Are School Feeding Programs in Low-Income Settings Sustainable? Insights on the Costs of School Feeding Compared with Investments in Primary Education

Abstract

BACKGROUND: School feeding programs are ubiquitous. Benchmarking expenditures for school feeding is an important component of program accountability and sustainability. OBJECTIVE: To analyze the costs of school feeding and the cost relative to education expenditure and other measures of economic growth. METHODS: Data on the costs of school feeding were collected from multiple sources, including United Nations databases, gray literature, and published reviews. Relationships between costs of school feeding, costs of education, and GDP per capita were analyzed through standard linear regression. RESULTS: Data on costs of school feeding were obtained for 74 countries, including 12 high-income, 40 middle-income, and 22 in low-income countries. School feeding programs were found to cost US$173 per child per year, ranging from US$54 in low-income countries to US$693 in high-income countries. In high-income countries, school feeding costs per capita were on average equivalent to 11% of the per capita investments in primary education, compared with 19% in middle-income countries and 68% in low-income countries. In middle- and low-income countries, school feeding programs covered on average 18% and 13% of the children enrolled in primary school, respectively. The total budget for school feeding in low-income countries was found to be on average 11% of the estimated total primary school education budget, compared to 4% in middle-income countries. CONCLUSIONS: School feeding investments are targeted in low- and middle-income settings, reaching only a portion of primary schoolchildren, with total costs only a fraction of the overall investment in education. As countries get richer, school feeding costs become a much smaller proportion of education costs. The findings of this study provide an updated framework for benchmarking school feeding programs.

Cite this work

No public copy is stored for this record, so only the metadata is cited.

Stored only when both papers are in the corpus — 29 citations exist upstream.

Papers that cite the same references — the relational fallback while embeddings are still backfilling.