Is Food the Last Thing to Worry About?

Our food system is woefully dependent on petroleum, as writers such as Richard Heinberg (1) and Michael Pollan (2) have eloquently pointed out. Soaring food costs have brought on riots in some countries, and in unstable nations, famine continues to be a regular visitor. Fears of empty grocery shelves have made food security the centerpiece of many a post-Peak Oil plan, and among those watching energy descent, a common refrain is that the best way to guarantee your food supply is to buy a piece of land and grow your own.

Yet in the developed world, especially the breadbasket nations such as the US, Canada, and other food-exporting countries, the food network may be one of the last systems to fail during energy descent. In developing a wise post-Peak strategy, assessing relative risks is critical. Devoting large amounts of time and resources to events that are less likely leaves us unprepared for more probable difficulties. I don’t want to discourage anyone from growing food—I’m a serious gardener myself and could list dozens of excellent reasons for doing it. But I think there are many reasons not to be focusing primarily on food as the system most likely to fail. This isn’t to say that industrial, oil-based agriculture is invulnerable, let alone sustainable. And we may see temporary shortages of specific foods. But there are many reasons why our fears of a food collapse—particularly when they lead us to a go-it-alone, grow-your-own response—may be distracting us from focusing on more immediate and likely risks.

First, two notes of clarification: This article is about net food-exporting nations such as the US, where I live. In the less-developed world, where food growing has been abandoned for export crops that are sold for cash to import commodity food, the food system is far more vulnerable. And by “food collapse” I mean a prolonged inability to produce essential foods, not brief or local shortages of certain items, or high prices while supplies are ample. Volatile commodities markets, weather, and the other gyrations of our uncertain era mean that temporary or local shortages can always occur.

Food gets a lot of attention in part because we need it to survive, but also because one solution to a food crisis—growing your own—seems doable. I suspect we focus on food in part because providing it appears much more possible than, say, keeping the financial, health care, or automotive industries running.

Why would I argue that food collapse in breadbasket nations is not likely, when today’s farming is so dependent on hydrocarbons? Our food system is complex—much more so than it needs to be—but many of our society’s other structures are far more complex, and thus more vulnerable. Joseph Tainter (3)  and others point out that complex systems need increasing energy inputs, and eventually reach a point of diminishing returns, so that the costs of complexity eventually outrun its benefits. When inputs decline, the most complex systems are often the first to fail, since they need vast resources to maintain them. With that in mind, we can ask what is likely to fail first during energy descent. That way, we’ll know what we should direct our energies toward preparing for.

Is it any wonder that one of the first complex systems to collapse has been our financial system? The energy and complexity used in Byzantine financial instruments such as collateralized debt obligations and credit default swaps, and in moving trillions of dollars through millions of highly orchestrated transactions each day, is immensely greater than what it takes to grow, process, and ship food. Another system teetering near collapse is health care, and it, too, is a fantastically complicated system needing sophisticated, expensive equipment and years of specialized training for practitioners, all administered by an insurance system of equally staggering complexity. Thus the most complex systems are already collapsing. When viewed through the lens of complexity, the relative robustness of the developed world’s food system, even as finance collapses and health care becomes increasingly unavailable, is less mysterious.

It would bolster my argument to show quantitative measurements of these systems’ relative complexity, and for these I’ll point to Howard T. Odum (4) and his concepts of emergy (not energy, but embedded energy) and transformity. Emergy measures the total solar energy used directly and indirectly to make a product or service. Transformity builds on this, and means the emergy of one type required to produce a unit of energy of another type. It describes conversion losses and energy quality. For example, think of a food chain. A million calories of solar energy can make a given quantity of algae. When plankton eat this, it might yield 1000 calories of plankton. These plankton, when eaten, become one calorie of fish. Thus the transformity of that one calorie of fish is one million calories: the amount of sunlight used at the beginning of the food chain divided by the one calorie of fish produced. The plankton, being lower on the food chain, have a lower transformity: 1000 calories, or a million calories of algae divided by 1000 calories of plankton produced.

Processes that have higher transformity don’t just need more energy per output. They also contain more energy conversion steps, which bring efficiency losses and places for the system to fail. Also, high-transformity systems usually need more complex technologies than processes of lower transformity. Plankton are simpler than fish.

