Food loss versus food waste: why the difference matters when you measure it

Two numbers about uneaten food circulate in Canada, and they rarely sit easily together. One says that more than half of the food produced in the country is lost or wasted. The other, from the United Nations, says that a fifth of the food available to consumers worldwide ends up as waste. Both come from serious work. They differ because they measure different things, starting with the words themselves.

"Food loss" and "food waste" are often used as if they meant the same thing. In the international framework they do not. The split decides which stages of the chain are counted, which parts of the food are counted, and where the food has to end up before it counts at all. Anyone who reads a food waste headline, sets a reduction target or runs a packing shed that reports its numbers is working inside these definitions, whether they know it or not.

Two words, two halves of the chain

The Food and Agriculture Organization of the United Nations set out the working definitions in its 2019 report on the state of food and agriculture, which was devoted to the subject.

• Food loss is the decrease in the quantity or quality of food that results from decisions and actions by food suppliers in the chain, not counting retailers, food service providers and consumers. In practice that means losses after harvest or slaughter, in storage, transport, packing and processing.
• Food waste is the same decrease when it results from decisions and actions by retailers, food service providers and consumers.

The dividing line is the shop door. Everything before retail is loss. Everything at retail and after is waste. The distinction is about where food leaves the chain, not about whose fault it is. A grocer who rejects a load for cosmetic reasons causes a loss upstream, at the packer's dock, while a load that spoils in the store's own backroom is waste.

The same 2019 report put the first global figure on loss under this framework: around 14 per cent of the world's food, measured by value, was lost after harvest up to but not including retail in 2016. FAO's later estimates have stayed in the range of 13 to 14 per cent.

The target that made the split official

The definitions matter because a global target was written around them. When United Nations members adopted the Sustainable Development Goals in 2015, target 12.3 read, in short: by 2030, halve per capita global food waste at the retail and consumer levels and reduce food losses along production and supply chains, including post-harvest losses.

Notice the two halves. Waste gets a numerical goal, cut in half per person. Loss gets a direction, reduce, with no number. The reason is partly practical. Waste at retail and in homes can be expressed per person and compared across countries. Loss depends heavily on what a country grows, how it stores it and how far it moves it, so a single percentage target would mean very different things in different places.

To track the target, the indicator was split in two, with a different UN agency in charge of each:

Losses before or during harvest, such as a field left unpicked, sit outside the Food Loss Index's core scope. Countries can include harvest losses in their own reporting, but the global index starts after the crop has been gathered.

What the UNEP reports have found

UNEP has published two editions of its Food Waste Index Report. The first, in March 2021, estimated that about 931 million tonnes of food were wasted in 2019, roughly 17 per cent of the food available to consumers.

The second, published in March 2024, estimated that 1.05 billion tonnes were wasted in 2022, close to one fifth of food available to consumers. That works out to about 132 kg per person per year, of which households accounted for 79 kg. By sector, households produced about 631 million tonnes, food service about 290 million and retail about 131 million.

It is tempting to read the move from 931 million to 1.05 billion tonnes as a sign that waste rose. The report itself cautions against that. More countries supplied measured data the second time, and the methods improved, so the two editions are better read as two snapshots taken with different cameras than as a trend line.

Two features of UNEP's approach are easy to miss. First, the index counts inedible parts that come with food, such as bones, peels, shells and stones. Second, it counts food by where it ends up. Food that goes to landfill, compost, anaerobic digestion, the sewer or controlled combustion counts as waste. Food that is redistributed to people, or diverted to animal feed or bio-based materials, does not.

The Canadian figure everyone quotes

The most widely cited Canadian estimate comes from a study published in January 2019 by Second Harvest, a food rescue charity, with the consultancy Value Chain Management International. It estimated that 58 per cent of the food produced in Canada is lost or wasted across the chain. It also estimated that about 32 per cent of the food produced, some 11.2 million tonnes, was avoidable loss and waste, meaning edible food that could have been eaten or rescued. The value of that avoidable portion was put at about $49.5 billion a year.

