Dealing with Food Waste
This is the third part of a three part series on food loss and waste. See the first part on preharvest losses and the second part on postharvest losses. Food waste, as opposed to loss, is food that reaches the intended consumer but is not eaten. Thus food waste is visible to the average person in a way that loss is not. Today we will review some statistics on food loss, present some solutions, and ponder what this whole discussion is for.
Food Waste Statistics
It is the conventional wisdom that waste is a counterpoint to loss, in that while the greatest losses occur in low-income countries, the greatest waste occurs in high-income countries. The reason is that loss is driven by infrastructural factors, such as a lack of access to irrigation, pesticides, mechanized equipment, and a reliable cold chain. Waste, by contrast, is driven by wealth: since people in wealthier countries spent less on food as a share of wealth, they may feel that they can buy more than they need, or simply to be picky.
The truth of all this is unclear. In 2021, a United Nations Environment Programme report found that household loss rates were 79 kilograms per capita per year in high-income countries, 76 kg/cap/yr in middle-high income countries, and 91 kg/cap/yr in middle-low income countries. The report considers three venues for food waste: household is what a family takes home, and there is also a food service category (e.g. restaurants, hotels) and a retail category. The country groups are delineated by income per capita thresholds.
For high-income countries, the report found that food service waste was 26 kg/cap/yr and retail waste was 13 kg/cap/year. These quantities could not be estimated for other income brackets, and none of these quantities could be estimated for low-income countries. Thus the report shows no obvious relationship between a country’s national wealth and food waste.
The report provides data from the surveys from which these statistics are derived. Individual country results show variation that is not apparent from aggregate groupings, and one cannot help but suspect that the variation is driven at least as much by study methodology as by actual differences.
A 2024 update did not find major differences and suggests that waste in the food service and retail categories is also roughly the same across income brackets. It also estimated 1.05 billion tons of food waste per year, which is 19% of total production. This report puts food waste at 17% of world production, compared to 22% for postharvest loss.
Here is another view of food waste by income level. The numbers appear to be significantly different from the above and show clearly that food waste levels are higher in high-income countries. I do not understand the reason for the discrepancy, though this paper uses an energy balance calculation (total food purchase as estimated from the Food and Agriculture Organization of the United Nations minus energy expenditure), whereas the UNEP measures are more direct. I don’t place great confidence in either.
![](https://substackcdn.com/image/fetch/w_1456,c_limit,f_auto,q_auto:good,fl_progressive:steep/https%3A%2F%2Fsubstack-post-media.s3.amazonaws.com%2Fpublic%2Fimages%2F79d51697-ece5-4211-b4e5-b74f68ac78fe_689x353.jpeg)
Waste Reduction Potential
In 2015, the United Nations adopted the Sustainable Development Goals, a sprawling set of goals and metrics toward, well, sustainability and development. One of them, Target 12.3, aims to reduce food loss and waste by half from 2015 to 2030. We are not on track to achieve the goal.
The waste reduction potential is there, at least on paper. This paper estimates that waste in Spain could be reduced by 76%. This paper claims a 20% waste reduction potential in the United States. The nonprofit ReFED has a database of solutions, which they claim would reduce waste by 27% (however, as some of the solutions relate to recycling, “waste reduction” might not mean exactly what the reader thinks it means).
Now that we have looked at the big picture, let us consider some specifics.
Shelf-Life
Shoppers overestimate how much they’ll need, or they simply forget about something for a while. Sometimes a long while. When I was in grad school, I served as a Graduate Student Representative for a year. One of that role’s duties was to clean out the refrigerator (yeah, it wasn’t as glamorous as it sounded), and my goodness, there were things in there that had been forgotten for so long that I couldn’t even recognize what they once were. Every office with a communal refrigerator battles, usually unsuccessfully, the scourge of people forgetting about their meals, sometimes by requiring that dates be written on stored items, or with empty threats to clean every weekend.
No enhancement on shelf-life would have helped here. But elsewhere it might.
