One of the most impactful things you can do to fight climate change is make a few small but meaningful tweaks to your diet. Nearly three-quarters of people in the UK and more than half of Americans think it is important to eat sustainably.
But the information on which foods are really "sustainable", "green" or "eco-friendly" is often confusing – and sometimes such terms are outright misleading. With products from beef to beer now being sold bearing a "carbon neutral" label, how can you make sure what you're eating is genuinely sustainable?
The first big challenge is knowing how to weigh up the many different factors that contribute to food's emissions. For instance, there's excellent evidence plant-based foods require less energy (and produce less emissions) than animal products. But what if those plant-based products have been flown halfway around the world before they get to your table? And how much do different cooking methods cut or increase carbon emissions?
To find out the answers, BBC Future worked with Sarah Bridle, professor of food, climate and society at the University of York in the UK, and freelance sustainability researcher Rebecca Lait to analyse our food emissions in a two-week experiment. We tracked the diets of a vegan, Zaria Gorvett, and a vegetarian, Martha Henriques – counting everything from lovingly home-made meals to furtive cupboard snacks. We also tracked dietary data from a volunteer omnivore, who provided us with a baseline to compare our vegan and vegetarian data.
The hypothesis for our experiment? We supposed that a vegan diet would have the lowest emissions, the omnivore the highest emissions and the vegetarian would come somewhere in the middle. After all, meats such as beef are known to be some of the highest-emission foods you can eat. Meanwhile other animal products such as cheese, butter and eggs also rank highly when it comes to climate impact. Some plant-based foods, on the other hand, can even have a negative climate impact (due to converting land from croplands to carbon-sequestering nut trees) – making them hard to beat.
Such an experiment might seem like a done deal – but our investigation yielded a few surprises. The role of air miles, cooking methods, food waste, and even the impact of eating out versus eating in all played a part. Some of our most revealing findings came from how these factors measured up against one another. In the end, the experiment also underscored some of the best-evidenced ways to reduce the climate impact from food.
I'm in my kitchen, basking in the glow of complacent superiority. I've been asked to take part in a sustainability experiment – and I have a sneaking suspicion that the results are going to make me look really good.
All I have to do is keep track of my normal meals for a week and see how my carbon emissions compare to those of a colleague. Though it's technically not a competition, my opponent – whoops, I mean co-worker – is a vegetarian, while I am a vegan. And though it's strictly just for educational purposes, it seems like the pole position is probably already mine.
Today – the day it begins – I have already clawed out a head start by skipping breakfast, though of course this is essentially cheating. Now it's lunchtime. Swiftly side-stepping the carbon peril of avocado toast, I opt for another toast-based delicacy – pan con tomate. This version is just rye bread slathered with a mixture of chopped tomatoes, fresh garlic, olive oil and salt. Since this is a public-facing meal, I add a handful of parsley for added panache. Not a bad start, at 196g (6.9oz) CO2e (or "CO2 equivalent" – meaning CO2 emissions as well as other greenhouse gases such as methane have been factored into the figure).
Skipping ahead a few days – and a few meals – things are looking promising. So far I've had herby pasta (356g/12.6oz CO2e per serving), mashed potato (589g/20.8oz CO2e), vegan yoghurt (69g/2.4oz CO2e), and several salads. Like I said, this is theoretically not a carbon contest, but if it was, I think I'm probably doing quite well already.
Now I'm having another meal that's so worthy, it seems almost contrived: quinoa and kale burgers (394g/13.9oz CO2e), which turn out to be delicious. But too late, I stumble across a surprising fact: in some cases, the majority of a food's carbon footprint comes from the way it's cooked – rather than what's in it.
Usually, emissions are calculated by looking at the way something is made, stored and transported before it reaches the consumer. However, this doesn't factor in what happens when people get round to eating it.
