Commodity share of retail food prices

Michael Roberts lays out some back-of-the-envelope math to back up the common (and true) assertion that “ Raw commodities make up a tiny share of retail food prices” in developed countries:

The numbers in the quote came from my own back-of-the-envelope calculations. There are about 6 lbs. of corn used to produce each lb. of beef. Corn sells for about $3.70/bushel. A bushel is 56 lbs. So I figured the value of corn in a pound of beef was about 40 cents, or 10 cents in a quarter pound hamburger. If corn when up to $37.00, a lot of people in poor countries would starve to death. But the price of a burger would go up just 90 cents.

Also, empirically, retail prices tend vary less than wholesale prices in absolute terms. In other words, I was being conservative--a pound of beef would probably go up a lot less than 90 cents.

On the topic of commodity prices, Michael also cites a recent paper on whether futures prices are effective predictors of commodity prices; the conclusion seems to be they they outperform a random walk, but that isn't necessarily significant in the statistical sense.