PAKISTAN’S education emergency is measured through attendance: out-of-school numbers dominate every plan. The Economic Survey 2025-26 reports that the proportion of out-of-school children fell from 38 per cent in 2023 to 28 per cent in 2025, with Balochistan declining from 69 per cent to 45. The gain matters, though the Bureau of Statistics, releasing the same survey in January, recorded the fall as from 30 per cent to 28 — nearer two points than ten. The Annual Status of Education Report 2025 puts the proportion below ten, though its coverage and methodology differ. The instruments do not agree. Enrolment, however counted, is an incomplete measure of inclusion. It records that a child crossed the threshold, not the condition of arrival. A child in Rajanpur receives a milk pack at mid-morning; in most Punjab districts no child does — the register cannot tell them apart.
That condition was last measured in 2018, when the National Nutrition Survey found 40.2 per cent of children under five stunted — later costings use 34 — 17.7 per cent wasted, above the emergency threshold of 15 and 53.7 per cent anaemic. That cohort now fills primary and middle classrooms. Household hunger, by contrast, is measured and worsening: the January survey found nearly a quarter of households in moderate or severe food insecurity, above 30 per cent in Balochistan and rising in Punjab from 14.4 per cent to 22.6. The deprivation is current; the instrument for locating it in a classroom is eight years old.
A child present, enrolled and hungry is counted as an attendance success though unable to learn. Malnutrition is not the sole cause, but it compounds weak teaching, poverty and poor infrastructure: ASER 2025 found that only about half of class five students could read a class two-level text. Exclusion deepens with progression, the Economic Survey recording net enrolment at 54 per cent at primary, 23 at middle and 16 at matriculation and for girls in Balochistan three, losses heaviest where stunting and food insecurity are worst. Three provinces have responded. Punjab’s domestically financed milk programme, launched in September 2024 in three southern districts, now covers thirteen and reaches 1.1 million children at seven billion rupees.
The budget speech of 16 June exposed its insecurity. Having justified it on the malnutrition of nearly half Punjab’s school-age children, the government proposed four billion rupees against seven billion in current cost — a cut of more than two-fifths inside a 750-billion-rupee education budget carrying 27 billion for a laptop scheme and 15 billion for scholarships. No government proposes cutting teachers’ salaries by two-fifths, because salaries are an operating cost of schooling. Meals are classified as welfare and welfare is easier to trim. Sindh’s approach reveals a different vulnerability. In February it launched an 80-million-dollar programme with the US Department of Agriculture: cooked meals and take-home rations for over 200,000 students in nearly 1,300 primary schools. It is among the region’s largest and entirely foreign-financed: a six-year commodity commitment made in Washington can be revised in Washington, against which the province holds no contingency line.
The argument for meals is not sentimental, though the evidence counsels modesty. A Cochrane review of December 2025 found school feeding slightly improves mathematics achievement on high-certainty evidence and raises enrolment by three percentage points, with attendance unmoved on low-certainty evidence. Balochistan, worst on every indicator above, approved a 1.58-billion-rupee Quetta pilot in June, of which 316 million is allocated this year; programme reporting there claims attendance rising from 68 per cent to 77. On cost the figures are firmer. Nutrition International’s Cost of Inaction Tool estimates undernutrition costs Pakistan at least 17 billion dollars a year, 4.6 per cent of gross national income; the World Bank’s 2024 framework puts returns near 23 dollars per dollar. Against such figures, seven billion rupees is not generosity but arithmetic. Pakistan has attempted this before. Tawana Pakistan, feeding girls in over 4,000 rural schools between 2002 and 2005, cut wasting by 45 per cent and raised enrolment by 40 per cent before administrative bottlenecks ended it; of 246 schools later surveyed, 221 reported enrolment falling. It worked; it did not survive. The political intention now exists; the institutional architecture does not. Commitments were made by every province in June 2025. In August last year the federal education ministry, with the planning ministry, BISP and the World Food Programme, pledged a National School Meal Roadmap with legal and budgetary force. A year on, no roadmap has been adopted and neither the national benchmark nor the common monitoring instrument it promised has appeared. The meal available to a hungry child therefore depends less on need than on province of birth.
Four measures would give that commitment effect. The Council of Common Interests, the one forum binding federation and provinces, should adopt the roadmap as a framework prescribing minimum nutritional content, coverage criteria and reporting obligations. The Ministry of Planning’s Nutrition Section should publish provincial coverage and expenditure annually. Provincial finance departments should reclassify meals from development to recurring expenditure; development spending is vulnerable because it can be deferred without amending a budget, which makes a meal collapsible. And a successor to the 2018 survey should be commissioned, since no framework can allocate against deprivation it has declined to measure. Without that architecture, governments will go on photographing chief ministers with milk cartons while surveys rediscover stunted millions. Before the lesson can begin, the child must be physically able to learn.
—The writer is Commoner from 44th Common Educationist — Founder of WHI Institute.based in Sargodha.
