The cost of a healthy diet and the economics of development

Production is not purchasing power

Food systems transform land, water, energy, capital, labor and technology into agricultural output. This supply can reduce production costs, yet the price paid by a household also depends on storage, roads, transport, trade policy, retail competition and the composition of the diet. Household access depends on wages, cash income, transfers and relative prices. National production and household nutritional security therefore measure different outcomes.

The FAO's State of Food Security and Nutrition in the World 2026 estimates that 645 million people faced hunger in 2025 and approximately 2.7 billion people could not afford a healthy diet. These are not interchangeable indicators: undernourishment, food insecurity and the cost of an adequate diet describe different constraints and may follow different trends. [1]

A causal pathway through human capital

Diet composition can affect health through both inadequate and excessive intake. Deficiency of essential nutrients presents a different clinical and economic problem from excess calorie intake combined with poor diet quality. Income, education, work schedules, food environments and medical access affect both diet and observed disease, requiring care before attributing a disease trend to one food or food-processing category.

A plausible long-run mechanism is household purchasing constraint → reduced dietary diversity → nutritional deficiency or poor diet quality → health and educational outcomes → future productivity and earnings. The size of each relationship is empirical, not automatic. The most informative comparisons distinguish changes in purchasing power from changes in supply and account for sanitation, schooling and health services.

Safety and nutrition are related, but distinct

The WHO's June 2026 food-safety assessment estimates 866 million illnesses and 1.52 million deaths per year due to contaminated food. These estimates measure foodborne disease, not deaths attributable to hunger, and should not be added to the FAO's estimates of food insecurity. Bacterial, viral, parasitic and chemical hazards have different exposure pathways and require different prevention measures. [2]

The economic transmission includes treatment costs, absenteeism, avoidable mortality, disruption of food trade and diminished confidence in supply chains. Enforcement of water-quality standards, inspections, veterinary monitoring and temperature-controlled logistics can thus have benefits that are not captured by farm-gate commodity prices.

Prices, institutions and distribution

Food-price increases are distributionally asymmetric. Low-income households generally devote a larger share of expenditure to essential goods and have fewer liquid resources to buffer shocks. Higher farm prices may benefit some producers while harming net food-buying households, including agricultural workers and small farmers who purchase more food than they sell. Policy evaluation must distinguish these groups rather than treating the national agricultural sector as one agent.

Taxes, subsidies, trade restrictions, strategic stocks, school meals and targeted transfers alter incentives and burdens in different ways. A policy that lowers urban retail prices may also reduce incentives for domestic producers if it is poorly designed. Conversely, productivity gains may fail to reach consumers if logistics, market power or income constraints remain binding.

Analytical comparisons to test

A responsible comparison would stratify households by income, geography and net food-buyer status, measure food prices relative to wages, and use repeated nutrition and health surveys. A temporal association between agricultural policy and life expectancy would not by itself establish causality.

Exposure to shocks and development trajectories

Conflict, extreme weather, energy price movements and transport interruptions can simultaneously raise costs and reduce income. Fiscal responses compete with other expenditure during crises. Import-dependent states may have adequate domestic purchasing power in ordinary periods while remaining exposed to foreign exchange, port capacity and trade restrictions. Exporters face a different risk: international price gains do not ensure affordable domestic diets.

The observation to test over several years is whether increases in real household income, diversification of affordable nutritious foods and safety-system capacity coincide with sustained improvements in health and educational attainment. Interpretation requires regional counterfactuals, exposure controls and transparent uncertainty rather than a single national ranking.

Institutions and evidence

FAO measures food security and diet affordability; WHO assesses health and foodborne hazards; Codex Alimentarius develops international food standards; domestic agricultural, sanitary and statistical institutions enforce and measure different parts of the system. The World Bank, national household budget surveys and labor statistics can help evaluate income and expenditure channels. [1, 3]

The central research implication is institutional rather than purely agricultural: higher production is valuable, but durable nutritional progress depends on whether productive gains become affordable, safe and sufficiently diverse household diets. The strength and distribution of that conversion remain measurable empirical questions.

What the 2026 evidence actually measures

The FAO-led SOFI 2026 assessment separates three statistical objects that often get conflated in public debate. Hunger is estimated from insufficient dietary energy consumption; moderate or severe food insecurity captures compromised access to food; and inability to afford a healthy diet compares a least-cost nutritionally adequate food basket with an affordability threshold. The units and methods differ. For 2025, the report estimates 645 million people experiencing hunger, around 2.1 billion experiencing moderate or severe food insecurity, and 2.69 billion unable to afford a healthy diet. One person may appear in several categories, so adding their totals would count overlapping populations. The release date is 21 July 2026, and the observation year is 2025. [1]

The time series carries a seeming paradox: the average cost of a healthy diet rose from 3.44 PPP dollars per person per day in 2021 to 4.28 in 2025, while the number unable to afford one fell from 2.97 billion to 2.69 billion. These findings are compatible because the affordability calculation also depends on household purchasing power and distribution. The two numbers do not demonstrate that every poor household became better off, or that nominal food prices declined. Purchasing-power-parity dollars are not US retail prices, and cross-year methodological changes or survey uncertainty must be acknowledged before using the results as direct national treatment effects.