So how complex is our food system? Odum’s work  tells us that food transformities in industrial cultures are on the order of 25,000 to 100,000 sej/J (solar emergy joules input per joule gained). This is low compared to nearly all other familiar goods and services. Odum says that the production of paper has a transformity of 215,000 sej/J; electricity, 200,000 sej/J; cement, 750,000,000 sej/J; and complex transactions based on digital technologies, such as investment banking, have transformities in the billions or higher. If complexity, transformity, and stability are related—and I think they are—then activities of great complexity and high transformities, including office jobs, electricity, communications, and nearly all social and economic services, will be disrupted before food production will be. We’re seeing that process unwind today. Training and supplying an investment banker or surgeon is more complex than doing the same for a farmer. As complexity plummets due to energy descent, jobs and products of lower transformity are more likely to remain.

But even if the food system isn’t all that complex, you might argue, we have paved over much of our farmland and use oil to make food. Let’s look at the numbers. The US is a net exporter of food, and produces roughly 4000 calories of food per person (5). To stock this larder, the US uses roughly 3 million barrels per day of petroleum, or 15% of our total consumption (6). Thus the US could cut the amount of oil used by the food system in half and still provide a basic 2000-calorie diet. That’s 1.5 million barrels per day or its equivalent, which should be available for some time. This means that neither complexity nor oil are likely to be limiting factors on food production in breadbasket nations until after the failure of other more complex, energy-intensive elements of our lives.

Cheap oil has freed us to pour staggering amounts of energy, both human and fossil, into non-essentials, such as the entertainment, recreation, tourism, sports, media, and other fuel-gobbling industries. Inexpensive oil lets much of the developed world endlessly buzz around in inefficient cars and jets. In other words, 85% of our fossil-fuel consumption is used for things other than food, usually wastefully. As oil becomes expensive we will choose to redirect a modest portion of that 85% away from long commutes, non-essential industries, and other symptoms of cheap oil, in order to feed ourselves. It’s likely that as we round Hubbert’s bend we’ll return to putting 30-50% of our energy use toward food production, as has been the case for most of human history (7). This reordering of oil priorities can buy us the time needed to reconfigure our grossly inefficient, hydrocarbon-based food system into something far more localized and sustainable, if we’re smart.

Another oft-cited argument for food collapse is that fossil-fuel supplies are unreliable. What if foreign producers cut us off? The US currently produces about 5.2 million barrels of oil per day. Canada and Mexico are the top two petroleum importers for the US, providing about 40% of our imports, or 3.8 million bbl/day (8). Thus 9 million bbl/day are currently available from nearby sources. That’s three times the oil used by our food system, and six times what is needed for a basic diet. Natural gas, used to make nitrogen fertilizers, is a critical agricultural resource that also comes from relatively stable sources. Canada provides 95% of America’s natural-gas imports. The continent’s intertwined economies and the realities of geopolitics make it probable that hydrocarbons will flow long enough for the US to shift to a less oil-intensive agriculture. Obviously, oil output will continue its decline, and there are bound to be periodic crises, but the numbers suggest that starvation in the US is far from a certainty.

Food production is truly the oldest profession. We’re good at it, we’ve been doing it for 10,000 years, and it is a relatively simple system to run. It is at the base of a large cultural pyramid, which makes it fundamental, so although disrupting it would be catastrophic, it is also more elementary and thus easier to keep running than all the systems above its level of complexity. There are gardeners in over 71 million American households (9), so there is a sizable knowledge base to help with the transition to more local food production.

Almost certainly, food will shift from being a minor piece of the US economy to once again requiring one-third to one-half of our labor and energy. The example of Cuba, which in a few years retooled its agriculture system after a sudden and near-total cutoff of oil, shows that food systems can be modified quickly. How long would it take us to convert the nearest city park, or a soybean field that’s growing feedstocks for newspaper ink and car lacquer, into food production if it were urgent? One season. The recent substitution of ethanol corn for soybeans over vast acreages in a single season shows how quickly farmers can respond to new markets. And as food prices rise, people thrown out of work by energy descent will find jobs growing food, as Sharon Astyk and Aaron Newton have suggested in their book, A Nation of Farmers.

