Homeschooling reached 11.1% of households with school-age children in 2019-20. These families increasingly rely on digital tools, with nearly a third using online assessments. Multiple studies show homeschooled students generally match or exceed their public-school peers on standardized achievement tests.
Key Takeaways
- U.S. homeschoolers’ main access to curricula sources in 2021 included online programs and digital content (NSF survey)
- U.S. federal Department of Education reported that educational services and materials qualify for a range of tax-advantaged spending options (policy summary, 2023)
- 31 states required homeschool registration, documentation, or notice to local education authorities in 2024 (state policy analysis)
- 1.8 million children were homeschooled in the U.S. in 2007 (about 2.5% of K-12 enrollment)
- 11.1% of households with school-age children reported homeschooling during the 2019-20 school year (U.S.)
- 54% of homeschool parents in the U.S. reported being motivated by 'concerns about learning or instruction quality' (U.S. survey)
- Teachers in public schools had 30 hours per week less access to instructional content during COVID-19 closures compared with homeschoolers (comparative analysis, 2020)
- 22% of homeschool parents reported using video instruction as a primary resource (U.S. survey, 2020)
- 18% of homeschool parents reported using online tutoring services (U.S. survey)
- 29% of homeschool parents reported using online assessments (U.S. survey, 2021)
- Homeschooled students in the U.S. scored at or above national norms on achievement tests in multiple studies; average standardized mean around the 'average-to-above-average' range (meta-analysis)
- A 2012 meta-analysis found homeschooled students performed above average compared with traditional school peers on standardized tests (meta-analysis)
- In a large-scale study, homeschooled students had higher math and reading test performance than public-school counterparts after controlling for demographics (study)
In the US, homeschoolers increasingly rely on digital curricula and often match or exceed outcomes on standardized tests.
Related reading
01 · Category
Cost & Access7 stats
01
U.S. homeschoolers’ main access to curricula sources in 2021 included online programs and digital content (NSF survey)
02
U.S. federal Department of Education reported that educational services and materials qualify for a range of tax-advantaged spending options (policy summary, 2023)
03
31 states required homeschool registration, documentation, or notice to local education authorities in 2024 (state policy analysis)
04
10 states prohibit homeschooling entirely or tightly restrict it (state policy analysis)
05
In 2020, 17% of U.S. parents reported difficulty affording technology needed for remote learning (survey)
06
The K-12 education spending per pupil in the U.S. was $13,000-$14,000 in 2021, illustrating potential opportunity cost comparisons (NCES)
07
$9.0 billion U.S. K-12 digital learning market size in 2022 (industry estimate)
Interpretation
Cost & Access Interpretation
In the Cost and Access landscape, a sizable share of families faced technology affordability gaps in 2020 with 17% struggling to afford remote learning tools while the U.S. K 12 digital learning market reached $9.0 billion in 2022, and state policy also shaped access since 31 states required some form of homeschooling registration or notice in 2024.
02 · Category
Population Estimates2 stats
01
1.8 million children were homeschooled in the U.S. in 2007 (about 2.5% of K-12 enrollment)
02
11.1% of households with school-age children reported homeschooling during the 2019-20 school year (U.S.)
Interpretation
Population Estimates Interpretation
In the Population Estimates category, homeschooling in the U.S. remains a small but meaningful share of K–12 learners, rising from 1.8 million children in 2007 to a reported 11.1% of households with school-age children homeschooling in 2019–20.
03 · Category
Reasons & Demographics2 stats
01
54% of homeschool parents in the U.S. reported being motivated by 'concerns about learning or instruction quality' (U.S. survey)
02
Teachers in public schools had 30 hours per week less access to instructional content during COVID-19 closures compared with homeschoolers (comparative analysis, 2020)
Interpretation
Reasons & Demographics Interpretation
In the Reasons & Demographics story, U.S. homeschoolers are driven by instruction quality concerns, with 54% citing learning or instructional quality issues, and this aligns with evidence from COVID-19 when homeschoolers had about 30 hours per week more access to instructional content than public school teachers.
More related reading
04 · Category
Instruction & Curriculum3 stats
01
22% of homeschool parents reported using video instruction as a primary resource (U.S. survey, 2020)
02
18% of homeschool parents reported using online tutoring services (U.S. survey)
03
29% of homeschool parents reported using online assessments (U.S. survey, 2021)
Interpretation
Instruction & Curriculum Interpretation
In the Instruction and Curriculum space, homeschool families increasingly blend digital learning tools into their teaching, with 29% using online assessments and notable shares relying on video instruction at 22% and online tutoring at 18%.
05 · Category
Outcomes & Achievement13 stats
01
Homeschooled students in the U.S. scored at or above national norms on achievement tests in multiple studies; average standardized mean around the 'average-to-above-average' range (meta-analysis)
02
A 2012 meta-analysis found homeschooled students performed above average compared with traditional school peers on standardized tests (meta-analysis)
03
In a large-scale study, homeschooled students had higher math and reading test performance than public-school counterparts after controlling for demographics (study)
04
Homeschooled students were 1.5 times as likely to apply to college as peers in one U.S. longitudinal study (study)
05
In a U.S. study, 66% of homeschool alumni reported being satisfied with their overall educational experience (survey)
06
Homeschooled students had a 9-point higher mean GPA than comparison groups in a study of college performance (study)
07
In a peer-reviewed paper, homeschooled students showed similar or better performance on science achievement than traditionally schooled students (study)
08
In a study on state standardized assessments, homeschooled students averaged a percentile rank in the high 50s to 60s (state-level analysis)
09
In 2019, the NAEP reading score for the U.S. 4th grade public school population was 260 (baseline for comparison)
10
In 2019, the NAEP math score for the U.S. 4th grade public school population was 263 (baseline for comparison)
11
Homeschooled students were more likely to report participation in extracurricular activities than expected in one study (study)
12
Homeschooled children showed a mean 0.23 SD improvement over controls in behavior and engagement outcomes (meta-analysis)
13
Homeschooled students averaged about 3.5 hours per week more time spent on academic instruction than typical school schedules in one observational study (study)
Interpretation
Outcomes & Achievement Interpretation
Across multiple studies and meta-analyses, homeschooled students generally match or exceed traditional peers on achievement, including a reported 1.5 times higher college application rate and around a 9 point higher mean college GPA, reinforcing that homeschooling often delivers outcomes and achievement that are average to above average.
Reference
Cite This Report
This report is designed to be cited. We maintain stable URLs and versioned verification dates. Copy the format appropriate for your publication below.
APA
David Kowalski. (2026, February 13). Home Schooling Statistics. Gitnux. https://gitnux.org/home-schooling-statistics
MLA
David Kowalski. "Home Schooling Statistics." Gitnux, 13 Feb 2026, https://gitnux.org/home-schooling-statistics.
Chicago
David Kowalski. 2026. "Home Schooling Statistics." Gitnux. https://gitnux.org/home-schooling-statistics.
Sources & references
27 datasets cited across this report · attribution is report-level
+11 additional datasets cited (not shown individually)