A food-waste number is only as useful as its boundary

Two households put different weights of material in their food-waste bins. One has cooked several meals from fresh ingredients and discarded peelings and bones. The other has thrown away a smaller quantity of unopened food. Which household has wasted more?

The scale can answer a question about mass. It cannot settle what was preventable, why it happened or which intervention would help. A useful food-waste measurement begins by deciding which of those questions it is meant to answer.

This is not an argument for discounting the problem. UNEP’s Food Waste Index 2024 puts retail, food-service and household waste on a global evidence base. Its widely reported estimate of 1.05 billion tonnes refers to 2022 and includes associated inedible parts. Removing that boundary from the headline turns a defined estimate into a different, unsupported claim about edible food alone.

UNEP’s summary of the same 2022 estimate assigns 60% to households, 28% to food services and 12% to retail. A household bin is therefore an important part of the account, rather than the whole explanation. Kitchen habits, restaurant portions and retail decisions enter the problem at different points.

Open the imaginary inventory

Consider a fictional week’s discarded material: carrot peelings, eggshells, part of a cooked meal, a spoiled vegetable and a packaged product that was never opened. This is an analytical example, not a sample collected by this publication.

Each item has a different history. A shell is associated with food preparation. A cooked remainder might reflect portions, preferences or a changed plan. A spoiled vegetable might have been forgotten, damaged before purchase or bought for a meal that never happened. The unopened product raises questions about timing, storage and the way it was sold.

Some boundaries are also cultural and contextual. A part used in one cuisine may be discarded in another. A measurement needs an explicit convention for “inedible,” rather than treating an observer’s personal cooking habits as a universal definition. That does not mean every discarded part should be eaten; safety, suitability and actual food practices matter.

The Food Loss and Waste Protocol’s definitions make room for inventories to identify food, associated inedible parts and destinations separately. Its standard does not impose one universal definition for every purpose. It supplies common terms so the chosen scope can be stated and understood.

This is useful because an inventory can be honest and still fail to answer the question being asked. A total intended to plan waste collection may reasonably include all relevant material entering that service. A prevention campaign needs to distinguish what could have remained food for people. A comparison of climate impacts needs more detail again.

The first improvement is therefore a better label. “Food waste collected” describes a service flow. “Edible food discarded” describes a narrower material category. “Avoidable waste prevented” makes an additional claim about what changed. The phrases should not trade places merely because the final number is convenient.

The bin does not contain the entire household

A collection sample can reveal what entered that bin. It may not include material put down a drain, composted at home or discarded through another route. A study needs to establish which pathways are covered before comparing households or years.

The boundary can also move when a new service begins. If a city introduces separate food-waste collection, more material may become visible in that stream without a corresponding increase in total generation. Calling the larger collection total proof of worsening behavior would confuse a changed observation system with a changed household.

The reverse mistake is possible too. A reduction in one bin could reflect diversion elsewhere rather than prevention. Composting can be a useful destination for appropriate material, but it is not the same outcome as avoiding the unnecessary purchase and disposal of edible food. The material’s destination and the reason it became waste are separate questions.

The FLW Protocol’s measurement resources distinguish direct weighing, composition analysis, diaries, surveys and other methods. They also make clear that packaging is outside the material definition. A full container weighed without subtracting its packaging is not an equivalent food-waste quantity.

Method choice affects what can be learned. A physical sort can show material categories in the sampled waste. A diary can add context about why something was discarded, but it depends on what participants record. A survey can investigate attitudes and reported routines. None should be quietly relabeled as a direct measurement of all food leaving every home.

It is often useful to combine evidence. An observed pattern can suggest where to ask questions, while reported reasons can help interpret the material found. The methods should retain their names so that uncertainty remains visible. More evidence is valuable when it answers a gap, not simply when another percentage can be added to a presentation.

The shopper is answering a different questionnaire

WRAP’s Household Food Management Survey 2025, published in March 2026, used 5,509 online interviews with UK adults responsible for shopping or preparation. Fieldwork took place from late June to mid-July 2025. The study examines knowledge, attitudes and behavior; the interview total is not a count of weighed household bins.

That distinction matters when self-reported improvement meets a collection record that changes less. People can become more attentive without estimating their waste accurately. A different sample or period can also affect the comparison. Apparent disagreement deserves investigation before either source is dismissed.

Households also operate under constraints. Work changes, care responsibilities, illness, unpredictable attendance and limited storage can disrupt a meal plan. A prevention message that assumes perfect control over the coming week will fit some lives better than others. Understanding the cause of waste should precede choosing the instruction.

