Diet and Global Warming

A paper recently published in Earth Interactions looks at the climate impacts of various North American diets, and in particular the greenhouse gas (GHG) impacts of meat consumption. I saw a poster for this research at AGU in the fall and was somewhat surprised by the findings. After much data-gathering and plenty of assumtions, the authors make the case that the difference in GHG emissions between the average American diet and a vegetarian diet is large — on the order of the difference in emissions between driving an SUV and a sedan.

More generally, animal-product consumption at typical American rates results in GHG emissions of 1-3 tons-CO2-equivalent per year per person. This is in the same range as driving larger vehicles (the consumer activity most commonly associated with climate impact), e.g. 3.6 tons/year difference between a hybrid and an SUV, 1 ton/year between a Camry and a Prius.

So if the whole country went vegan, how much GHG emissions would we avoid? About 6%, it seems. A huge quantity, to be sure, but clearly not the sole solution to climate change. But the point is taken that, on the scale of typical choices that consumers have personal control over, diet is right up there with vehicle choice and home energy efficiency as one of the most important.

Some other interesting tidbits fall out of the analysis. For instance, if one is going to eat a quantity of animal protein, the best type from a climate perspective is poultry. Slightly better than dairy and eggs, and significantly better than fish. Red meat is unsurprisingly worst — by a factor of 2.5 or so (see Figure 3). Still, the usual ethical hierarchy of meat consumption is upset. The emissions for fish are especially surprising since there are none associated with its production. Apparently the high emissions are due to the increasingly large distances that fish are transported.

In any case, I’m sure most of us were generally aware that meat consumption has environmental impacts. We’ve heard figures about how many pounds of grain, how many gallons of water, how many acres of land, are required to produce a pound of beef. Many have made the argument for vegetarianism in terms of resource conservation and availability of food for a growing population. But this is the first analysis I have seen specifically on climate impacts, and it challenges a few of my long-held notions about diet ethics.

As I have written earlier, Vaclav Smil’s book, “Feeding the World” had convinced me that the energy-optimal diet contains some meat, since not all livestock are competing with humans for primary energy. Seafood is the best example of this: there is more food available to a world that eats seafood than one that doesn’t. Hence, perhaps, the prevalence of pesco-vegetarianism. Again from an energy-efficiency perspective, consumption of dairy and eggs makes sense as a source of high-quality protein. But from a climate-impact perspective, things are more complicated. There is a strong case for veganism, if anything, and generally, less meat of any kind is better. And of course, red meat is still the worst.

Smil discussed some evidence that consumption of a small amount of animal protein has had significant health benefits historically (he suggests an optimum of eating meat twice a week or so). The authors of the climate and energy paper devote a section to surveying the health risks of high-protein diets, and assert that vegetarian diets are at least as safe as typical mixed diets. They do not contradict, necessarily, that moderate consumption of animal protein has benefits. Overall, however, I’m less sure of the best diet for good health and a clean conscience since reading this article.

  1. #1 written by Cortney April 30th, 2006 at 12:13


    I have to say, I wonder about the vegan diet — I’ve been told that health wise, it is not as good as a vegetarian diet (I believe in relation to heart health), as it is challenging to get the B vitamins, iron, and sufficient protein. There is of course, the issue of meat substitutes — products that require a fair amount of processing (and packaging) before they get to your local co-op/health store.

    For me, my mental model is to try and eat food (I’m not always successful and often too lazy) that hasn’t not been greatly processed (so beans, even canned beans, win out over tofu) and locally produced when available.

    As for fish, I think it’s a good rule that if you eat it, try and eat what is local to your region (to avoid the shipping issues).

    It’s a tough question, but it’s good to hear that something like diet could actually make a difference. Do you think this article is too controversial for GDRG? I’d certainly like to read it for the group!

    RE Q
  2. #2 written by joshuah May 4th, 2006 at 10:10

    That’s a good point about processed meat substitutes. I haven’t seen any numbers, but I wouldn’t be surprised if a frozen Boca pattie has more impact than an equivalent portion of fresh chicken.

    It may be that restricting diet by type of food isn’t the best approach, and that what we really need is dietary regionalism. Apparently there’s a new book about this.

    RE Q

Comments are closed.