Skip to main content
    School meals

    School meals: A meal that keeps a child in school.

    A daily meal at school brings more children into class, and it reaches the fewest children where hunger is worst.

    Try it in the sliderTalk to Ryan

    Ryan K. Hertel · heyitskenton

    The problem

    What we're up against.

    • Fewest meals where need is greatest.

      School meals reach as few as 27% of primary pupils in low-income countries, against 80% in high-income countries.

    • Millions never reached.

      Before the pandemic, school meals didn't reach 73 million children in need, and the world set out to reach them.

    • Meals are being cut.

      In Guinea-Bissau, the number of schoolchildren getting meals has fallen from 283,400 to around 152,000 amid aid disruptions.

    What works

    Proven, and ready to grow.

    • More children come to school.

      School meals raise enrollment slightly, by about 3.44%, the newest Cochrane review finds, with high certainty.

    • Learning, especially for girls.

      In a randomized trial of Ghana's program, scores rose by about 0.15 standard deviations, with larger gains for girls and the poorest children.

    • The world's largest program.

      In India, children with up to 5 years of free school lunches improved their test scores by about 10% to 20%.

    • More girls enroll.

      In a randomized trial in Burkina Faso, new enrollment of girls rose by 5% in villages that got school meals.

    • Children grow.

      Across trials, children who got school meals grew about 0.32 cm taller than those who didn't.

    What a family could fund

    Where the money goes furthest.

    • A benchmark price.

      About $64 a year gives a child decent school meals in the world's poorest countries.

    • What reaching everyone would take.

      Around $2.7 billion a year for five years would reach 162 million more children by 2030, by the World Food Program's estimate.

    • Groups already doing it.

      Mary's Meals says it feeds more than 3 million children every school day, at £19.15 per child for a school year. In Kenya, one program serves school meals at about $0.28 a plate.

    • Governments take over.

      More than 50 countries that once had help have fully taken over their own programs.

    Honest limits

    What money can't promise.

    • Learning gains are narrow.

      The same review finds school meals may have little or no effect on reading scores or attendance.

    • A costly way to raise test scores.

      An independent panel calls school meals effective but a relatively expensive way to improve learning, and meals of low nutritional value showed no effect on scores.

    • Gains for farmers are unproven.

      A randomized trial in Ghana found no strong evidence that buying school food locally raised small farmers' incomes.

    • The poorest countries depend on help.

      Low-income countries' own governments pay for only about 34% of their school meals.

    Hard questions

    Asked and answered.

    1. You might ask

      Does a meal really help a child learn?

      Ryan

      A little. The newest Cochrane review finds a slight gain in math and little or no effect on reading. The gains are larger for girls and the poorest children.

    2. You might ask

      Doesn't food aid undercut local farmers?

      Ryan

      Programs increasingly buy close to home: in Brazil, at least 30% of the food bought with federal school meal funds comes from family farmers, by law. Whether farmers gain is still unproven either way.

    3. You might ask

      Isn't this the government's job?

      Ryan

      Almost everywhere, governments already pay most of it; only in low-income countries do international donors still carry the main share. More than 50 countries have fully taken over their programs, and a family's initiative can bridge the handover.

    In the slider

    See what your fortune could do here.

    School meals

    The slider prices this cause with these numbers:

    • About $64 a year gives a child decent school meals in the world's poorest countries.
    • The whole need here is the 73 million children school meals didn't reach before the pandemic.
    • The World Food Program calls that price a tentative benchmark.
    Try it in the sliderTalk to Ryan

    More of the Blueprint

    Keep reading.

    23 more chapters on the way

    Back to Keep 100

    Sources

    Every number in this chapter, and where it came from.

    1. Source 1: School feeding programs for improving the physical and psychological health of school children experiencing socioeconomic disadvantage (Kristjansson and others) (opens in a new tab)

      Cochrane, 11 December 2025. Checked 16 October 2026.

      A systematic review of 40 studies. It was partly funded by the World Food Program's school meals service and Dubai Cares.

    2. Source 2: State of School Feeding Worldwide 2024 (opens in a new tab)

      World Food Programme, 2025. Checked 15 October 2026.

      A tentative benchmark for decent-quality school meals in low and lower-middle-income countries, in 2023 dollars.

    3. Source 3: UN agencies back bold plan to ensure every child in need gets a regular, healthy meal in school by 2030 (opens in a new tab)

      World Health Organization, with FAO, UNESCO, UNICEF and WFP, 16 November 2021. Checked 15 October 2026.

      Children school meals didn't reach before the pandemic: the latest count.

    4. Source 4: WFP warns of rising food needs in Guinea-Bissau amid aid disruptions (opens in a new tab)

      World Food Programme, 1 June 2026. Checked 16 October 2026.

    5. Source 5: Food for Thought? Experimental Evidence on the Learning Impacts of a Large-Scale School Feeding Program (Aurino, Gelli, Adamba, Osei-Akoto and Alderman) (opens in a new tab)

      Journal of Human Resources, January 2023. Checked 16 October 2026.

      A randomized trial of the Ghana School Feeding Program, which serves a free daily meal in government primary schools.

    6. Source 6: School Feeding and Learning Achievement: Evidence from India's Midday Meal Program (Chakraborty and Jayaraman) (opens in a new tab)

      IZA Institute of Labor Economics, July 2016. Checked 16 October 2026.

      A discussion paper, later published in the Journal of Development Economics. Uses the staggered rollout across states.

    7. Source 7: Educational and Health Impacts of Two School Feeding Schemes (Kazianga, de Walque and Alderman) (opens in a new tab)

      World Bank, 14 August 2013. Checked 16 October 2026.

      A summary of a randomized trial in northern Burkina Faso.

    8. Source 8: Impacts of school feeding on educational and health outcomes of school-age children and adolescents in low- and middle-income countries: a systematic review and meta-analysis (Wang and others) (opens in a new tab)

      Journal of Global Health, 4 September 2021. Checked 16 October 2026.

      A meta-analysis of 57 studies.

    9. Source 9: Mary's Meals (opens in a new tab)

      Mary's Meals. Checked 16 October 2026.

      Mary's Meals' own figures.

    10. Source 10: Food4Education (opens in a new tab)

      The Life You Can Save. Checked 16 October 2026.

      Figures reported by Food4Education. Parents and the county pay part of the cost.

    11. Source 11: School meals (opens in a new tab)

      World Food Programme. Checked 16 October 2026.

      WFP's own figure.

    12. Source 12: School feeding programs lead to modest but meaningful results (opens in a new tab)

      Cochrane, 11 December 2025. Checked 16 October 2026.

    13. Source 13: 2023 Cost-Effective Approaches to Improve Global Learning (opens in a new tab)

      Global Education Evidence Advisory Panel, May 2023. Checked 16 October 2026.

      An independent panel convened by the World Bank, UNICEF and others.

    14. Source 14: School Meals as a Market for Smallholder Agriculture: Experimental Evidence from Ghana (Gelli and others) (opens in a new tab)

      International Food Policy Research Institute, October 2021. Checked 16 October 2026.

      A three-year randomized trial in 58 districts; a discussion paper, not peer reviewed.