School Feeding Programs
School feeding reaches 418 million children worldwide — making it the largest social safety net by population coverage. India's mid-day meal scheme alone feeds 120 million children daily. Brazil's PNAE, the US National School Lunch Program, Korea's universal free meals — each reaches tens of millions. Programs combine nutritional, educational, and poverty objectives.
Key insights
India's mid-day meal is the world's largest program
India's PM POSHAN (formerly MDM) feeds ~120 million children daily across 1.3 million primary schools. Began 1995, scaled massively after 2001 Supreme Court directive. Operates with $5B+ annual budget. Effects: increased school enrollment by ~25%, improved nutritional outcomes, increased women's labor-force participation (school feeding reduces child-care demands). Quality remains uneven across states.
Brazil's PNAE is a model
Brazil's National School Feeding Programme (PNAE) feeds ~40 million students. Requires 30%+ of food purchased from family farms — connecting nutrition to local agricultural development. Began 1955, restructured and scaled under successive governments. The Brazilian model has been replicated across Latin America and parts of Africa.
Educational effects are real but modest
Meta-analyses: school feeding programs increase enrollment by 5-15pp, attendance by 5-15%, and test scores by ~0.1 standard deviations on average. Effects largest among the poorest. Nutritional effects: improvements in micronutrients, modest weight gain. Health effects: small reductions in absenteeism due to illness. The biggest effect is often participation in school itself; learning effects depend on instructional quality.
Children receiving school meals — top countries (2024)
Millions of beneficiaries
Key Finding: India, Brazil, US and China each feed tens of millions of children daily.
School feeding coverage by income level
% of primary-age children receiving school meals
Key Finding: High-income countries reach near-universal coverage; low-income countries reach roughly 20-30%.
Methodology & caveats
Program types
Universal (all students, e.g., Finland, Sweden, Korea), means-tested (US National School Lunch — free for poor, paid for others), targeted (UK Free School Meals — for low-income families), conditional (some Latin American programs requiring attendance). Universal programs have higher coverage and lower stigma but are more expensive. Means-tested are more targeted but have administrative costs and stigma.
Cost variation
Per-child annual costs vary from ~$20-50 in low-income country basic programs to $500+ in OECD universal programs. Cost drivers: food quality, kitchen infrastructure, staffing, local procurement requirements. The cost-effectiveness varies widely — most programs deliver more nutrition value than they cost in administrative overhead.
Why school feeding is politically robust
Programs serve multiple constituencies: children (direct beneficiaries), parents (reduced food costs), teachers (improved class behavior), farmers (in programs with local procurement), nutritionists, education advocates. The multi-stakeholder support has made school feeding programs unusually politically durable — they survive government changes and budget pressures that hit other programs harder.