The pandemic was supposed to be homeschooling's high-water mark: an emergency spike that would drain away once the buildings reopened. The drain never came. Tracking by the Johns Hopkins Institute for Education Policy's Homeschool Hub found homeschooling still growing in the 2024-25 school year at an average rate of 4.9 percent across states with data, more than double the roughly 2 percent annual growth typical before the pandemic. Eight in ten reporting states showed increases, and more than a third recorded their highest homeschool numbers ever, above even the peaks of the closure years (Johns Hopkins Homeschool Hub, 2025). Whatever this is, it is not a spike decaying. It is a base rate that moved.
The growth is quiet by structural necessity. Every other schooling sector has an institution whose job is to announce it: districts have communications offices, charter networks have advocacy arms, private schools have admissions marketing. Homeschooling's growth is the sum of hundreds of thousands of private decisions with no press release, visible only when a state agency tallies notification forms, and many states barely tally. The most consequential enrollment shift of the decade is documented approximately, in arrears, by researchers assembling scraps.
What changed in the who and the why
The composition of the movement has shifted under the numbers. The homeschooling of caricature was ideological, a project of religious separatism at one pole and countercultural conviction at the other. Those families remain, but the post-2020 entrants look different: parents who discovered during the closures that they could see the product up close, and did not like what they saw; families of children for whom the standard classroom was a bad fit, bullied children, anxious children, children parked in rows and bored past endurance; families assembling education from parts, a co-op two days a week, online courses, a tutor, sports through the town. The closures functioned less as an emergency than as a product trial. Some fraction of the customer base, having inspected the product, declined to return it to the shelf.
The boundary of the category has blurred accordingly. A child who spends three days a week at a twelve-student learning center staffed by a hired teacher is, depending on the state's forms, a homeschooler, a microschooler, or a private school pupil. The National Microschooling Center's sector data shows this adjacent world scaling on exactly the terms hybrid families need, with a median enrollment of sixteen students and, in the 2025 survey of 800 schools, 74 percent charging at or below $10,000 (Soifer and Soifer, 2025). "Homeschooling" increasingly names a legal status, not a location: education routed through the family rather than through an institution, wherever it physically happens.
What the exit is measuring
It is tempting for defenders of conventional schooling to read the numbers as a policy problem, and for libertarians to read them as vindication. The more useful reading is diagnostic. Exit, in any system, is information. Families absorbing the full cost of an alternative, forgone income, forgone convenience, social friction, are paying a price to escape something, and the price they will pay is a measure of the something. When exits accelerate after the emergency ends, the revealed judgment is not about the emergency.
Ricardo Semler built his critique of that something into an institutional alternative rather than a family one, but the diagnosis rhymes. As he has told it, his two decades of schools, some 3,000 children across six countries, ran without rows, bells, or grades for memorization, on the argument that school in rows is a factory from the 1800s in a world without those factories; the working unit is a project, a tutor, and a deadline. Homeschool families, mostly without reading a word of the literature, keep rebuilding that same unit on their kitchen tables: a thing the child is making, an adult who knows the child, a rhythm the family owns.
The honest caveats, and the signal anyway
The caveats are real. State data is patchy; definitions vary; growth rates on modest bases can flatter; and homeschooling contains real failure modes, isolation, neglect, curricula of pure ideology, that aggregate statistics cannot see. The sector's outcomes are as unmeasured as its size.
But the direction of the signal survives every caveat. Five years after the buildings reopened, the families keep leaving, faster than they did before anyone had heard of lockdowns. Institutions can argue with critics. It is much harder to argue with the back of a customer walking out, quietly, in record numbers, in thirty states at once.
Sources
- Johns Hopkins Institute for Education Policy, Homeschool Hub (2025). Homeschool Growth: 2024-2025. Average growth of 4.9% across reporting states, vs ~2% pre-pandemic; 80% of reporting states increasing; over a third at record highs. https://education.jhu.edu/edpolicy/policy-research-initiatives/homeschool-hub/homeschool-growth-2024-2025/
- National Microschooling Center / Soifer, D. & Soifer, A. (2025). American Microschools: A Sector Analysis. 800 schools surveyed; 74% at or below $10,000 tuition; median enrollment 16. https://microschoolingcenter.org/news-blog/2025-american-microschools-sector-analysis