Learning Poverty: Reading to Understand, the Forgotten Emergency

An entire generation of children is moving through primary school without acquiring its most elementary skill: reading and understanding a sentence. In West Africa, classrooms have filled up, diplomas are handed out, and yet the majority of pupils leave the cycle without mastering a text suited to their age. This crisis does not show up in enrollment statistics, which are reassuring. It hides in what school actually produces, and what it produces, when measured, is alarming: in Sub-Saharan Africa, 89 percent of 10-year-olds are in learning poverty. The cost of this blind spot is not paid in missing classrooms, but in decades of mortgaged growth.

A measured crisis, not a hunch

Learning poverty is not a field impression, it is a precise indicator. Developed by the World Bank and the UNESCO Institute for Statistics, it combines two realities: children out of school, and those who, although enrolled, fail to reach the minimum reading proficiency level by the end of primary school. Its reading is damning for West Africa. Benin posts a rate of 56 percent, Senegal 69 percent, Burkina Faso 74 percent, Ghana 80 percent, Togo 82 percent, Mali 90 percent and Nigeria 92 percent. In other words, in Togo four children out of five cannot read and understand an age-appropriate text by age 10, and in Nigeria nine out of ten are affected.

This indicator has a rare virtue: it makes visible a failure long masked by access figures. For two decades, the regional debate centered on the number of children sitting in class. Learning poverty shifts the question to what happens once they are seated, and the answer forces a rethink of the very definition of educational success. Enrolling a child who will not learn to read does not solve the problem, it simply makes it harder to detect.

A worsened situation, but not an immutable one

Placed in time, the crisis reveals two opposing movements. Globally, learning poverty in low- and middle-income countries rose from 57 percent before the pandemic to 70 percent in 2022, a 13-point jump driven by Covid-19 school closures. Sub-Saharan Africa, already at the top, climbed from 86 percent to 89 percent over the same period, 3 additional points. The region declined less than others, not out of resilience, but because it started from a level so high that it left little room for further deterioration. The contrast between Latin America's sharp drop and Sub-Saharan near-stability tells a single story: a learning foundation that was never firmly laid absorbs shocks differently from one that collapses.

The implicit message of these trajectories carries hope: if learning poverty can worsen by 13 points in two years under a shock, it can, by symmetry, recede under the effect of deliberate policies. In 2019 the World Bank set itself the goal of halving global learning poverty by 2030. At the current pace, that goal will not be met, but the fact that it was quantified changes the nature of the debate: the crisis is now steerable, provided it is measured.

Enrolling is not educating

The regional paradox fits in one sentence: access has advanced far faster than quality. Children enter school, but many do not complete the cycle, and those who stay do not learn the essentials. Primary completion rates confirm it, from Ghana (98 percent) to Senegal (63 percent), Benin (62 percent), Mali (48 percent) and Niger (41 percent). When one pupil in two leaves primary school without finishing in the Sahelian countries, and the majority of those who finish cannot read fluently, the system produces diplomas without producing skills.

This disconnect between presence and learning has a mechanical logic. Learning poverty adds together two populations: children never enrolled or who drop out, and those who stay in class but fail to reach the reading threshold. In the Sahelian countries, where completion plunges below 50 percent, the first component weighs heavily. In Ghana, where completion approaches 98 percent, the learning poverty rate is nonetheless 80 percent: almost all children reach the end of primary school, but four out of five leave it unable to read. This is the clearest demonstration that the problem is not only one of access. You can fill a school and keep failing.

No development policy is sturdier than the foundational skills of those who will have to carry it. A child who cannot read by age 10 is a promise of growth already compromised.

Why school does not teach reading

Behind the aggregate figure lies a chain of failures that accumulate, year after year, along the path of a single child. Four factors structure the gap between attending a class and learning to read in it.

• The language of instruction. In a large part of the region, the child approaches reading in a language (French, English) that is not spoken at home. The child learns to decode sounds without accessing meaning, durably delaying comprehension, the very core of the indicator.
• Effective learning time. Between teacher absenteeism, overcrowded classes and a truncated calendar, the number of hours actually devoted to reading is far below the theoretical time. The pandemic school closures worsened this deficit, adding 13 points of learning poverty worldwide.
• Teacher training and steering. Without a structured method for teaching reading or early assessment of skills, the teacher has no signal to catch up a pupil who falls behind from the first grade. The failure is discovered too late, at the end of the cycle.
• Dropout along the way. Where completion falls to 41 percent (Niger) or 48 percent (Mali), poverty, child labor, distances and insecurity remove the child from school before reading has been consolidated.