One 2020 study, led by food sustainability researcher Angelina Frankowska of the University of Manchester, found up to 61% of the total emissions linked to some foods are generated as they're prepared in the home, particularly with vegetables. Even toasted bread is significantly more carbon-intensive than the regular kind – this final act adds 13% to its footprint. For foods that have already been partially pre-cooked in the factory, such as tofu, certain meat substitutes – and presumably, quinoa burgers – finishing the job contributes around 42% of their total emissions.
And this is not all. Certain kinds of cooking are significantly more energy-intensive than others. Turning on an entire oven is an almost-ludicrously inefficient way to heat anything up, since you're not just warming your dinner, but the surrounding air. One step down is cooking on a hob – frying, for example – which involves a more efficient heat transfer. However, first place goes to microwaving, which specifically targets the water molecules inside the food, so you're not wasting it making the microwave itself hot. The latter also has the advantage that it only uses electricity, and if this is from a renewable source that's even better.
"For vegetables, cooking can cause up to 80% of the climate impact of the food – if they are roasted in the oven," says the University of York's Sarah Bridle. "This can be cut right back if they are boiled, steamed or cooked in the microwave oven instead. Pressure cooking and slow cooking are also very energy efficient."
However, it looks like I'm in luck. As sustainability researcher Rebecca Lait adds, the proportion of emissions down to cooking vary wildly depending what you're making – naturally, for lower-carbon ingredients, the same footprint from cooking is going to make up a larger fraction of the total. In other words, what you're eating is still the most important factor, Somehow, I'm still on the right track.
"If you're cooking spaghetti bolognese, we could compare the use of lentils or beef," says Lait. "Looking at the emissions from using 100g (3.5oz) of lentils and cooking them on the hob for 10 minutes, that would cause around 80g (2.8oz) CO2e from the food, and 60g (2.1oz) CO2e from the hob. It might look like the cooking emissions are very significant there. But if you used 100g (3.5oz) beef instead, that would cause around 4,500g (159oz) CO2e, and the hob just 60g (2.1oz) CO2e again," she says.
As a vegan, everything I'm cooking is already so sustainable, I decide to view low-carbon heating methods as something of an added bonus.
Anyway, back to my experiment. By Saturday, my resolution has cracked. For lunch I turn on the oven and have a vegan quiche with some chips (845g/29.8oz CO2e) – but here there's another snag. It turns out the emissions generated once food arrives in your house aren't just down to cooking, but storage: frozen chips have higher emissions than any other processed potato products, and this stage contributes significantly to their total.
From there, it seems like it's all downhill. That evening, I compound my previous transgression with a salad that turns out to have origins so decadent, it could have been ordered by a medieval lord. In addition to some locally grown tomatoes, I realise I have inadvertently used my consumer power to summon half an avocado from Peru, 9,777km (6,075 miles) away, a handful of olives from Greece, 2,969km (1,844 miles) away, and some diced smoked tofu made in an assortment of European countries including Germany, 857km (533 miles) away, Austria, 1,388km (863 miles) and France 840km (522 miles).
However, a quick internet search reassures me that this international concoction might not actually have been as carbon-intensive to transport as you would think.
In fact, the vast majority of foods are actually transported via land and sea, with just 0.16% of food miles coming from air travel on average. It's so expensive, it's usually reserved for ultra-perishable products like blueberries, raspberries and strawberries, and delicate vegetables such as green beans, asparagus and sugar-snap peas.
"Most foods are transported by boat, but sending that same food by air would cause 100 times as much climate change," says Bridle. "It's hard for us to know which foods came by boat and which by air – but a rough rule of thumb is that if it would last a week in the fridge, then it could also last a few weeks in carefully controlled conditions on a boat."
For everything else, a more leisurely journey is more cost-effective – some 60% of foods arrive via sea freight. "Dried pulses, apples, oranges and bananas will have come by boat, if they are from the other side of the world," says Bridle. "In that case, the climate impact of transporting them is likely less than the climate impact of growing the food in the first place."
