Custody, compliance, and credentialing — and why the whole thing has to be compulsory
Education is full of C's. There are the four C's of twenty-first-century skills: critical thinking, communication, collaboration, creativity. There are classroom-management C's — connection, consistency, compassion. There are inclusive-education C's, teacher-formation C's, content-context-child. Enough to lose count.
However, every one of them describes what school should emphasize. Not one of them describes what school does.
So I want to offer three more, with an apology for the crowding, but no apology at all for the collision, because the collision is the point. The three C's of compulsory schooling are custody, compliance, and credentialing. They are not aspirations. They are the functions the institution cannot survive losing.
These Are Not a List; They Are a Stacked Chain.
The temptation with a triad is to treat the three items as parallel, three things school does, listed side by side. That's not what these are.
Custody creates the time. The child is held in a supervised place, on a fixed schedule, for a large fraction of childhood.
Compliance is what gets built during that time. Not the content, but the behavior. Arrive when told. Sit until released. Work on someone else's task, on someone else's clock, evaluated by someone who isn't your parent. Wait your turn.
Credentialing is the receipt. It converts twelve years of compliance into a portable signal that a stranger who was never in the room can read and trust.
Custody, then compliance, then credential. Each step builds on the prior one. It is a production line.
Why It Has to Be Compulsory
Each of the three C's breaks under voluntariness, for a different reason.
Custody needs universal coverage. Its whole value to parents, employers, and the wider economy is certainty that you can build a commute, a job, an entire household structure on children being supervised elsewhere from morning to mid-afternoon.
Compliance needs the ability to exit to be unavailable. Authority is only real where the child cannot leave.
Credentialing needs everyone in the same apparent starting arrangement, because comparison is the whole point: if every child sits in the same kind of room for the same years, whatever comes out the other end reads as a difference between children rather than between circumstances. Compulsion supplies the sameness, the system attributes the remaining difference to effort, and grades, rank, and the diploma report the result.
Three independent reasons for compulsion converge, which is the best explanation I know for why compulsory schooling went near-universal across wildly different cultures, economies, and political systems.
The American dates showcase the shift to compulsory education. Massachusetts compelled education in 1647 — the Old Deluder Satan law, which required that children be taught, not that they be sent anywhere. Massachusetts compelled attendance in 1852: children between eight and fourteen, twelve weeks a year, six of them consecutive. That two-century gap is the custody function being born. The obligation moved from the home to the building. Every other state followed, Mississippi last, in 1918, and the model was mostly Prussian, dating to 1763.
The industrial timing is obvious enough that it's worth being careful with. The tempting story is that factories wanted punctual, procedure-following workers and built schools to produce them. I don't think that story is accurate. Factory owners were often the opposition, not the demand: they were losing cheap labor, and keeping children out of the workplace was among the stated purposes of the laws. Alabama went so far as to temporarily repeal its law under pressure from a textile company. Whatever compulsory attendance was, it was not industry's project.
There's also a stranger finding that matters more. The laws did remarkably little to increase attendance. Schooling was already close to universal in several northern states before their legislatures got around to requiring it, and economists estimate that enforcement accounts for no more than about five percent of the rise in attendance across the period. What the laws changed was not the number. It was the norm.
Compulsion did not create the arrangement. It arrived after the arrangement had already emerged, and locked it in, thereby converting a thing most families were doing into a thing no family could decline. An arrangement that delivered custody, compliance, and credentialing proved so useful to so many parties at once that it spread on its own, and the law followed to make it permanent.
The Functional Fiction, and Why This Isn't an Accusation
Custody, compliance, and credentialing are loaded words, and I know how they read.
Compulsory schooling did not become the default condition of childhood across the entire planet because it is a scam. It became universal because it works, delivering real, valuable, expensive-to-replace outputs to nearly every sector at once.
Parents receive guaranteed daytime coverage, without which two-earner household formation is essentially impossible. That is one of the largest subsidies to workforce participation ever constructed, and it is invisible because it's never invoiced. Employers receive a pre-sorted, punctual, procedure-following labor pool at very little screening cost to themselves. The state receives continuous contact and influence with the entire rising generation: for nutrition, health screening, language, safety, civic formation. And children receive real goods: meals, supervision, literacy, friendships, and exposure to a world larger than the accident of their family.
The fiction isn't that the official story is false. Learning happens. Teachers care. Some students are transformed. The functional fiction is in the claim that those are the load-bearing purposes, that is, the ones that explain the funding, the compulsion, the architecture, and the persistence.
Which means that custody, compliance, and credentialing are what the institution cannot survive losing, what I called the operative function. Instruction, mentorship, curiosity, the moment a kid catches fire — those are real, and they happen daily. But they aren't structural. The system survives their absence, and we know this because it has been surviving their absence for a long time.
The spring of 2020 is when the machine said this out loud. Schools closed nationwide in March, and how long they stayed closed varied enormously. Some districts were back in person that fall, others remained remote for a full academic year or more. The tell isn't the duration. It's which argument moved policy. Learning loss was measured, publicized, and largely absorbed as a regrettable cost. What produced genuine urgency was that parents could not work, that essential workers had nowhere to put their children, that households and employers could not function without the school day. Schools reopened, and the case made in public was substantially an economic one. That was not cynicism. That was the operative function speaking in its own voice, briefly, before the narrative closed over it again.
The Mixture Varies
One refinement worth making: the three C's are not blended in equal proportion everywhere.
Elite private schools and selective magnets soften custody and compliance considerably: more autonomy, looser scheduling, fewer bells, and they lean almost entirely on the strength of the credential signal. Under-resourced schools lean the other way, heavily toward custody and basic behavioral compliance, with the credential doing comparatively little work in the market.
That's a pattern usually attributed to funding or culture. I'd suggest it's better read as different mixtures of the same three functions, chosen according to which one each population actually needs the institution to deliver.
Failures and Responses
Here my framework makes a prediction. Damage to an operative function, something the institution actually rests on, will produce urgency. Damage to the narrative purpose, the one the story places at the center, will produce reports.
If a decade of collapsing literacy had produced restructured schedules, altered calendars, and emergency legislation, while chronic absenteeism drew only commentary and concern, the framework would be wrong, and I would have to give it up. Conveniently, we don't have to speculate.
Start with instruction, which sits outside the operative stack entirely, making it a functional fiction: "learning is what we are here for." The 2024 national assessment showed reading declining again in fourth and eighth grade, steepening a decline already underway before the pandemic, with no state improving in either grade and fewer than a third of students reading at the proficient level. Twelfth-grade reading sits about ten points below where it was in 1992, and the lowest-performing students posted their worst scores in more than thirty years. A key detail is what the announcement revealed without meaning to: the acting director of the federal research institute noted that his predecessor had warned of this trend, and that her predecessor had warned of it as well, and that he was now warning of it too.
Three officials, a decade of measured decline, front-page coverage each time. And the building opens at the same hour with the same bells. No emergency session, no structural change, no interruption to the arrangement. In any other institution, that would be a scandal.
We can be careful here. Reading isn't a fiction, and the decline isn't trivial. Children are genuinely worse off, and the teachers working against it are doing real and difficult work. The fiction is only in the claim that instruction is what the institution rests on. And notice that the evidence for that claim isn't the decline itself, but the response to it: reports.
Next: grade inflation has run for decades alongside flat or falling achievement, which is the credential quietly loosening from the thing it certifies. Grade inflation has been absorbed with little more than commentary, because a degraded signal still issues a receipt and still sorts, just less precisely. AI is different, because it doesn't degrade the signal; it severs it. And the response has been immediate: honor-code rewrites, detection software, proctoring, a return to handwritten exams under supervision. Now notice what all that urgency is defending. Not one of those measures teaches a student to write better. Every one of them restores the conditions under which compliance can be witnessed. Faced with a threat to the credential, the institution did not move to improve learning. It moved to make the work verifiable again.
Compliance is also operative and load-bearing, and here the tone is sharp. Teachers leave over student misconduct at rates that falling test scores have never come close to producing. Phone bans arrived across states with a speed and near-unanimity that no reading initiative has achieved in fifty years. Districts that absorbed a decade of literacy decline without altering a schedule have rewritten policy within a single year over classroom disorder. And compliance is under attack from a second direction: teachers report, with a consistency that's hard to wave off, that sustained focus on an imposed task has become dramatically harder to obtain. Whatever is producing it (doom scrolling?), the capacity being reported as degraded is precisely the raw material compliance training works on. Sitting still, attending to something you didn't choose, on someone else's clock. Urgency here.
Custody is the framework's base, and it's the only failure that gets the law. Chronic absenteeism remains elevated years after the pandemic closures, and the response is not reports or policy debate. It's truancy enforcement, court referrals, and legal consequences for families. Urgency again.
