In the mid-2020s, the Global South is being described as the world's growth engine. That phrase hides an uncomfortable detail. An engine is useful even when the driver is disposable.

Across large parts of Africa, South Asia and fragile island economies, the education story is no longer "catch-up". It is divergence. A slice of young people is accelerating into the AI era, while a much larger slice is not being equipped for modern work at all. The gap is not merely unfair. It is strategic. It shapes who gets to invent, who gets to obey, and who gets to emigrate.

The warning lights have been blinking for years. One of the cleanest global signals is blunt: hundreds of millions of children and youth remain out of school. UNESCO and partners have put that number at 244 million in 2022, rising to 251 million in 2024.

A second signal is even more brutal, because it includes children who are enrolled: "learning poverty", the share of children unable to read and understand a simple text by around age 10, has been estimated at about 70% in low- and middle-income countries.

So the education crisis is not only access. It is quality. It is time. It is the slow violence of classrooms that exist, but do not teach.

These numbers expose a system that is diverging rather than converging. The out-of-school population has decreased by only 1% in nearly a decade despite 110 million more children entering school since 2015. Completion rates tell a similarly stagnant story: primary completion has increased by less than 3 percentage points to 87% since 2015, while secondary completion has increased by less than 5 percentage points to 58%.

Progress is not impossible. It is simply too slow. At current rates, universal education by 2030 has moved from ambitious to mathematically implausible. Regional disparities remain stark. More than half of all out-of-school children and adolescents in the world are in Sub-Saharan Africa alone. Less than 30% of youth in the region complete upper secondary education.

But the numbers tell only half the story. The other half is quality. Of the children who are in school, 70% in low- and middle-income countries cannot read and understand a simple text by age 10. In the poorest countries, this figure reaches 91%. This is not illiteracy as exception. This is illiteracy as norm.

Governments can deliver speeches about "human capital" for decades. Budgets reveal what they actually believe.

UNESCO's own Education 2030 financing guidance repeats a benchmark that policymakers quote and then quietly ignore: countries should aim to allocate 4–6% of GDP and/or 15–20% of public expenditure to education.

That benchmark matters because it is not a moral slogan; it is a capacity threshold. Below it, systems struggle to hire and retain teachers, maintain facilities, fund learning materials, and modernise curricula at scale.

The World Bank's education-spending indicator exists precisely because the question is measurable: how much is actually being spent?

When countries consistently undershoot, they do not merely "invest less." They choose a different model of society.

The financing crisis extends beyond government budgets. In many low- and middle-income countries, households bear the burden that public systems cannot carry.

In Nigeria, Haiti, and Lebanon, households contribute over 70% of total education expenditure. In some African countries, enrolling a child in a private school costs families 1.5 to 5 times more than public school fees. The share of household spending on education is inversely related to public spending, a predictable but devastating pattern: where governments invest least, families pay most.

This creates a secondary exclusion mechanism. Even in countries with nominally free public education, the hidden costs (uniforms, transport, materials, exam fees, informal payments) push poor families toward impossible choices. Education becomes expensive precisely where incomes are lowest.

The result is a quiet sorting machine. The poor are excluded early. The middle class finances credentials through debt. The elite secures global mobility. What appears as educational diversity is actually systematic stratification by income.

A pattern repeats across developing economies, including small island states and commodity-dependent countries.

Education systems become pipelines into narrow labour markets: hospitality, low-wage services, basic manufacturing, plantation-linked supply chains, outsourced back-office work, security, and logistical support for someone else's value chain. In this model, "skills" means compliance and repetition. The brightest students often exit the system physically (migration) or psychologically (credential-chasing with minimal belief).

This is why education underfunding is not a side issue. It becomes the upstream cause of everything your readers track downstream: low productivity, weak innovation, fragile tax bases, and a population that sees the state as a manager of limits rather than a builder of futures.

Every technological shift creates winners and losers. AI is different because it compresses time.

In one part of the world, teenagers are learning to prompt, code, automate, design, and build businesses at almost zero marginal cost. In another, teenagers are leaving school early, cycling through under-resourced classrooms, or graduating without literacy strong enough to benefit from technology at all, which turns AI from "opportunity" into "replacement".

If learning poverty is widespread, and the World Bank's estimate suggests it is, then the Global South is not simply facing a "skills gap." It is facing a comprehension gap, which is far harder to repair later.