Even luxury fruits such as avocados rarely arrive by jet. According to the Danish investigative media and research centre Danwatch, most of those that end up in the country will have arrived from Chile via a three-week voyage across the Atlantic in cooling containers, followed by a seven-to-14-day layover in a ripening chamber in the Netherlands, before they're transported to where they'll be sold.
The biggest surprise – and in a way, the most satisfying discovery so far – is that the trendy practice of "eating local" has very little impact. One study found that, in the US, the proportion of a food's emissions linked to transport works out at just 11%. In comparison to the carbon required to make it, even travelling long distances – the average product has covered 6,760km (4200 miles) by the end of its life – makes a relatively small contribution.
As soon as you make animal products a part of your diet, there's no escaping the fact that your carbon emissions are going to rise. As the experiment's resident vegetarian, I assumed I would be somewhere in the middle of the dietary emissions spectrum. But what I really wanted to know was, just how much higher would my emissions be than our vegan baseline? By eating dairy but avoiding meat, was I making a small dent in my dietary emissions or slashing them to near-vegan levels?
With some trepidation, I began taking notes of my daily meals.
The first meal I recorded was one I would struggle to live without: a staple breakfast of toasted pitta bread with a scrape of butter and marmite with boiled eggs. More than 500g (17.6oz) CO2e came from the two eggs. The pitta, butter and marmite combined, however, made up just 30g (1oz) CO2e. Although butter is an animal product too – and has 3.5 times more emissions than plant-based spreads – I was intrigued to see that using just a little meant it contributed fewer emissions than the eggs.
It wasn't long before I was confronted with another animal product. Cheese is not my greatest weakness, but it comes high on the list. For every kilogram of cheese you buy, 23.9kg (0.84oz) CO2e is released. Cheese is more carbon-intensive even than pork, poultry and farmed fish. For a vegetarian, it is likely to be greatest source of emissions that finds its way onto your plate. Intriguingly, hard cheeses have higher emissions than soft cheeses as they require greater quantities of milk to make – and how the dairy in question deals with its waste products can also have a significant effect on emissions.
But even the softest of cheeses have relatively high emissions. A light lunch of rice cakes with cottage cheese and sauerkraut came in at 766g (27oz) CO2e – one of my most carbon-intensive meals of the week.
Soon came a bigger surprise. It had happened one afternoon when the fridge had started to take on a certain odour, and I couldn't put off a clear-out any longer. Rootling around in the vegetable drawer, I found three ancient parsnips and two broccoli that were going brown and giving off a pungent cruciferous vegetable smell. Despite my best efforts not to over-buy fresh produce, there was also an incriminating mouldy orange in the fruit bowl. Regretfully, they all went in the bin.
As sad as it is to throw away food, I'm not alone. The average household in the UK threw away the equivalent of eight meals a week in 2018 – curbing that waste could save as much as £60 ($75) a month on food costs. In the US, the average household throws away the equivalent of $156 (£124) of food each month.
Clearing out my fridge of waste, it turned out, led to more emissions than any single meal I ate that week (bar one – only a large and delicious meal out at a Greek restaurant had greater emissions, at 2,001g/70.6oz CO2e). I was very aware that waste is a major contributor to emissions from food, but it hadn't occurred to me that that afternoon's food waste would amount to double the emissions of the bean and vegetable stir-fry I had that evening (713g/25.2oz CO2e).
Ordinarily, vegetables like broccoli and parsnips aren't high-carbon foods – they take a lot less energy to produce than an animal-based product. Gram-for-gram when eaten, a broccoli releases 3.7g (0.13oz) of CO2e less than an egg.
The problem was, I didn't eat those vegetables. It wasn't so much the carbon expense of growing or transporting the fruit and veg that was the main problem, but the fate it met in my dustbin.