None of these factors is a geographic fate. They are public policy parameters: choice of the language of literacy, allocation of school time, training arrangements, retention safety nets. The fact that rates range from 56 percent to 92 percent between neighboring countries, subject to the same climatic and economic constraints, demonstrates that these choices produce different results. Above all, these factors do not add up, they multiply: a child who approaches reading in a foreign language, in an overcrowded class, with a teacher unequipped to spot the falling behind, accumulates three handicaps, none of which will be corrected if the child leaves school mid-cycle. Learning poverty is the end product of this cascade, not the sum of isolated causes that could be treated one by one.

A continental divide, a West African epicenter

The imbalance does not play out only inside schools, it weighs on the map of the world. Central and West Africa is home to almost a quarter (around 24 percent) of the planet's out-of-school children aged 6 to 18, even though it accounts for only 12 percent of the world's child population. Conflict, poverty and gender inequality combine to durably exclude millions of children, with Nigeria alone counting 10.5 million primary-age children out of school despite schooling that is in principle free and compulsory. At the Sub-Saharan scale, the picture comes down to a single figure: 89 percent of children are in learning poverty.

The double disadvantage of girls

Learning poverty strikes both boys and girls, but the exclusion that feeds it is not neutral. In low-income countries, only 38 percent of girls complete lower secondary school, compared with 43 percent of boys. The gap widens precisely at the moment when reading should consolidate and deepen. Early marriage, domestic burdens and family trade-offs under the constraint of poverty remove girls from school before boys. Any strategy to reduce learning poverty that does not disaggregate its results by sex risks leaving this half of the cohort even further behind.

The cost of inaction

A reading crisis at age 10 does not stay confined to school: it becomes a productivity crisis twenty years later. The World Bank puts the present-value loss of earnings this generation risks suffering worldwide at 21 trillion dollars, the equivalent of 17 percent of today's global GDP. A child who does not acquire the fundamentals only with difficulty develops the technical and higher-order skills demanded by increasingly demanding labor markets. For West African economies counting on their demographic dividend, the arithmetic is harsh: a youth that is numerous but poorly literate is not a growth asset, it is a liability.

The demographic dividend is not automatic. A large cohort that cannot read turns into a liability where a cohort that can read becomes an engine.

The cost is not limited to the pay slip. A population that does not read fluently has less access to health, agricultural and civic information, votes on more fragile foundations, negotiates its place in the formal economy less well. Learning poverty thus feeds the cycle it prolongs: a poorly literate parent supports a child's schooling less effectively, passing the disadvantage to the next generation. It is an invisible tax, paid every year, that appears on no budget yet weighs on all downstream public policies. And because it is paid silently, over twenty years, it escapes the annual budget trade-offs that nonetheless decide the future of each cohort. The paradox is cruel: the investment with the highest long-term return, reading the fundamentals, is also the one whose payoff is most deferred, and therefore the easiest to sacrifice under pressure.

What national averages hide

A national learning poverty rate is an average, and every average conceals as much as it reveals. Behind Benin's 56 percent or Ghana's 80 percent lie massive gaps between urban and rural areas, between girls and boys, between wealthy regions and conflict zones, between well-resourced schools and ghost schools. A ministry that steers on the national average alone acts blindly on the pockets where the crisis is most acute, and risks concentrating its resources where they change the situation least.

This is precisely where the value of fine-grained measurement comes into play. A program such as AIM4Learning, deployed by the World Bank in Eastern and Southern Africa, illustrates the logic: it targets more than 70 million children and has set itself the goal of bringing the number of out-of-school children down from 14 to 7 million by 2034. Such quantified and dated targets are only possible because the data exists at the right level of granularity. You cannot steer a goal you cannot measure.

The CRAD angle: measuring where the decision is made

CRAD's conviction is that the exit from learning poverty depends first on a revolution in measurement. A regional rate of 89 percent is outrageous; it tells a school principal or a mayor nothing about what to do on Monday morning. Useful data is disaggregated (by sex, age, mother tongue, area), geolocated (down to the municipality and the school) and early (from the first grade, through early grade reading assessments, and not at the end of the cycle when failure is already sealed). It is this fine-grained data that turns a global finding into a local action plan: it identifies the school that is falling behind, the cohort of girls evaporating, the area where the language of instruction blocks comprehension. Measure early, measure low, measure disaggregated: that is what separates an advocacy statistic from a steering instrument. Building these measurement and evaluation systems in the service of West African decision-makers is CRAD's core business.

← All insights