Read the balance sheet first, then the headline. The headline about China’s students being short on sleep is two years old and counting; the balance sheet behind it is what changes. The 2026 national white paper on sleep among secondary and college students, based on a survey of 30,000, puts the deficit in numbers an analyst can actually use: 72.5 percent of middle schoolers sleep less than eight hours, 67 percent of high schoolers get less than seven, and 50.8 percent of college students also fall below seven. Then come the two lines that should make an investor in human capital wince: 94 percent of college students report sleep problems, and only 8 percent wake up feeling fully rested.
One more observation for the commute: nobody argues that students should sleep less on purpose. The deficit is not a preference; it is a by-product of how the day is structured. And a by-product of a structure is something a structure can stop producing. That is the entire business case for reform — the highest-return asset available is the one already in the building, asleep.
Let me push the analysis one stage further, because a balance-sheet reading is only useful if it reaches the operating room. On the income side of the ledger, education promises returns that show up decades later — in careers, incomes, innovation. On the cost side, sleep is not even a line item; it is treated as background noise, the free resource that needs no accounting. But no resource is free, and the white paper is, in effect, a creditor statement: the system has been drawing on its students’ sleep at a rate the accounts cannot sustain, and the debt is now visible in 94 percent of college students reporting problems. A liability that large, unacknowledged on the books, is the definition of an off-balance-sheet risk.
There is a second market story hiding in the same numbers, and it concerns where the money goes when a family decides to respond. When a system generates a chronic deficit, households start self-medicating: parents buy tutoring for the daytime and supplements for the night, schools buy extended curricula, everyone buys the story that more is better. The aggregate effect is that money flows toward compensating for a schedule problem rather than fixing the schedule. An editor’s verdict is blunt: this is paying interest on a debt that should be refinanced at the source. The cheap fix — an earlier bedtime, a trimmed homework load — is the one almost nobody sells, because nobody profits from it.
What would a hard-nosed manager actually do with these numbers? First, treat the 66.8 percent homework driver as the operational finding it is, and commission the obvious experiment: a school that cuts homework by a quarter and holds quality constant. Second, treat the college 94 percent as a training failure, not a lifestyle footnote — which argues for teaching sleep as a skill in the same curriculum that teaches time management. Third, measure what matters: if the system genuinely wants learning, it should track retention and attention, not seat-hours. The math works on paper; it needs to be run in practice.
How does the math work here? Let me do what a business editor does on a train: decompose the problem into its line items. The white paper itself points to the biggest driver for middle schoolers — 66.8 percent say they go to bed late because of too much homework. That single line item explains most of the deficit. It is not a random distribution of late nights; it is a scheduled, homework-driven shortfall, predictable enough to appear in a survey of thirty thousand people. When a deficit is that systematic, it stops being a personal matter and becomes a structural one.
The quarterly rhythm tells you more than the press release, and the rhythm here is striking. The numbers worsen from middle school to high school, then shift shape rather than disappear in college. Middle schoolers are short on sleep because homework extends the evening; high schoolers are short because the academic day itself is longer and denser; college students are short for a different reason — 94 percent report sleep problems in a context where no one is forcing a schedule at all. The pattern is a depreciation curve with three distinct stages, and each stage needs a different fix.
Let me pause on the stage that is hardest to explain and most interesting: the college numbers. A university student has, in principle, the most control over a bedtime of anyone in the education system. And yet 94 percent report sleep problems. This is the tell. It suggests that the habits built during the homework years do not dissolve when the homework ends; they persist as an internal schedule. The system trained a generation to treat sleep as the residual — the thing that gets whatever time is left over — and the students have internalised the training. The math works exactly as designed, and the design was wrong.
I want to be healthily sceptical about one thing before I go further: the sample and the source. A white paper issued around an education conference, based on self-reported sleep, has methodological limits — people round their sleep times, and self-reports lean toward the tidy. Fair enough. But the direction of the error does not rescue the conclusion. Even if every figure were a few points optimistic, the pattern — a majority of middle schoolers under eight hours, driven largely by homework — survives any reasonable discounting. Scepticism adjusts the numbers; it does not delete the story.
