ASK three sources how many Pakistani children are out of school and the answers vary dramatically. The figure repeated in policy speeches and donor briefs is roughly 25 million. The Pakistan Economic Survey 2025–26, citing household data, says the out-of-school share fell from 38 percent in 2023 to 28 percent in 2025. ASER Pakistan 2025 finds 92.2 percent of children aged six to sixteen enrolled, implying roughly five million out of school. Those estimates deserve comparison, but they do not measure precisely the same population. The divergence is largely methodological. A percentage cannot become a national headcount without an agreed denominator, while different age bands produce different results. HIES is the Pakistan Bureau of Statistics’ own survey; ASER is an independent household assessment, while the familiar 25 million estimate comes from a different statistics series and reference period. Surveys also differ in counting madrassa and non-formal enrolment. The gap between five million and 25 million therefore calls for transparent reconciliation, not a verdict that one source is wrong.
This is more than a statistical quarrel. A district planning new classrooms needs to know how many children have never enrolled, how many have left and where they live. A national announcement may celebrate a falling percentage while a rapidly growing population keeps the number of children outside school stubbornly high. Both measures belong in every official progress report. Article 25-A commits the state to free and compulsory education for children aged five to sixteen. Federal and provincial departments, development partners and the public need a comparable measure of progress towards that obligation. A shared baseline would help target budgets, identify gaps and give credit for genuine improvement wherever it occurs.
Precision matters within provinces too. Punjab’s 2026 baseline estimates that some 6.4 million children have never enrolled, while another 3.16 million enrolled and later dropped out. These groups require different responses: outreach and access for the first, sustained attendance and support for the second. A single provincial total conceals the reason a child is missing from class and therefore the remedy. Yet counting children through the gate answers only half the question. ASER 2025 finds that only 7 percent of Class 3 children can read a simple story in Urdu or Sindhi and only 8 percent can perform division. Even by Class 5, only about half can read a Class 2 text and 48 percent can divide. Outcomes are strikingly similar in government and private schools alike and girls perform as well as boys and sometimes better. The central message is unmistakable: attendance alone does not ensure learning.
If enrolment rises while literacy and numeracy lag, the education emergency changes shape. The child who sits through years of lessons without learning to read faces narrower choices in every later subject. Parents may see a uniform and a timetable, while the system quietly misses its most basic promise. Provincial systems already collect enrolment and facilities data. A regular, district-level account of what children can read and calculate would complete the picture. It would also help officials distinguish a school with empty seats from one with full classrooms but little learning.
Resources remain tight. The Economic Survey records education spending of 0.8 percent of GDP for July–March 2024–25, a partial-year figure. Annual public education spending has long remained below the widely cited benchmark of 4 to 6 percent of GDP and shortfalls in electricity, drinking water and toilets further constrain schools. Better measurement cannot substitute for adequate funding, but it can make the case for investment more exact and more persuasive. Three practical changes would move the debate forward. First, establish a common official baseline. Federal and provincial education authorities should agree on the age band, the definition of enrolment, the treatment of madrassa and non-formal learners and the reference year. The Pakistan Bureau of Statistics could publish a reconciled estimate, with a clear, published bridge from other major surveys. Independent assessments such as ASER should remain visible as checks on the official account.
Second, publish a comparable annual measure of foundational learning by district. Provinces have assessment bodies and should jointly agree on a short, sampled reading and arithmetic assessment for Classes 3 and 5. Results should be disaggregated by gender and location and released on a fixed annual calendar. Reporting both the share attending and the share learning would prevent success on one measure from disguising failure on the other. Third, use the findings to direct support. Districts with weak retention or learning outcomes need diagnostic visits, teaching materials and sustained practical assistance. School councils and district officers should be able to see the results and explain how they respond. Funding formulas may reward improvement, but must also protect districts facing poverty, distance and other barriers beyond a school’s immediate control. Otherwise, a well-intended incentive could punish the children who need help most. Pakistan has made gains in access and public attention to children outside school has grown. The next task is to make estimates comparable and ensure that the children counted as enrolled actually learn. A common baseline would make progress credible; a reliable learning measure would make it meaningful. Together they could turn education plans into a public account of where each child stands and what help is needed next.
—The writer is Commoner from 44th Common Educationist — Founder of WHI Institute.based in Sargodha.