This paper finds that a 40% increase in poultry shelf-life decreases waste by 6-7%, whereas increasing bread’s shelf-life by 20% decreases waste 5-6%. This article claims that shelf-lives could decrease consumer food waste by 50%.
The most effective way to extend the shelf-life of produce is to freeze it. Produce that would have a fresh shelf-life in days might last for up to a year in the freezer. This paper finds that frozen foods induce 47% less waste than fresh foods. A more recent analysis finds a reduction in waste, sometimes more than 90%, for freezing across various foods. A consumer survey from the American Frozen Food Institute, a trade organization, confirms that shoppers waste less with frozen food and also value frozen food for the ability to eat when desired, and not to feel compelled to eat it right away.
The same survey also found that, by modest margins, consumers perceive fresh food as more nutritious than frozen food. In the case of produce, nutrients degrade from the moment it is harvested, though the degradation is slowed dramatically when the produce is frozen. Furthermore, frozen produce is typically picked at peak ripeness, while fresh produce is picked before peak ripeness to prevent it from spoiling by the time it reaches the consumer. Generally speaking, fresh produce is more nutritious if the supply chain is very short, such as harvesting from one’s personal garden. Otherwise, and in particular for what one will find in the grocery store, usually either frozen produce has the nutritional edge, or it makes little difference.
Two weeks ago, I discuss the role of genetic modification for pest control to reduce preharvest losses. Biotechnology can also extend shelf-life and thus reduce food waste. From the abstract of this paper,
Biotechnological interventions also include the use of biocontrol agents and beneficial microorganisms to suppress post-harvest pathogens, thereby reducing spoilage and decay.
The paper goes on to suggest that biopreservatives can serve as an alternative to chemical preservatives. Other papers have found the shelf-life potential of biotechnology.
Overeating
About 800 million people are undernourished in the world today, whereas 2.5 billion people are overweight. Today, obesity is viewed with mostly negative moral judgment, but in many (not all) pre-industrial cultures, obesity was seen as a sign of wealth and social privilege. Overeating is usually not classified as food waste, but it should be.
This would be as good a place as any to comment on the urban legend that upper-class ancient Romans would, at a feast, gorge themselves, and then deliberately vomit so that they could gorge themselves some more. The myth originates in part due to the word “vomitorium”, which, unlike what you might imagine, was merely an entrance or exit to a stadium or theatre. The myth also derives from a play by sloppy 19th and 20th century authors on the trope of the Roman elite as decadent among poverty and growing dysfunction within the Empire.
Anyway, this paper estimates that worldwide metabolic food waste, defined as food consumption in excess of healthy levels, is about 140 million tons worldwide per year. Recall that more traditionally defined food waste is just over 1 billion tons per year.
Obesity is a complicated public health issue and beyond the scope of this post. Proposed solutions include reducing restaurant meal sizes, physical education for children, and sugar taxes. None of these are easy solutions; if they were, the problem would be solved by now.
Junk Food
Related to overeating, junk food, or “discretionary food” as this paper by Michalis Hadjikakou uses as a more appropriate academic term, is food that is processed in a manner so that it does not provide adequate nutrition. The paper estimates that, depending on the metric, 33-39% of the environmental impact of food in Australia is from junk food. Correspondingly, another paper finds that junk food comprises about a third of the food intake in Australia. This paper finds that added sugars comprise ~15% of the caloric intake of the American diet, in contrast to recommendations of less than 10%.
Junk food cannot simply be eliminated without substitution, but among many other health problems, junk food induces overeating. There are several reasons: first, junk food lacks important nutrients, leaving the body to crave more food, and second, junk food tastes good.
There is some evidence that Americans are more health-conscious now than in past decades. Smoking rates in the United States have fallen from a peak of 45% around 1955 to 11.5% in 2021, and there is hope for a similar generational shift in diet.