Hopefully, you see the pattern. The one visible failure that is part of our functional fictions about education produced reports and nothing else, for a decade, across three commissioners. The other three, actual operative and load-bearing functions, each produced urgency. What we say we are doing and what we are actually doing in schooling becomes pretty clear.
The Squeeze
Put the pieces together and something uncomfortable emerges.
Compliance is being eroded from underneath by whatever is happening to attention. The credential is being severed from the top by AI, which makes the compliance it certified free to fake. Instruction was already ornamental long before either, which is the reading data, and it isn't a crisis so much as the background condition.
Neither crisis touches custody. If anything, they strengthen it: children who are harder to manage make guaranteed supervision more valuable, not less, and a household that cannot rely on the school day cannot function at all.
Consider who each C benefits. Custody benefits parents, employers, and the state: the guaranteed supervised day that households and economies are built on. Compliance benefits employers and the state too: the punctual, procedure-following disposition that makes a workforce and a citizenry manageable — and the institution itself, since it's what makes the room governable. Two of the three C's, then, benefit parties outside the arrangement. So two of the three C's benefit parties outside the arrangement. Credentialing is the only one that benefits the family and the child: comply for thirteen years and receive a portable receipt that converts into a livelihood.
The credentialing damage doesn't propagate downward. So the compulsion continues, the compliance is still extracted, and the one thing that was the family's consideration stops being worth much.
AI doesn't remove the differences between students; it removes their legibility. The work no longer testifies to who did it. But employers still have to sort, so the signal moves to whatever stays verifiable, which is likely to increasingly be provenance: which school, which selective program, which admission decision, made by which gatekeeper. And provenance is purchasable. A hardworking student at an indifferent school could once produce work that outperformed, and the receipt reported it. That was the channel by which effort escaped circumstance. Close it, and sorting reverts to where you came from. Circumstance is all that's left to read.
So yes: bifurcation. Parents moving districts, buying private school, competing for selective admission. And those purchases becoming more valuable, not less, as the generic signal decays. Though I'd be careful about blaming AI for it. The gulf has been widening for decades, and AI significantly accelerates a trend, although it didn't start it.
I can't predict how this resolves. But I can say what each path requires. If some witnessed, technology-free assessment becomes normal, effort becomes legible again without provenance, and the bifurcation slows. If nothing replaces the receipt, provenance becomes the sorting mechanism and the advantage compounds across generations.
Either way, custody is unaffected. Whatever happens to the diploma, whatever happens to attention, whatever happens to reading, the children will still be in the room at eight in the morning. That was always the point.
Addendum: What the 3 C's Explain About Homeschooling
Homeschooling has been regarded with suspicion, regulated heavily, and in some countries prohibited outright, and the hostility has never tracked evidence about learning outcomes. The three C's explain why, because each one generates a different objection, from a different enforcer.
Custody produces the legal objection. Homeschooling removes a child from the custodial net, and this is the one that gets written into statute. Look at what the actual legal fights concern: registration, notification deadlines, attendance reporting, home visits, oversight. They are almost never about curriculum or actual learning. That's custody speaking in operative urgency.
Compliance produces the cultural objection, and this finally decodes the "socialization" paradox. The socialization complaint has always been strangely immune to evidence. You can produce a homeschooled child with a dozen friends, a soccer team, and a part-time job, and the objection survives entirely intact. That's the tell that friendship or social interaction were never the core issues. What's actually absent is institutional compliance training: deference to non-parental authority, working on an imposed schedule, sitting still when you'd rather not. That can't be said out loud, so it comes out as socialization.
Credentialing produces the market objection, and this is the one that lands on the family as real cost. It's not exclusion, since homeschoolers do get into college and do get hired. The cost is that they must manufacture a substitute receipt at private expense: standardized tests, dual enrollment, portfolios, competition results. Which means the homeschooling family has often been quietly running a one-household version of the exam-prep industry: building a private credential apparatus because the public one wasn't issued.
Here is the mismatch that makes the whole debate so frustrating for everyone in it: homeschoolers defend themselves on intellectual growth. They point to test scores, reading lists, curiosity, capability — and they are frequently right. But that's an argument conducted entirely on the narrative layer, against objections that were never coming from there. They aren't naive for doing it. They're arguing on the layer the institution says it operates on. Taking a narrative at face value isn't a failure of intelligence; it's what a narrative is for.
