Industries / Government, education and community / Education / Homeschooling
Records and diplomas
Turning learning at home into papers that colleges, the military, sports bodies and aid offices accept, and counting how many children learn at home.
6
Milestones achieved, from 1998 to 2025.
1
To watch for: announced, not yet confirmed outside.
2
Open deliverables. Nobody has shipped them.
The road
AchievedWatch forOpen
Achieved
1998
Federal aid without a diploma
Congress made students who completed secondary school "in a home school setting" under state law eligible for US federal student aid. Today's rule accepts either a state home-school completion credential or, where a state issues none, home schooling that counts as an exemption from compulsory attendance.
Credit
US Congress, Higher Education Amendments of 1998; US Department of Education
1999
A federal count begins
The National Household Education Survey, run by the National Center for Education Statistics (NCES), estimated 850,000 homeschooled students aged 5 to 17 in 1999, or 1.7% of students. The Hechinger Report says the Education Department began counting homeschoolers that year.
Credit
NCES
2011
Equal footing for recruits
A 2006 law told the Defense Department to exempt home-school graduates from the diploma-or-GED rule for enlisting, and the defense law for fiscal 2012, signed on 31 December 2011, requires each service to treat anyone who "completes a program of secondary education in compliance with the education laws of the State" the same as a high school graduate.
LessonTying recognition to state law rather than to a diploma let one rule cover every state's homeschool system.
Credit
US Congress
2020
Homeschooling doubles in a pandemic
The Census Bureau's Household Pulse Survey found that 5.4% of US households with school-age children reported homeschooling in spring 2020, and 11.1% by fall 2020.
LessonThis survey counts households, not students, so it cannot be compared directly with the Education Department's figures.
Credit
US Census Bureau (Casey Eggleston and Jason Fields)
2023
England starts counting
England's Department for Education published its first elective home education figures in May 2023, from councils reporting voluntarily, with an estimated 80,900 children in autumn 2022. Reporting became mandatory in autumn 2024, and the January 2026 release counted 126,000 children, 1.5% of those aged 5 to 16.
LessonThe department says part of the rise reflects better data, not only more home education.
explore-education-statistics.service.gov.ukexplore-education-statistics.service.gov.uk
Credit
Department for Education, with local authorities
2025
A federal count stalls
NCES released a preliminary figure in September 2024: 3.4% of US students were homeschooled in 2022–23. The contract for the survey's final data was cancelled in February 2025 and the staff who oversaw it were dismissed in March, and the full data has not been released.
LessonThe US national count depended on one survey contract.
Credit
reported by The Hechinger Report
Watch for
Counts from England's registers
The same 2026 Act lets the Secretary of State direct councils to hand over information from their new registers, "whether that is information relating to an individual child or aggregated information", which would give a national count built from registers rather than estimates.
Achieved whenThe registers are in force and the department publishes a national count drawn from them.
Claimed by
UK Parliament and the Department for Education, April 2026
Open
01
Parent transcripts in standard form
An open-source converter that reads parent-made transcripts (PDFs, scans, spreadsheets) and writes a standard machine-readable record, such as the PESC high school transcript or the 1EdTech Comprehensive Learner Record, with a public test set of at least 300 synthetic transcripts in at least 30 layouts, each with an answer key. The NCAA's required fields serve as the checklist: ninth-grade start date, course titles, grades, credits, grading scale and the administrator's signature.
The briefCredit · unclaimed02
How districts credit home courses
A public dataset of school districts' written rules on credit for high school work done at home, for when a teenager enrolls in a public high school, coded from published district policy manuals, many hosted on BoardDocs, NEOLA or Simbli. Each district is coded on eight fields: whether a written rule exists; who decides; what evidence it asks for, such as a portfolio or exams; any exam pass mark; any cap on credits; whether grades are kept or shown as pass only; the minimum credits a student must earn in the district for its diploma; and how a family can appeal. Every code carries the policy number, revision date, link and quoted clause. It describes published rules only and gives no advice to any family.
The briefCredit · unclaimed
Open 01 · the brief
Parent transcripts in standard form
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
An open-source converter that reads parent-made transcripts (PDFs, scans, spreadsheets) and writes a standard machine-readable record, such as the PESC high school transcript or the 1EdTech Comprehensive Learner Record, with a public test set of at least 300 synthetic transcripts in at least 30 layouts, each with an answer key. The NCAA's required fields serve as the checklist: ninth-grade start date, course titles, grades, credits, grading scale and the administrator's signature.
Why it is still open
Commercial tools read school transcripts; Softdocs says its product "can read any transcript, regardless of format", but neither it nor the other tools found mention homeschool transcripts or publish a test set. The NCAA still asks home-school students for a hand-filled Core-Course Worksheet for each course.
Built with what exists
AI models that read page images, which now handle one-off layouts without a template for each school; Tesseract OCR; Python; the Faker library for made-up names; Word or LaTeX templates to render varied layouts. No real student's records.
First step
Write the answer-key schema from the NCAA toolkit's field list and render 50 synthetic transcripts in 10 layouts.
Done looks like
At least 95% field-level accuracy on layouts held out of development, 100% of outputs passing schema validation, and the test set and scores published.
Open 02 · the brief
How districts credit home courses
The first to publish is credited on this line, and the deliverable moves from Open to Achieved.
What to build
A public dataset of school districts' written rules on credit for high school work done at home, for when a teenager enrolls in a public high school, coded from published district policy manuals, many hosted on BoardDocs, NEOLA or Simbli. Each district is coded on eight fields: whether a written rule exists; who decides; what evidence it asks for, such as a portfolio or exams; any exam pass mark; any cap on credits; whether grades are kept or shown as pass only; the minimum credits a student must earn in the district for its diploma; and how a family can appeal. Every code carries the policy number, revision date, link and quoted clause. It describes published rules only and gives no advice to any family.
Why it is still open
North Carolina tells families that a principal "has no legal obligation to accept home school credit" and has "the final authority" on credit, and written rules differ where they exist: New Albany Floyd County, Indiana, for one, decides "on a case-by-case basis" after a petition. The only district-by-district review this research found is EducationNC's 2016 reading of North Carolina's 115 policy manuals, which found 23 with credit rules; it covers one state and published no dataset.
Built with what exists
An AI agent with web browsing and a long-context model, which can find a district's policy manual, read the whole credit or transfer section and quote the clause behind each code; the NCES district directory to draw the sample; Python with requests and pdfplumber; a CSV or JSON schema and git to track revisions.
First step
Have the agent recode North Carolina's 115 district manuals and compare its count with EducationNC's 23 from 2016.
Done looks like
A random sample of at least 500 districts in at least 10 states with all eight fields coded and quoted; a second person codes 10% blind and reaches a Cohen's kappa, a standard agreement score, of at least 0.8; the release reports the share of districts with no written rule.