School is a bundle. Families are starting to open it.
Something quiet is happening at the level of enrollment data, and it is worth reading carefully before reaching for a frame.
In the United States, approximately 3.4 million children are now homeschooled — roughly 6.3 percent of the K–12 school-age population, growing at about 5.4 percent per year. That rate is nearly triple the pre-pandemic baseline. Thirty-six states are at all-time highs. Running alongside this: an estimated 95,000 microschools now serve more than one million students. Combined, these numbers represent something like one in twelve American children whose families have chosen to renegotiate, in some material way, what school means for their household.
The instinct is to sort these families into political bins. Left-leaning, right-leaning. Secular, religious. Urban, rural. The sorting is satisfying because it gives the data a story with sides. But it is not the most interesting thing the data is doing.
The more interesting thing: families are unbundling school.
What school is actually a bundle of
A school is not one thing. It is several things held together by the architecture of the school day — a bundle assembled over roughly two centuries and delivered as a unit so consistently that we have almost forgotten it is a bundle at all.
At minimum, a conventional school contains: supervised childcare (somewhere for children to be while parents work), academic instruction (the transfer of curriculum), peer socialization (sustained exposure to age-cohort peers), credentialing (the diplomas, transcripts, and records that downstream institutions recognize), and something harder to name — the cultural transmission function, the slow deposit of shared references, rituals, and identities that locate a child inside a community.
These five things are not naturally the same thing. They happen to be delivered together. The bundle holds because, for most families, the cost of separately sourcing each component is higher than the cost of simply sending a child to school. The bundle is a convenience — a genuine one, and for most families, still the right call.
What the exit data is showing is that the convenience calculus is shifting at the margins. Not dramatically. Not irreversibly. But measurably, and in a direction that seems to be accelerating rather than stabilizing.
What families are exiting from
The question that sits inside the 6.3 percent figure is not "why are these families against schools?" Most are not against schools in any abstract sense. The more precise question is: which specific components of the bundle stopped working for them, and which components are they trying to source differently?
This is where the data gets interesting, because the exits are not uniform. Some families are exiting the academic-instruction component — finding, through microschool or home curriculum, that they can deliver their child's core academic work in fewer hours and with more pacing flexibility than a thirty-student classroom allows. The two-hour learning thesis that Alpha School built its model around — the idea that adaptive, individualized academic practice can compress a full school day's core instruction into approximately two hours — is at least a partial articulation of why a family might conclude that conventional classroom pacing is not the best use of their child's morning. The thesis remains contested and the performance data is self-reported, but the underlying intuition (synchronous group instruction involves a lot of waiting for others to catch up or falling behind if you need more time) is not obviously wrong.
Other families are exiting the socialization component — not because they want their children to have fewer peers, but because they are sourcing peer contact differently, through co-ops, community sports, religious community, neighborhood life. The microschool form, at its most coherent, is a deliberate attempt to keep the peer-socialization component while reducing the administrative overhead of a full institutional school.
Still others are exiting the cultural-transmission component. Not all cultural transmission, but the specific version of it that a state curriculum delivers. A family that believes their child's deepest cultural formation should come from somewhere other than a standardized national syllabus is making an identity-track argument: school is a statement about what a community believes matters, and we want that statement to be made by us, not by a bureaucratic syllabus process we have no access to.
This last group is the most interesting one to sit with — not because their position is correct, but because it names something true about what school is. A curriculum is an identity claim. It is a community's decision about what is worth remembering, what habits of mind are worth forming, what history is worth knowing. That decision, in the conventional school, is made upstream of the family: by curriculum committees, by state standards, by textbook publishers. The exit from it is a statement that the family wants to make that decision themselves, or at least to be much closer to who is making it.
What they are reaching for
The geography of the exit is not random. South Carolina recorded 21.5 percent year-over-year growth in homeschool enrollment. Several Sun Belt states show similar curves. The growth is also statistically undercounted: hybrid-enrolled students — children who attend a conventional school part-time and homeschool the rest — often do not appear in the federal homeschool numbers. The actual exit from full-time conventional schooling is larger than the visible figures suggest.
What these families are reaching for varies. Some are reaching for compressed academics and more control over their child's time. Some are reaching for a peer environment they trust more. Some are reaching for the explicit right to make identity-formation decisions — religious, cultural, national — that the institutional school cannot make for them.
The Alpha School model, the Acton Academy lineage it descended from, the broader microschool category — these are attempts to answer the bundle problem with a different bundle. Not no school, but a differently-shaped school: one that keeps the peer contact and the academic instruction but sheds the institutional overhead, the administrative credentialing machinery, and in some cases the curriculum authority.
Whether these alternative bundles are working — whether students in microschools are learning as much or more, socializing as well or better, forming identities as coherently or more so — is a genuinely open question. The evidence base is thin because these schools are largely invisible to federal data collection. The exit is happening faster than the measurement system can track it.
Why the identity question is the lasting one
The components of the bundle that are most amenable to unbundling are the administrative ones. Supervision can be distributed. Academic instruction can be mediated by software. Credentialing is starting to be disrupted by alternative credentials and employer preferences that are moving away from transcript-reliance. These are real shifts, and they are worth watching.
The component of the bundle that is not easily unbundled is the cultural transmission one. This is not because families cannot choose their own cultural formation — clearly they can, and many are doing exactly that. It is because cultural transmission, at its deepest, is a community act, not a household act. The school, at its best, is a place where a community's accumulated sense of what matters passes from the people who hold it to the people who will hold it next. That passing requires the community to be in the room.
A family that exits conventional school entirely has to build the community context from scratch, or find it in existing communities (religious, ethnic, local) that can hold the transmission function the school used to perform. Some families do this brilliantly. Most find it harder than the academic unbundling.
This is the question the exit data is ultimately asking, beneath the enrollment percentages and the political attributions. Not "should school survive?" — it will. Not "are homeschoolers right?" — they are making different bets, not a single coherent argument. The question is: what happens to the parts of school that only work when a community does them together, if enough families exit that the community is no longer assembled?
We are not at a point where that question has a crisis attached to it. The exits are marginal in the aggregate, even where they are dramatic in specific states. But the direction of the unbundling is clear, and the identity-transmission component of the bundle is the one that no adaptive software, no AI-powered tutor, and no home curriculum product can fully replace. It requires people who share something — a community, a culture, a set of stakes — to be in proximity.
School, at its core, has always been that. The current unbundling is a reminder to name it explicitly: the parts of school that are bundle-components can be unbundled. The part that is community cannot. The interesting design question, for families and for the education field alike, is which part of the bundle they are actually trying to keep.
Sources
Wiki pages drawn from
concepts/two-hour-learning— compressed-academic model; Alpha School's morning block; Acton Academy lineage; evidence base and open questions.entities/companies/alpha-school— Alpha School detail: campuses, tuition, 2-hour claim, AI framing vs. actual deployment, reception.concepts/identity— Thewhat's Identity R&D path; the coincidence-moments framing; cultural transmission as the core Identity bet.
External sources
- National Home Education Research Institute (NHERI) — 2024–25 homeschool enrollment estimates. https://www.nheri.org/
- National Microschooling Center — microschool count and student estimates. https://nationalmicroschoolingcenter.org/
- Johns Hopkins Institute for Education Policy — 2024–25 homeschool growth summary. https://edpolicy.education.jhu.edu/