A tonne of rice spilled during storage and a tray of cooked food discarded by a household are both food-system failures, but global statistics do not count them in the same indicator. Understanding the boundary between food loss and food waste is essential before comparing countries, commodities or reduction claims.
Key takeaways
- Food loss generally covers production, harvest, storage, transport, processing and wholesale up to—but excluding—retail.
- Food waste covers retail, food service and households.
- Weight, economic value, edible content and climate impact answer different questions.
- Modelled estimates can show scale, while repeated direct measurement is needed to track progress.
Where the measurement boundary sits
SDG Target 12.3 separates the food chain into two reporting components. The Food and Agriculture Organization of the United Nations (FAO) is responsible for the Food Loss Index, while the UN Environment Programme (UNEP) leads the Food Waste Index.
FAO defines the Food Loss Index boundary as the stages from production and harvest through storage, transport, primary processing and wholesale, stopping before retail. UNEP measures waste at retail, food service and household stages. The official FAO food-loss measurement platform explains the split.
| Stage | Typical classification | Examples |
|---|---|---|
| Harvest and post-harvest | Food loss | Crop damage, incomplete harvesting, handling loss |
| Storage and transport | Food loss | Moisture damage, pests, breakage, cold-chain failure |
| Processing and wholesale | Food loss | Trimming, sorting rejection, spoilage before retail |
| Retail | Food waste | Unsold or damaged products discarded by stores |
| Food service | Food waste | Kitchen preparation waste and uneaten meals |
| Households | Food waste | Expired, spoiled or uneaten food discarded at home |
The boundary prevents double counting, but it can create gaps if surveys use different definitions. Manufacturing is especially complicated when several commodities become one processed product. A study must state which stage it covers rather than relying on “food loss and waste” as a single label.
How the Food Loss Index works
The Food Loss Index is designed to measure change over time. It uses a basket of key commodities and compares percentage losses with a base period. FAO recommends collecting information at critical points such as harvesting, storage, transport, primary processing and wholesale so interventions can target the stage where losses actually occur.
The index is not simply the mass of all lost food added together. Commodity weights and production values matter because losing one tonne of a low-value crop is not equivalent to losing one tonne of a high-value product. The FAO SDG data portal publishes the methodology and progress information.
FAO estimates that 13.3% of food was lost globally after harvest and before retail in 2023, compared with 13.0% in 2015. FAO cautions that the small change sits within expected model variation and reflects updates to national data as well as revised estimates. That is a useful reminder: a modelled global trend is not the same as a directly measured change at a warehouse or farm.
How the Food Waste Index works
UNEP’s Food Waste Index Report 2024 covers households, food service and retail. Its methodology uses three levels.
- Level 1 uses modelling and extrapolation where countries lack direct national measurements. It helps indicate the scale of a problem but is not suitable for tracking small changes over time.
- Level 2 uses national measurement and is the recommended basis for reporting progress.
- Level 3 adds detail such as edible and inedible fractions, destinations and other information needed to design policy.
UNEP excludes methods known to have large biases from its preferred evidence base. For example, household diaries can under-report what people discard. Waste-composition analysis, direct weighing and administrative data may provide stronger evidence, but each method has its own cost and coverage limits.
Why percentages can tell different stories
Suppose a packing facility receives 100 tonnes of vegetables. It ships 92 tonnes, diverts three tonnes to animal feed and sends five tonnes to disposal. A study may report an 8% reduction from the human-food stream, a 5% disposal rate or a different figure after moisture changes are considered. Each number can be correct if the boundary and destination are stated.
Four choices shape the result:
- Denominator: food produced, food entering a stage or food available for consumption.
- Material included: edible food only or both edible and normally inedible parts.
- Destination: landfill, composting, animal feed, anaerobic digestion or sewer.
- Unit: fresh mass, dry mass, economic value, calories or greenhouse-gas impact.
A mass-based figure is useful for logistics. Economic value may better reveal farmer income loss. Calorie measures relate to nutrition, while greenhouse-gas estimates help identify climate priorities. Good reporting can present more than one metric without blending them into one headline.
Why this matters in Southeast Asia
Southeast Asia’s food chains range from small farms and wet markets to export processors and modern supermarkets. Tropical heat and humidity increase the importance of rapid handling, ventilation, drying and cold storage. Islands and long transport routes add further complexity. A regional percentage cannot show whether the largest opportunity is a rice dryer, a fish landing site, a wholesale market or a household refrigerator.
Commodity-specific evidence is therefore more actionable. For rice, water and field management can affect both yield and methane emissions; see our explainer on alternate wetting and drying in rice production. For drought-prone supply chains, storage and procurement decisions must also account for water availability; our drought explainer shows why rainfall alone does not describe supply risk.
How to assess a food-loss or waste claim
Before accepting a headline figure, ask seven questions:
- Which stages of the supply chain are included?
- What year and geography does the estimate represent?
- Was the food measured directly or modelled from other places?
- Does the figure include inedible parts?
- What destinations count as loss or waste?
- What denominator and unit were used?
- Can the same method be repeated to track change?
A credible reduction target should also identify a baseline, measurement frequency and intervention. “Cut waste by 30%” is incomplete if the organisation changes its stores, product mix or counting method along the way.
The practical conclusion
Food loss and food waste describe connected problems at different stages. The split helps governments and businesses match solutions to causes: better harvesting, storage and cold chains before retail; improved ordering, portioning, labelling and household practices after retail.
The most useful number is not always the biggest number. It is the one with a clear boundary, a transparent method and enough detail to show what should change next.
Image note: The featured image is an original AI-generated editorial illustration created for ClimateCrisis.net. It depicts generic produce handling in Southeast Asia and is not a photograph of a specific market or facility.