. . . is small and visually unimpressive. At the moment, it's in my living room. On an end table beside my favorite chair. It's table 3.1 in Chapter 3 of Ecological Economics: An Introduction  by Michael Common and Sigrid Stagl. (It's written as a college-level textbook. Greenback's library didn't have it, but the Backboro public library did. Go figure.)
Anyway, Table 3.1 has the tremendously engaging title of "Energy Accounts for Food Production". The data presented comes from a volume that somehow missed the NYT best-seller lists -- Energy and Food Production , by Gerald Leach. (Sexy stuff, eh?)
The table includes three columns:
- Column 1 reflects the energy inputs required, and the energy outputs produced, by a hectare of land utilized by a hunter-gatherer society. The numbers were generated based on empirical observation of !Kung in the early 20th century.
- Column 2 shows energy in and energy out for a hectare of land used in rice production as practiced in China during the 1930's. Since only animal and human labor are involved, it's intended to represent pre-mechanical agriculture as practiced over the bulk of human history.
- Column 3 depicts energy in and energy out in modern mechanized rice production, as practiced in Louisiana during the 1960's. (Hey, Leach's book came out in 1975. How modern do you want?)
Energy inputs are broken down in terms of human labor, animal labor, machinery (creation and use), fertilizer (ditto), pesticides (same) and irrigation (likewise). Energy out is based on the energy content of the food produced. Ratios are shown for total energy out/total energy in, and total energy out/human labor in. Everything is expressed in megajoules, but it's easily converted to calories by anyone sufficiently interested.
In a nutshell, what Leach discovered, what Common and Stagl summarize, is that hunter/gatherers reap 7.8 times as much energy from the food they acquire as they spend in the process of aquisition; since no animal or mechanical labor is involved, all input energy is in the form of human labor. Pre-industrial agriculturalists reap 41.1 times as much energy as they invest and, since they use animal labor (the accounting for which fully reflects the need to feed the animals), the return on human labor is 49.7 to 1. But when agriculture (or, at least, rice cultivation) is industrialized, the return on human labor invested jumps to a phenomenal 4206 to 1. Impressive!! But . . .
But the ratio of energy harvested to total energy consumed by the food production process -- including machine work, fertilizer, pesticides and irrigation -- drops to a miserable 1.3 to 1. Less than 20% as efficient per unit energy invested as hunter-gathering! About 3% as efficient in terms of energy utilization as pre-industrial agriculture.
Oh, and total output per hectare of land dropped from 281,100 Mj (pre-industrial) to 84,120 Mj (industrialized). You don't have to be an expert in the metric system to do that math.
Now I'm not suggesting we all move into the forests and become hunter/gatherers. The simple truth is that there's not enough land (even if the land were appropriately forested) to feed us all that way. But I am suggesting that any argument that only industrial-scale agriculture can feed the ever-burgeoning human population is fundamentally flawed. Modern agricultural techniques, I'm sure, achieve better results than Leach's rice farmers from the 1960's. But how much better? Ten times as good? (Gee, that would be a ratio of 13 to 1 -- almost a third as good as achieved by non-mechanized farmers.)
More important, Table 3.1 emphasizes just how important it is that we question the inevitability of modern production methods of all sorts. And challenge assertions about efficiency. Efficiency is a ratio, not an absolute. It depends entirely on what you're dividing by -- efficiency per unit time, unit land, unit energy, unit capital, unit whatever. Industrial agriculture, by these numbers, is incredibly efficient per unit of human labor. And it's effectively designed to maximize the investment of financial capital (payment for which consumes most or all of the financial return generated). But human labor is in over-supply, as is financial capital. Those are not the efficiencies that any rational manager, looking at the big picture, would identify as important. (Rule of thumb -- design your process to achieve the best possible efficiency in use of the least-available resource. That's the way to maximize output.)
The least available resources are energy (particularly fossil energy, used to power machinery and irrigation, and to manufacture fertilizers and pesticides) and land (not only are they not making any more of it,the supply of good agricultural land is decreasing daily due to erosion, infertility, compaction and salinization from deep-aquifer irrigation). If we want to maximize food production in a world where land and energy are the key resources, the non-mechanized farm (or, in some cases, the minimally-mechanized farm) looks amazingly attractive. We could more than triple our production per unit of land, while decreasing the energy consumed in agriculture by over 95%.
Oh, and since more of that energy would be in the form of human labor, we could decrease the greenhouse gas emissions generated by agriculture by something like 99%.
All of which, of course, is based on extrapolation and over-generalization from a very limited set of data. The numbers for other forms of agriculture -- and other industrial activities -- are doubtless different from those for rice-farming.
One little table doesn't hold all the answers. But it certainly does raise some interesting questions. And it can force us to question a lot of the putative advantages of industrialization in all its forms.