The movement in the first series is an improvement in estimated affordability, while the second represents a higher minimum cost. The series have different units and are intentionally not combined into one chart. Exact definitions and aggregation procedures are those published by FAO rather than a locally invented composite index.

The supply chain between the field and the plate

A farm price is only one component of the final consumer price. SOFI 2026 estimates that activities after the farm gate account for roughly 70–75% of consumer food expenditure, with processing, logistics and wholesale contributing up to 40% of total costs. These are global structural estimates, not universal margins for every crop or country. Perishable foods require grading, refrigeration, frequent transport and short retail turnover. The opportunity cost of missing a delivery can be larger than the cost of transporting an equivalent quantity of a shelf-stable staple.

The composition of the cheapest adequate diet matters. Staple cereals supply calories relatively cheaply but cannot, on their own, supply all required food groups and micronutrients. Fruits, vegetables and some animal-source products carry different cost and spoilage structures. If a highway investment reduces grain transport costs but not the cost of refrigerated fresh produce, calorie affordability may improve more than diet diversity. Public investment evaluation should therefore examine food-group-specific retail prices, not only national staple price indexes.

Neither processing nor industrial scale should automatically be classified as damaging: pasteurization, fortification, drying, freezing and hygienic packaging can preserve nutrients, prevent infections or reduce losses. Some formulations, marketing environments and patterns of consumption raise different nutritional questions. The public-health effect depends on the substance, dietary substitution and duration of exposure, not a binary opposition between modern and traditional food.

The distribution of affordability across regions

SOFI 2026 estimates that 66.6% of Africa's population could not afford a healthy diet in 2025; across the world, the estimate was 32.7%. In Latin America and the Caribbean the minimum diet cost was estimated at 4.91 PPP dollars per person per day, compared with the global 4.28. These observations describe separate indicators: one is the proportion of households or persons unable to pay; the other is the estimated price of a nutritionally suitable basket. A region can have higher diet costs but better affordability if incomes are also higher.

Geographic analysis must avoid converting regional means into assumptions about countries or individuals. Within-region variation in rainfall, infrastructure, governance, urbanization and wage levels is large. Local data are necessary to evaluate any proposed intervention, particularly where informal markets and subsistence farming are significant.

Childhood, maternal nutrition and long economic horizons

Early childhood makes the distinction between food quantity and nutritional adequacy especially consequential. SOFI 2026 reports approximately 150 million children under five with stunting and only 30.8% of children aged 6–23 months receiving a minimally diverse diet. Stunting is a population indicator associated with accumulated adverse conditions, not a direct diagnosis of one cause. Poor diet, recurrent infections, maternal health, sanitation and access to health care interact over a child's development. [1]

UNICEF, WHO and World Bank joint estimates provide another measurement system, tracking stunting, wasting, overweight and related indicators across years and regions. They should be used with their age definitions and confidence intervals intact rather than merging incompatible survey waves. [4]

The economic horizon exceeds the electoral cycle. A shock during pregnancy or early childhood may be associated with later educational and health differences, but longitudinal evaluation has to control for family income, neighborhood conditions and access to services. This evidence supports analysis of early-life interventions without justifying a specific fiscal return on investment absent local program data. A school meal policy, maternal health service and clean-water project cannot be assumed to have interchangeable effects.

Food safety as productive infrastructure

The WHO's revised foodborne-disease estimates were released on 4 June 2026 and describe the burden observed in 2021. They cover 42 food hazards and estimate 866 million illnesses, 1.52 million deaths and US$310 billion in combined productivity and medical losses. The different release and reference years matter: calling these figures 'new 2026 cases' would be incorrect. The WHO emphasizes disproportionate effects on children and populations in lower-income countries. [2]

In 2018, the World Bank separately estimated US$110 billion in annual losses due to unsafe food in low- and middle-income countries. The older figure covers a different geography, model and research period; it cannot be treated as a new additive component of the WHO global US$310 billion estimate. Comparison requires harmonizing geographic scope, cost definitions, currency years and method. [5]

The arrows identify possible channels rather than statistically proven causal coefficients. An economic assessment would pair these mechanisms with contamination surveillance, household budget surveys, clinical registries and labor outcomes while considering reverse causation and measurement error.

Pesticides, residues and the difference between hazard and exposure

Debates around pesticides frequently confuse the intrinsic hazard of an active ingredient, the probability of exposure and the magnitude of health risk. Occupational spraying, contaminated water, drift and residues in food are different exposure pathways, with different duration and dose. A substance's ability to cause harm under a particular experimental condition is not sufficient to estimate harm at population-level exposures. Equally, compliance with one limit cannot establish zero risk under all mixtures and chronic conditions.

