I was struck by this argument, which I had never heard:
The overriding advantage of meat is that demand for it is elastic. People don't need it but they like it, and up to a point, however much you produce, they'll keep on buying it. The demand for cereals for human consumption, on the other hand, tends to be inelastic. People need their pound of grain a day, but they don't need much more, and they won't buy any more unless they have sufficient wealth to invest the grain in animals, either to produce higher value food, or else to keep it "on the hoof" for a rainy day (or a drought).
The existence of meat means that a farmer can sow wheat, barley, oats, beans, maize, and so on with reasonable confidence that, in the event of a good harvest, someone will buy it, because even if everybody has sufficient, it can be fed to animals. This dynamic is not restricted to a money economy. It works exactly the same for Melanesian subsistence farmers who can sow enough sweet potato and manioc to cover a bad year knowing that it is not a waste of effort, because in a good year the surplus can be fed to pigs.
Take the animals, the elastic element, out of the equation and the business of sowing grain suddenly become far more risky…This elementary matter of the need for a feed buffer fails to feature in most of the literature that is written about meat-eating and vegetarianism…
That quotation is from Simon Fairlie's quite interesting Meat: A Benign Extravagance.
As they used to say on the U. Chicago Ph.d. qualifying exams, true, false, or uncertain? And under what conditions?