publication/230092738_Inspection_of_Home_Education_in_European_Countries/links/59fb3e5eaca272347a1d1e24/Inspection-of-Home-Education-in-European-Countries.pdf — Blok & Karsten 2011: Inspection of Home Education in European Countries [13] https://www.government.nl/themes/education/compulsory-school-attendance/exemptions-from-compulsory-school-attendance — Government.nl: Exemptions from compulsory school attendance (Netherlands) [14] https://datacatalog.tusla.ie/dataset/no-of-children-on-the-register-for-home-education-2024 — Tusla: No. of Children on Register for Home Education 2024 [15] https://services-eu1.arcgis.com/FKG93wmpArM17gAf/ArcGIS/rest/services/LiveMetrics/FeatureServer/1403/query?where=1=1&outFields=*&f=pjson — Tusla LiveMetrics data: Home education register 2024 (quarterly) [16] https://www.cardus.ca/research/diverse-outcomes-for-a-diverse-population — Cardus: Diverse Outcomes for a Diverse Population (Cheng & Watson, 2025) [17] https://homeschooledvoices.com/2025/12/31/recent-homeschooling-research — Homeschooled Voices: Recent Homeschooling Research (2023 CES numbers) [18] https://www.researchgate.net/profile/Odette-Gould/publication/232544669_The_Impact_of_Schooling_on_Academic_Achievement_Evidence_From_Homeschooled_and_Traditionally_Schooled_Students/links/0c9605228ea23bc4bc000000/The-Impact-of-Schooling-on-Academic-Achievement-Evidence-From-Homeschooled-and-Traditionally-Schooled-Students.pdf — Martin-Chang, Gould & Meuse 2011: Impact of Schooling on Academic Achievement (Canada) [19] https://www.education.gov.au/national-assessment-program/national-assessment-program-literacy-and-numeracy — Australian Dept of Education: NAPLAN overview [20] https://www.gov.uk/government/statistics/youth-justice-statistics-2024-to-2025/youth-justice-statistics-2024-to-2025 — Youth Justice Statistics 2024-25 (YJB/MoJ) [21] https://safeguardingchildren.co.uk/professionals/procedures-practice-guidance-and-one-minute-guides/criminal-exploitation-and-county-lines — NYSCP: Criminal Exploitation and County Lines guidance [22] https://hic.crhe.org — CRHE: Homeschooling's Invisible Children (US) [23] https://nheri.org/homeschooling-and-child-maltreatment-no-significant-news — NHERI: Homeschooling and Child Maltreatment — response

    Back to top