The gap between 58 and 32 per cent is the measurement question in miniature. The larger figure includes unavoidable material: peels, bones, trim and the other parts nobody expects to eat. The smaller one counts only food that could have fed someone. Neither is wrong. They answer different questions, and the headline usually uses the first while the policy discussion is mostly about the second.

The Canadian study also counted the whole chain, from farm through processing, distribution, retail, food service and households. So it bundles loss and waste together, while the UN indices keep them apart. Comparing "58 per cent in Canada" with "19 per cent worldwide" is comparing a whole-chain measure, with inedible parts, against a retail-and-after measure. The numbers are not on the same scale.

Following one crate through the accounts

An illustrative example shows how the definitions change the answer. The figures below are round, invented numbers for teaching, not data from any real farm or store.

Say a field crew picks a crate of 20 kg of field tomatoes in late August. Here is where the tomatoes go.

1. At the farm packing line, 2 kg are culled. Of these, 1 kg is split and soft and goes to compost, and 1 kg is sound but too large for the grade and goes to a neighbour's hogs.
2. On the truck to the distribution centre, a refrigeration fault softens 1 kg, which is sent to an anaerobic digester.
3. At the store, 2 kg pass their best quality and are pulled from the display. Half goes to a food bank, half to the store's organics bin.
4. A household buys the remaining 15 kg over a few weeks. It eats 12 kg, trims 0.5 kg of cores and stems, and throws out 2.5 kg that went soft in the fridge.

Now count it four ways.

The same crate produces totals from 3.5 kg to 8 kg depending on the rules, and the home's share moves from under two fifths to more than two thirds. Nothing about the tomatoes changed. Only the accounting did.

Three choices drive most of the difference: whether inedible parts count, whether food sent to animal feed or donation counts as lost, and whether the stages before retail are included. Value is a fourth. If each kilogram is valued at the price at that stage, the 2.5 kg thrown out at home weighs far more in dollars than the kilogram culled at the packing line, because the household paid for the washing, packing, trucking and retail margin built into it.

Why the distinction is useful, not pedantic

Splitting loss from waste does more than tidy up statistics. It points to different fixes.

Loss is mostly an engineering and contracting problem: cold storage, better handling, faster transport, grading standards and buyer specifications. The people who can act are growers, packers, carriers and processors, and the fixes often pay for themselves, a theme taken up in a page on cutting food waste as a business decision.

Waste is more about behaviour and retail practice: portion sizes, date labels, promotions, menu design and how households plan meals. The policy tools there are campaigns, labelling rules and support for redistribution.

Separate measures also let a country see whether progress at one stage is pushing the problem to another. A retailer that tightens its specifications to cut in-store waste may simply move the culls back to the packer. Under a single combined number, that shift would be invisible. Under the split, waste falls and loss rises, and someone can ask why. A stage-by-stage look at where food is wasted along Canada's value chain shows how many hands that crate passes through.

Common questions

Does food given to a food bank count as waste?

Under the UNEP framework, no. Food redistributed to people stays in the human food supply. Some business reports count donations as diverted waste, which is why company figures and national statistics may not match.

Is food fed to livestock lost?

It depends on the framework. The UN indices do not count food sent to animal feed as waste. A study focused strictly on human food could count it as lost, because something people could have eaten went elsewhere. The hogs in the example above are fed either way.

Why do Canadian estimates look so much higher than global ones?

Mostly because of scope. The 2019 Canadian study counted the whole chain and included inedible parts in its headline figure. The UNEP index counts only retail, food service and households. The two numbers measure different slices.

Which figure should a business use?

The one that matches the decision. A packer weighing a new cooler needs loss in its own shed, by weight and value. A grocer setting a waste target should say which destinations it counts, so that sending more to animal feed is not mistaken for buying better. Writing the definition next to the number is the cheapest improvement any report can make.