When you throw away food, you might think its carbon emissions would be just the same as if you had eaten it. But in fact, eating it makes all the difference. When food breaks down in a low-oxygen environment like a landfill, much of its carbon turns to methane instead of CO2. Methane is around 80 times more potent a greenhouse gas than CO2 in its first 20 years in the atmosphere (though over time methane breaks down, whereas CO2 is much longer-lived). Methane emissions are responsible for almost a quarter of global warming and an all-out, rapid effort to reduce them could slow the rate of current warming by 30% and avoid a 0.5C temperature rise by the end of the century.
One way to think about it, is that if you eat food rather than throwing it away, its emissions essentially stop there, says Lait. "But if it's chucked away, the total for emissions released doesn't stop until that food has decomposed," she says.
Where I live there is no municipal food waste collection service, so household food waste goes in the ordinary bin destined for landfill. As my discarded parsnips, broccoli and orange were buried beneath other household waste, they entered an environment starved of oxygen and produced large quantities of methane.
Bridle and Lait did the maths to calculate the magnitude of my food waste mistake: one head of broccoli releases approximately 153g (5.4oz) CO2e if it's eaten, versus the equivalent of 2.7 times that much if it goes to landfill (423g/14.9oz of CO2e). If I had acted a few days earlier I could have turned that waste into a meal (roast parsnip and broccoli with an orange dressing, perhaps?).
"There are great alternatives to throwing away food," says Lait. "Being able to compost food is one. I would encourage that."
As luck would have it, soon after my food waste incident an opportunity came up to become a compost volunteer at a local community scheme, giving me access to shared food waste bins at the end of my road. Food waste which is composted releases just 14% the greenhouse gases of food that goes to landfill. For people living in other areas where there is no food waste collection service, it is possible to make your own compost if you have space, or join one of many community schemes, such as MakeSoil, ShareWaste or CompostNow.
As I fork compost to aerate it on a warm spring day, I think of all the greenhouse gases saved, and try not to breathe in any flies.
Another unexpected insight came from a satay paneer dish (1,049g/37oz CO2e). It was going to be a dinner for my partner and I. As I don't particularly enjoy cooking and do it as infrequently as possible, I made double portions that we could eat over the next few days.
Energy from cooking on the stove made up more carbon emissions (45g/1.6oz CO2e) than any of the vegetables in the dish. By making enough for around four people, the emissions from cooking per portion were one-quarter what they would have been if I was making a meal for one. However, the biggest change I could have made to this particular meal would have been to cut out the paneer (795g/28oz CO2e) and replace with a plant-based alternative.
Taking batch cooking one step further, I was interested to see that a meal out in a Greek restaurant was particularly efficient in terms of the energy used in cooking (an estimated 18g/0.6oz CO2e for the whole meal).
It might seem surprising, given the image of a hot, steaming professional kitchen, that this can be quite an efficient way to cook. But, as Bridle points out, if 20 people eat out at a restaurant, that's dozens of ovens, hobs and microwaves that aren't being fired up at the diners' homes instead.
A lot will depend, though, on the restaurant kitchen's sustainability practices, where they source their energy, and the types of food they're cooking. Certain types of restaurant are more likely to engage in sustainable practices than others – one study of 93 US restaurant chains found that sit-down eateries were more likely than fast-food operators to have sustainability initiatives in place (including how they source, produce, market and serve food).
Glancing down my traffic-light colour-coded spreadsheet of emissions meticulously prepared by Bridle and Lait, I puzzled over an entry in prominent red from my last Friday evening of the experiment. It was my highest-emission meal that week and, to my great surprise, it was vegan mushroom pizza.
As resident our resident vegan found, the method of cooking can greatly affect the emissions of a given meal. As my pizza was baked in an oven, the emissions from cooking skewed this vegan pizza to be one of my most carbon-intensive meals of the week.
So, how did we do?
A few weeks after our experiment, BBC Future attends a grand unveiling over Zoom, during which our total emissions during the experiment are revealed.
