food

Cheese, lentils, and carbon

I’ve been vegan for about 8 years now.  Primarily this was an animal welfare decision, but as I’ve become more climate-change conscious, I’ve also become aware of the fact that vegan foods are lower-carbon than meat or dairy.

More recently, I’ve been considering the issue of local eating and sustainability.  You can’t (sadly) easily grow in the UK the pulses I use for most of my protein (lentils, chickpeas, kidney beans, soya beans (tofu)1).  So that’s all being shipped in from — I don’t even know where, to be honest.  Somewhere Else.  On the other hand, I can get ‘local’ (within 100 miles) cheese or milk down the road at Borough Market.  Would that be better in terms of carbon footprint?

Probably not, it turns out.  The study described there was conducted in the US, but the figures won’t be far off for the UK.  Food miles turned out to be only a small part (around 11%, with only 4% being the producer-retailer leg) of the carbon footprint of any given food.  Most of it was in the production stage, and both red meat and dairy are high-carbon-producers.

The graph they have at that link is irritatingly uninformative, as it doesn’t (seem to) allow for quantities of consumption.  (Broadly speaking, what is interesting isn’t what percentage of food-related greenhouse gas emissions are related to red meat, but how that compares with the percentage of red meat that is eaten with food.  If the Average Diet is 30% red meat and red meat produces 30% of the carbon output, that’s probably fine.)  However, the fact that a 21-23% shift away from red meat towards chicken and fish would cut as much carbon as buying all-local would indicates that the carbon footprint difference between red meat/dairy, and pulses, is genuinely significant.

I then managed to locate a chart showing the carbon cost of various foods.  It doesn’t include pulses but they’ll be somewhere down there with the carrots: very obvious that the carbon footprint is tiny compared to cheese.  

Of course, there’s another factor: if you’re eating for protein, how much protein do you get for your carbon?  Turns out that the protein content of cheese and pulses is close-enough to the same.  Around 100g protein per pound of cheese (exact rate depends on what cheese); 115g/lb lentils, 102g/lb (raw) kidney beans; an impressive 166g/lb for (raw) soya beans2.  So the high carbon cost of cheese isn’t compensated for by higher protein content (although it is higher-calorie).  Milk is low-carbon; but it’s also low-protein (15g/lb or so). 

So I don’t have a good climate-related excuse to start eating cheese again, which is a shame!  The figures might be a bit different if I had my own goat/cow, on otherwise not agriculturally useful land, and was making my own cheese, but unfortunately I don’t think I can fit a ruminant of any sort on the balcony.

Here’s another couple of links for further reading, if you’re interested:

  • The carbon footprint of cheese (theory only, no numbers).  This is less accurate if you’re buying organic artisan cheese from a proper dairy, but there’s still a lot of CO2-emitting there which doesn’t apply to pulses (and it’s accepted that it’s more efficient to put the pulses straight into the humans rather than detouring them via a cow). 
  • An assessment of the carbon cost of a cheeseburger (headline conclusion: the US cheeseburger consumption is responsible for the same sort of quantity of carbon as is the US SUV habit).


1. If you want to try soya beans in the UK, try Elena — the yield isn’t great though for any pulses of this sort
2. 1lb of soya beans would make about 2 medium-sized blocks of tofu.  Not sure exactly the weight of that, but there’s not enough difference to seriously screw up the figures.  Soya beans before being made into tofu are not particularly tasty.

Leave a Reply