Now let me do the balance-sheet reading properly, because that is the point of this column. Consider what the education system is actually producing. It takes in children and, over twelve years, outputs young adults with credentials and habits. On the asset side of that ledger, we count knowledge and skills. On the liability side, if we were honest, we would count the accumulated sleep debt — a liability that compounds, because chronic sleep loss impairs memory consolidation, attention and emotional regulation, which is to say it degrades the very asset the system is trying to build. You do not need to be a neuroscientist to see the incoherence: the system spends the most valuable resource it has — the students’ capacity to learn — on the theory that more hours on the task compensates, when the evidence says the opposite.
I found myself thinking about where the money goes in this trade, because that is the question a business editor always asks. Families spend on extra tutoring, on the assumption that more instruction equals more learning. Schools schedule longer days on the same assumption. But if the marginal hour of instruction lands on a student who has been sleep-deprived for years, the return on that hour is poor — the brain is not recording it well. In investment terms, the system is pouring capital into a channel with declining marginal productivity, while the highest-return asset available, the student’s rest, is systematically underfunded. That is a misallocation on the books.
Let me correct myself before an over-enthusiastic reading takes hold: I am not arguing that sleep explains all of educational performance, or that homework alone is the villain. Some homework is necessary; some late nights are unavoidable; individual variation is real. The claim is narrower and, I think, more robust: the scale of the deficit — three quarters of middle schoolers — is not a collection of individual choices but a system-level outcome, and system-level outcomes respond to system-level levers. Where the hours go is a policy question, not just a parenting one.
There is a market angle here too, and it is not trivial. Where a large cohort carries a chronic sleep deficit, a whole economy of supplements, devices and “sleep solutions” grows up to service the debt. Every percentage point of the deficit is a market in miniature: products that promise to repair what the schedule broke. A sceptical editor notes the irony — the system that creates the deficit also creates the demand for the cure — and the deeper point: repairing the schedule would shrink that market far more effectively than any product will. The healthiest outcome for the economy is the one that makes the symptom industry unnecessary.
A concrete moment to anchor the numbers. Imagine a middle schooler on a Tuesday night: homework done at eleven, alarm set for six-thirty, seven and a half hours of sleep — already under the eight-hour mark, and that assumes she fell asleep instantly, which almost no one does. Now multiply that evening by a school year, then by three years of middle school. The single night is unremarkable; the product of all those nights is the 72.5 percent. The white paper is not reporting exceptional cases; it is reporting the ordinary schedule. That is what makes the reading structural rather than anecdotal.
What would a serious fix look like on the ledger? It would start where the deficit starts: the homework line. If 66.8 percent of middle schoolers identify homework as the cause, then homework volume is the first line item to audit — not to abolish, but to examine for actual learning value. It would also treat the college stage as a habit problem rather than a workload problem: the university numbers point to schedules, not assignments, which suggests the fix is not fewer courses but better sleep architecture — fixed windows, later starts, institutional respect for rest. And it would measure outcomes the way any good manager would: not hours in the day, but quality of output — which, in education, is learning, and learning is exactly what sleep protects.
Fair enough — but don’t call this a moral lecture. The students are not failing; the schedule is. And the schedule can be changed, which is the one genuinely encouraging fact in the whole story. A deficit that is structural can be repaired structurally; a deficit that were only personal would be far harder to fix at scale. The white paper, read as an annual report, is a complaint about the books — and the books can be corrected. Where the hours go is a decision, and decisions can be revised. That is the difference between a hope and a plan.
In the end, the question an editor on a train keeps asking is simple: does the math work? The math of the current schedule does not — it produces three quarters of a cohort short on sleep, and it finances a symptom industry on the back of it. The math of the alternative is straightforward: restore the eight hours, and the marginal hour of tutoring becomes redundant, the waking hours become more productive, and the depreciation line stops growing. Read the balance sheet first, then the headline. The headline says students are tired. The balance sheet says we are spending our most valuable asset on its own decline.