Quick-and-dirty calculation of ag footprint

Following a comment discussion on resource footprints over at Greed, Green and Grains, I felt the urge to do a very rough calculation of the relative per capita agricultural footprints of a few countries (U.S., China, India, and Brazil). Here’s what came out:

	U.S.	Brazil	China	India
Caloric consumption (Kcal/capita/day)	3,855	3,118	2,970	2,348
Ag footprint (Kcal/capita/day)	6,024	4,819	3,979	2,597
Relative ag footprint (U.S. to country)	1.0x	1.3x	1.5x	2.3x
Relative oil footprint (U.S. to country)	1x	~5x	~12x	~30x

I’m not surprised by this result – back-of-the-envelope is that U.S. eats ~70% more calories, gets 23% of calories from meat/dairy compared to 6% in India, and uses 3.4 kg of feed per calorie of animal product vs. 1.8 kg since U.S. is more heavily weighted toward meat. Add these factors up and the overall gap not much bigger than double.

I didn’t include waste because while consumer waste is much lower in developing countries, harvest and post-harvest waste is much higher. Even if we do assume the U.S. wastes 40% and India only 10% of food production, the ratio only jumps to 3.3x – still an order of magnitude less than the oil consumption differential of ~30x.

India is about as extreme an example as you can get – not only low average caloric intake (people tend to forget that India has worse malnourishment than Sub-Saharan Africa), but abnormally low meat consumption when adjusted for income level due to cultural/religious factors. So unless my calculations are totally off (and I’ll lay out the details below, so someone can point out if they are), we can say with confidence that rich-poor differential in per capita agricultural footprint is an order of magnitude smaller than the rich-poor differential in energy footprint.

(If you adjust calories consumed for local yields, the gap will be even smaller due to low agricultural productivity in many developing countries… but that's an entirely different topic.)

Calculation details
To simplify I made the assumption that a calorie of grain has a fixed “agricultural footprint.” I used FAOStat data on per capita caloric consumption and FAPRI data on per capita meat and dairy consumption. Conversions for calories/kg and feed conversion are from internet and personal knowledge – please let me know where they’re wrong.

Animal product	Kcal/kg	kg feed/kg product
Beef	1,500	8.0
Pork	2,800	4.0
Mutton/Goat meat	2,070	4.0
Chicken	1,440	2.0
Milk	600	1.5
Cheese	3,300	1.5
Butter	7,720	1.5

I’m missing fish and eggs. There are probably other errors as well – tell me!

Oil footprint, by the way, is based on barrels of oil consumed per capita per day.

Update: How convenient - the FAO's 2009 State of Food and Agriculture report (entitled "Livestock in the balance") has a chart showing calories of livestock products consumed by region. My estimates were within 100 calories for each, which gives me incremental confidence in my answer above.