The product and shop participate in the decision. Pack size, promotions, available portions and useful storage information can influence whether a purchase fits the household. The consumer makes choices within that offer, rather than inventing every quantity independently. A program focused exclusively on personal discipline can miss changes businesses are able to make.

Price pressure complicates the picture further. Buying a larger unit can appear economical while creating more food than the household will use. Buying smaller quantities can reduce that risk while costing more per unit. There is no universal answer hidden in the phrase “shop smarter”; the actual options and household routine determine the trade-off.

This is why the imagined unopened product should provoke a question rather than a verdict. Was it an unnecessary impulse purchase, an unsuitable promotion, a plan interrupted or something else? The same item in the bin can come from several mechanisms. Different mechanisms call for different responses.

A date is information, not a dare

Food-waste campaigns can become unsafe when they encourage people to disregard dates indiscriminately. In the UK context, the Food Standards Agency distinguishes use-by dates, concerning safety, from best-before dates, concerning quality. The distinction should not be generalized to differently worded labels in every country, nor replaced by an informal smell test for a use-by product.

The useful prevention opportunity lies before disposal becomes the only appropriate option: clear information, suitable purchasing and storage, and planning within the product’s instructions. A campaign should reduce unnecessary waste without making safety the household’s sacrifice for a better metric.

Label knowledge is also only one part of the problem. A person can understand the date perfectly and still lack the time, facilities or reliable schedule to use every item as intended. A retailer can print a clear label while selling a quantity unsuited to a small household. Information is necessary in many cases, but it is not a substitute for examining the rest of the system.

For evaluation, this suggests measuring the intended behavior carefully. A campaign about date-label understanding should establish whether comprehension improved and whether the relevant waste changed, without assuming that a higher knowledge score proves a particular reduction. The link is a research question, not an automatic conversion.

A kilogram has a composition

Mass is useful because it can be measured consistently. It does not make every kilogram nutritionally, economically or environmentally equivalent. Different foods embody different production and processing histories, while their water content can strongly affect weight.

This creates a choice for any program claiming a benefit. A mass-reduction target should report mass. A monetary estimate needs appropriate prices and a stated valuation method. An emissions estimate needs the food composition and relevant life-cycle assumptions. The largest weight category is not automatically the largest source of every other impact.

An aggregate can remain useful without carrying all these meanings. The problem arises when a report takes a reduction in tonnes and implies an unmeasured improvement in nutrition, household finances and emissions at the same rate. Each outcome deserves its own evidence or a clearly identified estimate.

The system boundary extends upstream as well. FAO’s food-loss measurement explanation distinguishes losses from harvest up to, but excluding, retail from the retail-and-consumption focus of UNEP’s Food Waste Index. These are related parts of a food-system goal, not interchangeable measures that can be added casually without checking coverage.

A household prevention measure could have effects elsewhere in the chain, but those effects should be investigated rather than assumed. Similarly, reducing a farm or storage loss is valuable without being counted as a measured change in household behavior. The point of the boundary is to locate responsibility and progress accurately, not to divide the system into unrelated moral territories.

The denominator can change the apparent comparison too. Imagine one person discarding two kilograms and a four-person household discarding four. The larger household has the greater household total but the lower per-person total in this invented example. Neither calculation describes the edible share or explains the cause. A comparison needs the denominator suited to its purpose, with household composition visible where it affects interpretation.

The sampling period matters in the same way. An ordinary week and a celebration week may contain different food and routines. Extrapolating either into a year needs justification. A short diary can be useful for learning without being presented as a precise annual estimate for every household like it.

Write the useful conclusion

A useful local study could state the sampled period, households, material categories and disposal routes. It could distinguish weighed material from self-reported behavior and describe what falls outside its scope. Its conclusions would connect the observed problem with changes that a household, retailer or public service can realistically make.

It should also preserve comparison over time. If a campaign changes the definition, sampling or covered routes, a new total needs an explanation. A smaller number produced by a narrower boundary is not the same as a smaller quantity of wasted food. Good measurement prevents a public promise from being met only on paper.

Return to the two bins with a better inventory: what was edible, what was discarded and what household routine put it there? One family may need a way to use a surplus; another may simply have the unavoidable remains of meals cooked from scratch. Weighing the material helps distinguish those situations when its composition stays visible. The useful result is the change that prevents edible food being lost next week, rather than a league table that confuses the two households.