Scientific and regulatory analysis should consider the compound, route, duration, susceptible population, biomonitoring evidence and uncertainty around mixture effects. The FAO/WHO Joint Meeting on Pesticide Residues supports international evaluations, while the Codex Alimentarius hosts food-safety standards, including pesticide-residue provisions. National authorities remain responsible for enforceable law, monitoring and worker protections. [6, 3]

Economic evaluation also has to address the reasons for chemical use: yield protection, pest pressure, labor cost and the feasibility of alternatives. A restriction may reduce one exposure while shifting costs or encouraging different practices. These possibilities are questions for comparative agricultural, toxicological and labor evidence, not grounds to assume that existing regulation is sufficient or that all pesticide use has the same consequence.

Institutional responsibility through the chain

The policy problem is joint accountability without a single omnipotent regulator. A health ministry may observe malnutrition while a transport authority controls a costly road link and a fiscal authority determines transfer coverage. Fragmentation can allow individually reasonable decisions to produce poor aggregate outcomes. Coordination should be assessed through budgets, implementation and actual measured change, rather than legislation alone.

Policy alternatives and counterfactuals

An unrestricted general price subsidy can deliver immediate relief but may disproportionately subsidize families that could already purchase adequate food, and can become fiscally expensive. Targeted support has different trade-offs: it may reach disadvantaged groups more efficiently yet suffer from exclusion errors, administrative cost and irregular payments. Supply-side investment in rural roads, sanitation, storage and cold chains has a different time profile from cash relief. Public-health regulation is another instrument whose economic return depends on enforcement, compliance and hazard reductions.

To evaluate these policies, compare plausible counterfactuals. Did districts receiving cold-chain infrastructure show a larger decrease in losses and fresh-food prices than otherwise comparable districts? Did a transfer improve real dietary diversity rather than merely nominal spending? Did tighter sanitary enforcement reduce verified outbreaks or mostly increase detection and reporting? More reported cases immediately after enhanced surveillance may signal better observation instead of worsening safety.

A measurement agenda for Brazil and other middle-income economies

Brazil provides a relevant research setting because agricultural export capacity coexists with heterogeneous household incomes, urban food environments and large geographic distances. The analytical challenge is to distinguish production leadership from the distribution of safe, healthy food. Household budget surveys can identify food-group expenditure and income constraints; price series can distinguish staple and fresh-food inflation; sanitary monitoring can examine exposure and contamination; and health records can test disease associations with socioeconomic controls.

Comparisons with import-dependent countries should not assume a common exposure to global grain prices. An exporter may experience currency depreciation, increased farm revenues and more expensive domestic food simultaneously. Consumers and producers occupy different positions in the same shock. Econometric work should specify what is observed, what is a modeled mechanism and what remains a possible confounder.

Relevant public datasets and institutional responsibilities include FAOSTAT [7], World Bank development indicators [8], Brazil's IBGE statistics portal [9], and the WHO Global Health Observatory [10]. These resources offer compatible starting points, but their definitions, reporting delays and sampling frames must be checked before cross-country regressions.

Conditions that would change the assessment

The joint supply-and-income interpretation would weaken if consistently measured nutritional outcomes across different regions responded almost entirely to production volume after controlling for real earnings, food-group prices, sanitation and distribution. The institutional argument would strengthen if comparable harvest gains repeatedly failed to produce better affordability in areas with weaker logistics or lower household incomes. Neither analysis should be made from national aggregates alone.

A second uncertainty is substitution: cheaper calories may coincide with less diverse diets, while higher spending on food may reflect prices rather than better nutrition. A third is timing: educational and chronic-disease outcomes can evolve more slowly than market prices. Finally, surveillance gains may increase measured disease incidence before actual risk declines. The correct unit of analysis changes with each proposed mechanism.

Food policy cannot be assessed only by tons harvested, prices on commodity exchanges or export receipts. Its development effect depends on which people can obtain which foods, at which cost, under which sanitary conditions, and with which long-run consequences.

References

1. FAO et al. (2026). The State of Food Security and Nutrition in the World 2026: official findings. Official source.
2. WHO (2026). Foodborne disease estimates, 2026 edition. Official source.
3. FAO/WHO. Codex Alimentarius: institutional overview. Official source.
4. WHO, UNICEF and World Bank. Joint child malnutrition estimates. Official source.
5. World Bank (2018). The Safe Food Imperative. Official source.
6. WHO. Codex Alimentarius: food standards and scientific advice. Official source.
7. FAO. FAOSTAT: agriculture and food statistical database. Official source.
8. World Bank. World Development Indicators: open data. Official source.
9. IBGE. Brazilian official statistics. Official source.
10. WHO. Global Health Observatory. Official source.