Institutions rarely die of criticism. They die of arithmetic. The classroom in rows has absorbed a century of eloquent critics without shifting a single desk, because its inputs were guaranteed: children arrived by law and demography, and its outputs were redeemed by an economy that hired what it produced. Both guarantees are now expiring at once, and a third force, exit, is compounding the squeeze. The question stops being whether the industrial model is good and becomes whether it can keep filling the rows.
The demographic squeeze
The quietest pressure is the birth rate. Across most developed systems, cohorts are shrinking, and a schooling model built as mass logistics loses its core subsidy when the masses thin: buildings sized for the past century's children, funding formulas that pay per enrolled head, closures that turn districts into managers of decline. Shrinkage changes the politics of schooling in a way growth never did. When every child is scarce, every family's threat of exit carries real fiscal weight, and systems that could once ignore dissatisfied parents must now compete for them. Scarcity converts parents from conscripts into customers, and customers get product changes that petitioners never do.
The labor-market squeeze
The second pressure attacks the model's warranty. Industrial schooling's implicit promise was always vocational: sit in the rows, store the information, and the economy will buy the result. The employer side of that trade is being repriced in public. The World Economic Forum's survey of over a thousand employers projects 22 percent of jobs disrupted by 2030, with 170 million roles created and 92 million displaced, and nearly 40 percent of required skills changing in the window (WEF, 2025). More surgically, Stanford payroll-data research finds employment for 22-to-25-year-olds in the most AI-exposed occupations already down roughly 16 percent relative to trend, while their experienced colleagues hold steady (Brynjolfsson, Chandar and Chen, 2025). The rungs being removed are precisely the ones schooling's memorize-and-certify pipeline fed: the junior roles made of retrievable knowledge and supervised routine. A model that sells preparation cannot survive indefinitely the visible unemployment of the prepared.
The exit
The third pressure is the one already measurable at the schoolhouse door. Tracking by the Johns Hopkins Homeschool Hub found homeschooling growing 4.9 percent on average across reporting states in 2024-25, more than double its pre-pandemic rate, with over a third of reporting states at all-time records (Johns Hopkins, 2025). Beside that runs the microschool wave: the National Microschooling Center's estimates put between 750,000 and 2.1 million American children in microschools as their primary schooling (National Microschooling Center, 2024-2025). The counting is soft, and honest analysis says so. But soft counting cuts both ways; what is not in doubt is the direction, and the fact that the exits accelerated after the emergency that supposedly caused them had ended.
Ricardo Semler's line names what the leavers are leaving. School in rows, as he has told it, is a factory from the 1800s operating without the factories; his own two decades of schools, some 3,000 children across six countries, replaced the rows with a project, a tutor, and a deadline, the shape, as he says, of work itself. The families building kitchen-table hybrids and twelve-student schoolhouses are mostly not quoting him. They are independently arriving at his arithmetic.
What the squeeze selects for
None of these forces makes an argument. That is what gives them power. Demography defunds the rows; the AI labor market devalues their product; exit depopulates them, one unmarketed decision at a time. Critics spent a century asking the industrial school to change its mind. The pressures now assembling ask it only to balance its books.
Predictions of the death of school have a long, embarrassing genealogy, so precision matters: buildings, teachers, and public funding are not ending. What is ending is the automatic cohort, the guaranteed child who arrives because there is nowhere else to go and leaves toward a job that no longer exists in the old form. Some cohort, and it may already be enrolled, will be the last for whom rows, bells, and memorization face no competing offer. After that, the industrial classroom joins every other industrial product in the condition it long escaped: having to be chosen.
Sources
- World Economic Forum (2025). Future of Jobs Report 2025. 22% of jobs disrupted by 2030 (170M created, 92M displaced); ~40% of skills changing. https://www.weforum.org/publications/the-future-of-jobs-report-2025/
- Brynjolfsson, E., Chandar, B. & Chen, R. (2025). Canaries in the Coal Mine? Stanford Digital Economy Lab. ~16% relative employment decline for early-career workers in AI-exposed occupations. https://digitaleconomy.stanford.edu/publication/canaries-in-the-coal-mine-six-facts-about-the-recent-employment-effects-of-artificial-intelligence/
- Johns Hopkins Institute for Education Policy, Homeschool Hub (2025). Homeschool Growth: 2024-2025. Average 4.9% growth across reporting states, double the pre-pandemic ~2%; a third of states at records. https://education.jhu.edu/edpolicy/policy-research-initiatives/homeschool-hub/homeschool-growth-2024-2025/
- National Microschooling Center (2024-2025). Sector analyses and enrollment estimates. Estimated 750,000-2.1 million US microschool students. https://microschoolingcenter.org/news-blog/how-many-students-do-microschools-serve