* Vegan CO2e emissions per week: 9.9kg
* Vegetarian CO2e emissions per week: 16.9kg
* Omnivore CO2e emissions per week: 48.9kg
What do these figures this translate to in everyday terms? US Environmental Protection Agency's equivalencies calculator, provides some useful comparisons:
* Vegan, 9.9kg (21.8lbs): 24.6 miles (39.6km) driven in a petrol-powered car, or 1,204 smartphones charged
* Vegetarian, 16.9kg (37.3lbs): 41.9 miles (67.4km) driven in a petrol-powered car, or 2,056 smartphones charged
* Omnivore, 48.9kg (107.8lbs): 121 miles (194.7km) driven in a petrol-powered car, or 5,948 smartphones charged
As expected, our vegan scored the lowest emissions, coming in at less than two-thirds the emissions of our vegetarian and just one-fifth that of the omnivore's emissions.
If it were a competition, our vegan would certainly be the winner.
But the result is not quite as clear cut as we had imagined. It turns out that on some days, vegetarianism came out as the diet with fewer emissions:
* Vegan day: Toasted crumpets with beans, followed with a lunch of mashed potato and salad, rounded off with yoghurt and most of a bag of fennel taralli snacks: 1.9kg (4.2lbs) CO2e – or 231 smartphones charged
* Vegetarian day: A bowl of porridge with a splash of milk, an apple, a banana, a bowl of carrot and coriander soup with bread and butter, a bar of milk chocolate and a dinner of spinach and ricotta tortellini: 1.7kg (3.7lbs) CO2e – or 207 smartphones charged
It seems that having the lowest carbon footprint as a vegan isn't guaranteed – it depends on what you eat. What is clear from our experiment (and more importantly, from rigorous scientific research) is that on average a plant-based diet has significantly fewer emissions. Eating large amounts of meat, especially beef, is a sure way to increase your emissions many times over.
The tweaks to our diets that would result in the greatest fall in emissions were:
1. Reducing animal products – eating fewer of them, or replacing with a plant-based alternative
2. Focusing on what you eat rather than food miles
3. Cooking efficiently, and saving ovens for special occasions rather than everyday use
4. Batch cooking to prepare food using a fraction of the energy
5. Avoiding food waste, through careful planning and creative cooking
And what did our researchers think of our experiment? "The biggest surprise for me is how many different foods we all eat in a day," says Bridle. "It's complicated for anyone to figure out for themselves what their food climate impact is. But despite this, the usual trends emerged – that the most important factors are usually the quantities of each animal product, and any long cooking times. I was impressed with your honesty about the food waste, and surprised how much the waste added to the total climate impact."
For Lait, it was how quickly decisions about food start to add up. "I've done lots of calculations on individual meals or school menus," says Lait. "But seeing over a couple of weeks what a difference your diet makes – it really reminded me how powerful our food choices can be in affecting climate change. We have the opportunity to make these powerful decisions several times a day."
The world's food system is immensely complex, and emissions come from many different sources. Many of these happen before we even pick food off a shelf: land use, farming, packaging and transport, and pre-retail waste among them. But a few simple rules of thumb can be helpful for finding our way through this maze, to make sure our food choices really do help curb emissions.
And, while our experiment has been illuminating, the scale of the challenge of food emissions is deeply sobering. One 2020 review found that even if we had stopped burning fossil fuels immediately, humanity still wouldn't be able to meet the 1.5C limit for global warming set out by the Paris Agreement. Our emissions from food are so high, they alone could tip us over this threshold. At 2C warming, coral reefs are almost extinct – more than 99% are expected to vanish – small islands and coastal communities will disappear, and the Arctic will have an ice-free summer once every 10 years.
So for now, our vegan will continue avoiding animal products, feeling rather pleased with herself, and our vegetarian will fork her compost with renewed determination.