Other Solutions
There are many other solutions from ReFED and from WRAP, a UK-based environmental NGO. ReFED’s database has 73 solutions on topics including composting, inventory management, package design, and many others. I can’t discuss them all, but I’ll mention a few more.
ReFED claims that 7% of consumer food waste in the United States results from confusing date labels, leading consumers to discard food that is still good. This article explains the difference between “best before”, “use by”, and “expiration date” labels. Now, we obviously don’t want a solution that will make people sick, but in most cases we could probably get rid of the “best before” labels.
ReFED calls for retailers to make food near or just past the sell date available at a discount. I take advantage of those discounts occasionally, if it’s something that I intend to eat that day. This paper goes into the logistics from the retailer’s standpoint. I can imagine that a grocery store, for the sake of its image, would not want food near the sell date to be on prominent display, but it would be good for environmental cred and also a way to sell products that would otherwise be thrown out.
A small amount (350,000 tons in the United States) of food waste might be saved with the First Expired, First Out inventory management system, rather than the more common First In, First Out. FEFO makes intuitive sense, but it does add cost for the warehouse by requiring more than a simple queue to track food items.
Looking over the ReFED solutions, particularly those that show very favorable returns, the obvious question is whether the private sector is already implementing these solutions, and if not, why not? I haven’t looked deeply enough to know, and ReFED doesn’t do a good job of documenting their methodology.
More troubling is what is lacking. ReFED has nothing to say about biotechnology. The reason why lends itself to speculation. Ultimately, I am left the impression that ReFED has made a half-hearted attempt to make their numbers look rigorous, but the potential to fudge them is glaringly obvious, and I place little confidence in the numbers. The contrast between the lack of progress on SDG 12.3 and the apparent ease of addressing food waste that ReFED portrays is telling.
I hate to keep piling on, but another problem with ReFED’s presentation is that I have no sense of what the big picture is. Instead I see a lot of solutions that individually seem piddly and may or may not be workable. If I was running a grocery store and wanted to reduce food waste, I don’t know what I should take away from ReFED. If I was in the state legislature and wanted to draft legislation to address food waste, I don’t know what I should take away from ReFED.
The Moral Dimension
Food waste, especially that which is visible to the average citizen, naturally provokes a negative reaction. In this survey from the American Frozen Food Institute, on various questions related to the perceived moral implications of food waste, 81-88% of respondents were bothered by it. Most of us grew up with some variant of the admonishment to “eat your peas because there are starving children in Africa”.
This article asserts that food waste, which has gone through ups and downs over the generations in American history, dropped during the Great Depression, first because refrigerators became widespread, and second because poverty compelled families to minimize waste. I haven’t found good statistics, but we all know people who lived through the Depression or economic hardships and are fastidious about waste. This should be counterbalanced with an understanding that the Depression was a time of great waste on farms because farmers could not sell their crops at a profit, and the destruction of crops and livestock was abetted by the Federal Government with the Agricultural Adjustment Act.
Of course, we all know that there is not a direct line between cutting food waste and solving world hunger, despite the fact that many of the articles I cited today invoked the subject of world hunger.
Incidentally, I have tried this on my cat. When he gets picky and refuses to eat a certain cat food, I lecture him about all the stray cats in Israel and Turkey who have to hunt for their meals. So far, this strategy has not been effective.
Concluding Thoughts
It is time to bring this series on food loss and waste to a close with some concluding remarks.
I accused ReFED of putting forward a bunch of ideas without painting a big picture, but this series suffers from the same flaw. I talked about managing droughts and pests, farm mechanization, grain storage, the cold chain, shelf-lives, and other solutions, and I threw in a few historical vignettes, but I don’t have a better sense of the takeaway message now than I did three weeks ago. This may be the nature of the beast, that there simply isn’t a silver bullet solution, and that a bunch of silver buckshot solutions is the best we can do.
There is a larger problem, though, and one that I wish I had dealt with at the start. Why is food loss and waste important? The best motivation I can think of is that food waste provokes a visceral reaction, at least for me. I dislike watching something throw food into the garbage. But good policy is more than gut reactions.
Reducing food waste will have significant environmental benefit, rebound effects notwithstanding, but food waste is not the most important environmental angle on food. Intensification; reducing meat consumption, especially beef; and biotechnology would all be far more important environmentally. And despite popular admonishments, it is not at all clear how reducing food waste will help with food security for the most needy. For world hunger, I would look at questions like bad infrastructure in poor countries, corrupt governments, and barriers to access to world markets well before I look at food waste in wealthy countries.
Food waste suffers from Goodhart’s Law, which is that “when a measure becomes a target, it ceases to be a good measure”. In other words, governments and NGOs have become so fixated on food waste metrics that they forget what the measures are for and lose sight of the bigger picture on the environmental impacts of agriculture and on food security. And what greater waste is there than that?
Quick Hits
Last week, I argued that reducing world greenhouse gas emissions by 1.8 billion tons CO2 per year from developing a cold chain for food transport, as this paper claims, is implausible. The basis of my assertion is that world GHG emissions from agriculture is 5.87 tons, and 30% of that figure is far more than overall food loss rates, let alone losses relevant to refrigeration. This analysis does not account for emissions from land use change, which would nearly double agricultural overall emissions rates. I still find the paper’s claim implausible, but less implausible that it appeared before.
Also last week, I discussed farm mechanization in the context of reducing food losses. Farm mechanization is a big topic, and there are many angles to it that I was unable to discuss, including possible rural unemployment, environmental concerns such as cropland expansion and erosion, relative benefits to smallholder vs. larger farmers, financing, and many other issues. This paper discusses these issues in the context in sub-Saharan Africa, and why so many African governments undertook unsuccessful mechanization campaigns in the 1960s and 1970s.
There was a terrorist attack this week in Dagestan, which entailed attacks against a synagogue, a church, and a police station and killed 21 people. Some readers may recall the Chechan Wars, pitting the Russian government against a separatist movement in Chechnya which fought for independence in the 1990s and by the 2000s had devolved into a terrorist insurgency, whose acts included the Beslan school siege. Readers may also remember the recent Crocus City Hall attack, which was perpetrated by ISIS-K and marks a major escalation of that group’s capabilities and threat to the rest of the world. It is not clear if/how ISIS-K is directly involved in this week’s attack, but in any case, there is a major risk of growth of Islamist insurgency in the North Caucusus, especially with the Kremlin distracted by the war in Ukraine.
Another interesting artificial intelligence paper this week introduces DKPROMPT, which combines prompting of a vision-language model with the Planning Domain Definition Language, one of the premier languages for AI planning. Responding to “situations”, e.g. ways in which a robot’s actions are not exactly as intended, DKPROMPT scores significantly better than other systems on the OmniGibson simulator benchmark of everyday tasks. This is another illustration of how contemporary large models can perform well when paired with more classical approaches.
Finally, I would like to comment on last Wednesday’s Gospel reading in the Catholic Church, Matthew 7:15-20.
15 “Watch out for false prophets. They come to you in sheep’s clothing, but inwardly they are ferocious wolves. 16 By their fruit you will recognize them. Do people pick grapes from thornbushes, or figs from thistles? 17 Likewise, every good tree bears good fruit, but a bad tree bears bad fruit. 18 A good tree cannot bear bad fruit, and a bad tree cannot bear good fruit. 19 Every tree that does not bear good fruit is cut down and thrown into the fire. 20 Thus, by their fruit you will recognize them.
This has great relevance to how we think about politics. In undergraduate philosophy class, we learn that we should evaluate a statement strictly on the merits of the statement itself, and not to consider the character of the speaker. In real life, though, nothing makes sense if we do not understand where it comes from. If an idea comes from the intention of sowing social discord, stoking the pride of the listener, or win-at-all-costs partisanship, then one should judge it a bad idea, regardless of the specifics. Too often, we fall into thinking that the issues of the day are of such great importance that we must take moral shortcuts, and this is always bad.