This is how a society ends up with two futures living in the same street: one child building digital leverage, another locked into a low-wage treadmill that the global market can undercut at any moment.

The AI era does not wait for education systems to catch up. It accelerates past them. And when 70% of children in low- and middle-income countries cannot read by age 10, the question is not whether they can learn to code. The question is whether they can learn to read the instructions for the technology replacing them.

Behind the statistics of children out of school and learning poverty lies a more fundamental crisis: the global teacher shortage.

UNESCO estimates that 44 million primary and secondary teachers are needed worldwide by 2030. Seven out of ten will be needed at the secondary level, and over half of existing teachers will have left the profession by decade's end.

Sub-Saharan Africa faces the most severe shortage, needing 15 million new teachers. Ninety percent of countries in the region face serious secondary teacher shortages; 70% face primary teacher shortages. Southern Asia follows with the second-largest deficit: 1.7 million primary teachers and 5.3 million secondary teachers needed.

The numbers translate into classroom reality. In low-income countries, primary teachers face an average of 52 pupils per class, double the global average of 26. In some countries, the ratios are extreme: Central African Republic (80 pupils per teacher), Malawi (69), Ethiopia (64), Chad (62).

But quantity alone does not capture the crisis. In 31 of 96 countries with recent data, less than 80% of primary school teachers were trained according to national standards. More than half of these countries are in Sub-Saharan Africa. In Madagascar, there are 250 pupils per trained teacher. In Central African Republic, 138. In Chad, 96.

These are not teachers who need refresher courses. These are classrooms where instruction has become impossible, where the concept of "teaching" collapses into crowd management, where learning poverty becomes inevitable not because students cannot learn, but because there is no one adequately trained to teach them.

There is a second trap: even where education is available, it is increasingly costly to families. Fees, uniforms, transport, tutoring, exam preparation, informal payments, opportunity costs. The result is a quiet sorting machine. The poor are pushed toward early work; the middle class toward debt-financed credentials; the elite toward global mobility.

Then comes the cultural ritual: degree, job, loan, car, house, marriage, children, and repeat. It produces social stability on paper, but not intellectual sovereignty. It trains young people to climb ladders that lead to crowded rooms, not to build new rooms.

Meanwhile, the world that matters economically is being re-coded. If your population cannot read well, it cannot bargain well. It cannot audit. It cannot innovate. It cannot negotiate with the future.

In institutional language, underfunding is "insufficient public expenditure." In human language, it is teachers stretched thin, underpaid, and exhausted. Schools functioning as storage, not instruction. Curricula tuned to passing exams, not mastering capability. Young people treated as labour supply rather than creative capital. A permanent sense that the state expects obedience, not brilliance.

That is why the education emergency belongs in an Outlook about stability. It shapes legitimacy.

A government can deliver GDP growth and still lose its young if schooling does not translate into dignity. And when belief collapses, the state learns a costly lesson: coercion is always more expensive than credibility.

The Global South does not lack intelligence. It lacks investment that matches the rhetoric.

If 251 million children and youth remain out of school, that is not an "education sector problem". It is a macroeconomic emergency, a security risk, and a moral failure, all at once.

If learning poverty sits near 70% in the countries that house most of the world's children, then the world's development story is being written on top of a comprehension crisis.

If primary completion has increased by less than 3 percentage points in a decade, if secondary completion lags at 58%, if one country alone (Nigeria) contains 20.2 million out-of-school children, then universal education by 2030 is not delayed. It is abandoned.

And if states continue to miss the financing benchmarks they themselves endorse, if high-income countries invest 155 times more per learner than low-income countries, if teachers in Madagascar face 250 pupils per trained teacher while earning less than $7,500 annually, if households in Nigeria must finance 70% of education costs, if the economic cost of educational failure approaches $21 trillion while the financing gap is only $97 billion, then the "lost generation" is not a tragedy.

It is a policy outcome.

By 2026, the defining divide will not only be between rich and poor countries. It will be between societies that treat education as sovereignty and those that treat it as decoration.

The warning lights are blinking. They have been blinking for years. The data is unambiguous. The cost of inaction is quantified. The financing gap is measurable. The teacher shortage is documented. The learning poverty is tested. The household burden is surveyed. The completion rates are tracked.

The only question now is whether anyone with power is still watching.