As cheap shipping disappears, can we feed ourselves locally? To gauge this, we need to know if there is enough farmland near cities to feed their populations. Researchers at Cornell University found that the basic calories to feed Rochester, New York’s population of 225,000 could be grown on existing cropland within 16.5 miles (26.6 km) of the city limits and would cover 36,000 hectares (90,000 acres) (10). This admittedly simplistic analysis looks only at caloric needs, not overall nutrition. To provide a balanced and diverse diet might require a larger area, so let’s say we’d need twice as much land, or 180,000 acres. That area is still within 25 miles of the city, close enough to easily bring goods to market. This could save much of the fuel used today to transport the infamous 1500-mile salad. Plus, the Cornell analysis assumes wasteful conventional agriculture techniques, not high-intensity ones that use local nutrient sources such as composted waste and animal and human manure, as well as other resource-saving methods that people dependent on local food would readily use. It’s probable that the largest cities, such as New York, would be unable to feed themselves locally, but it is likely that for them we will set fuel priorities to ship food from more distant farms.

And it is the reordering of fuel priorities that leads us to one of the most powerful reasons that food supplies are less likely to run out than almost any other resource. Politicians understand that hungry people topple governments. We’re deeply imbued with cultural lore reflecting this. Most people know little else about Marie Antoinette other than the apocryphal taunt to starving peasants that ensured her rendezvous with the guillotine, “Let them eat cake.” Trotsky noted that every society is only three meals away from a revolution. History shows that any functional state short of a kleptocracy will allow almost every other service—health care, banking, sanitation, schools, transportation—to languish before it allows its people to go hungry. Preserving the flow of at least 1.5 million barrels of oil per day for food will be a critical priority of the US government.

Let me be the first to admit that there’s still some chance of food collapse. Perhaps stupid or corrupt leaders will choose to direct energy resources not toward food but to the military or the rich. Or it’s possible that the link between the financial sector and food, via the futures and commodities markets, may play havoc with food supplies. And it’s certain that adjusting from today’s food consuming 10% of the average family budget to the historical norm of 30% to 50% will be disruptive.

Whatever your chosen post-Peak scenario, it’s smart to keep emergency food and water on hand, as much as makes you feel comfortable. But focusing on preparations for a food-system collapse reminds me of the story of the fellow searching for his keys under the streetlight. He didn’t lose them there, but that was the only place where the light was bright enough to see. In crisis, we often default to doing what we know even if it’s not the wisest action. We can’t individually fix the economy or health care, yet we certainly can grow some food, and that may be why it is central to many post-Peak plans. And I agree: growing food is simple. It’s an ancient skill that is at the heart of human culture, and even in its industrial manifestation, it is a robust system that is less complex and energy-intensive than most of society’s other activities. That’s why I suspect the food system will last longer than much of the rest of the oil society. Although brief disruptions are certainly possible, in breadbasket nations food is more likely than many other aspects of our culture to make it through the transition.

But for a thousand other reasons, plant a garden anyway.


1. Heinberg, Richard. “What Will We Eat as the Oil Runs Out?”

2. Pollan, Michael. “Farmer in Chief,” New York Times Magazine, October 8, 2008.

3. Tainter, Joseph. The Collapse of Complex Societies.

4. Odum, Howard T. A Prosperous Way Down.

5. Putnam, J, J Allshouse, L. S. Kantor. U.S. Per Capita Food Supply Trends

6. I’m taking the middle of estimates that vary from 19%, (see Michael Pollan, above), to 10% (see Martin C. Heller and Gregory A. Keoleian; Life Cycle-Based Sustainability Indicators for Assessment of the U.S. Food System.

7. Braudel, Fernand. The Structures of Everyday Life.

8. Energy Information Administration. Crude Oil and Total Petroleum Imports, Top 15 Countries.…

9. National Gardening Association, 2005. Environmental Lawn and Garden Survey.

10. Peters, Christian J., Arthur J. Lembo, and Gary W. Fick, 2005. A Tale of Two Foodsheds: Mapping Local Food Production Capacity Relative to Local Food Requirements.

Previous post:

Next post: