Education inequality: a seat in class is only the beginning

First, count who can take part

UNESCO's 2026 Global Education Monitoring Report estimates that 273 million children and young people of primary and secondary school age were out of school in 2024. That is an access measure covering several age groups; it does not describe how well enrolled students can read or calculate. Source: UNESCO's 2026 education monitoring summary.

Three questions belong beside any school statistic: who is included, what is being measured, and which year does it describe? Enrolment, attendance, completing a grade, and demonstrating a skill are related ideas. They are not interchangeable answers.

Then, ask what a school year makes possible

Imagine two fictional pupils who have each completed the same number of school years. One can read a short set of instructions independently. The other can recognise the words but needs help understanding what to do. Their calendars match; their next learning steps differ.

Now give both pupils the same science activity. The first can get started while the second is still untangling the instructions. That imagined difference says nothing about who is more curious, inventive, or capable of improving. It shows why counting years alone can miss something essential about an opportunity to learn.

What “learning poverty” actually measures

The World Bank's learning-poverty indicator combines reading proficiency near the end of primary school with the share of primary-age children outside school. It assumes out-of-school children are below the reading threshold. That is a measurement assumption, not an individual assessment of every child. Source: World Bank indicator methodology.

A joint report published in June 2022 estimated that 70% of ten-year-olds in low- and middle-income countries could not understand a simple written text. The pandemic-era estimate is often repeated, so its date and population matter: it is neither a 2026 reading assessment nor a percentage of all children worldwide. Source: World Bank and partners' 2022 update.

Keep that estimate separate from the global out-of-school count above. The age groups, coverage, years, and questions differ. Combining them into one percentage would make the result less informative, not more.

A more useful character sheet

Try sketching an imaginary learner with four spaces: access to lessons, time to practise, help when stuck, and skills already mastered. Leave room for their favourite subject, too. This is a thought experiment, not a statistical model or a description of how the app calculates outcomes.

Change one detail: the learner now has someone who explains a confusing paragraph patiently. What could tomorrow's lesson look like? Then change another: the learner misses a week but receives a useful catch-up activity. Thinking in concrete situations is more revealing than assigning a single permanent “education score.”

Keep the person larger than the indicator

Following a pixel life in 1 in 8 Billion can prompt the same question: what has this person had a chance to learn so far? A statistic can point toward unequal circumstances. It cannot tell us the full story of an individual's interests, judgement, humour, or future.

Read the numbers carefully, and leave the fictional character, and the real people behind the numbers, room to surprise you.