The Archdruid Report - Archive

I opened last week’s post by pointing out that many people nowadays fail to grasp some of the most basic realities facing us as the industrial age comes to an end. That turned out to be a rich irony, for a great many of the comments I received in response to the post displayed a blind spot even bigger than the one I attempted to address. It’s a convenient irony, though, as it offers a useful way to start talking about an underexplored dimension of the predicament of our time.

The post in question pointed out that today’s much-hyped "information superhighway," far from being the wave of the future so many of its promoters claim it to be, was a temporary product of the last hurrah of the age of cheap energy and can't be expected to survive for long as that age winds down. Instead, as the economic burden of the internet's immense energy usage begins to bear down, other technologies less dependent on huge energy inputs will become more economical, driving a spiral in which rising costs and restricted access will cut into internet service while simpler technologies absorb a growing range of its current economic roles. Finally, when economic contraction and social disintegration have proceeded far enough, the internet will simply drop out of use altogether because the economic basis for its operation will have gone away.

Most of those who objected to this sketch of the future, in turn, relied on a very curious logic. The internet will remain viable and widely accessible, they claimed, because the economic advantages of keeping it are so great. Those few who addressed the issue of costs at all simply insisted that technological progress would allow the internet to use less power than it does at present, and left it at that. The same arguments, interestingly enough, were deployed in earlier discussions about railroad technology: most critics simply insisted that railroads were efficient and economically advantageous, while a few suggested that they could be run more efficiently than they are now.

All this is true, but it misses the central issue I've tried to raise in the last few posts – the impact of energy and resource scarcity on the relative costs and benefits of different technologies – and it also dismisses the even broader issue of whether such energy-intensive technologies are sustainable at all in the future ahead of us. It's a dizzying departure from reason to insist that the advantages conferred by the internet mean that the internet must continue to exist. The fact that something is an advantage does not guarantee that it is possible.

An example from one of the most famous cases of social collapse is relevant here. On Easter Island, as I think most people know by now, the native culture built a thriving society that got most of its food from deepwater fishing, using dugout canoes made from the once-plentiful trees of the island. As the population expanded, however, the demand for food expanded as well, requiring more canoes, along with many other things made of wood. Eventually the result was deforestation so extreme that all the tree species once found on the island went extinct. Without wood for canoes, deepwater food sources were out of reach, and Easter Island's society imploded in a terrible spiral of war, starvation, and cannibalism.

It's easy to see that nothing would have offered as great an economic advantage to the people of Easter Island as a permanent source of trees for deepwater fishing canoes. It's just as easy to see that once deforestation had gone far enough, nothing on Earth could have provided them with that advantage. Well before the final crisis arrived, the people of Easter Island – even if they had grasped the nature of the trap that had closed around them – would have faced a terrible choice: leave the last few big trees standing and starve today, or cut them down to make canoes and starve later on. All the less horrific options had already been foreclosed.

Further back in Easter Island's history, when it might still have been possible to work out a scheme to manage timber production sustainably and produce a steady supply of trees for canoes, this would have required harsh tradeoffs: one additional canoe per year, for example, might have required building or repairing one less house each year. Both the canoe and the house would have yielded significant economic advantage, but it wouldn't have been possible to get both. In a world of limited resources, in other words, it's not enough to insist that a given allocation of resources has economic advantages; you must also show that the same resources would not be better used in some other way or for some other need.

The survival of the internet in an age of dwindling energy supplies is subject to the same hard logic. The internet demands huge inputs of energy and resources. Those were easy to provide during the quarter century from 1980 to 2005, when the price of energy was artificially forced down to the lowest levels in human history, and the same glut of cheap energy made it possible to build and power the internet without impacting other sectors of the economy. As energy becomes scarce and costly in the not too distant future, on the other hand, the demands of the internet will begin to conflict with the demands of other economic sectors. The task of managing those conflicts will likely be the supreme economic challenge of the century ahead of us, not least because we are so utterly unused to thinking in terms of hard tradeoffs; we assume, blindly, that we can have it all.

Now it's true, of course, that the internet could be operated more efficiently than it is today. Efforts to increase efficiency, however, are subject to a law of diminishing returns; a range of limits ultimately rooted in thermodynamic laws put a ceiling on just how efficient any process can get. Such gains also have costs of their own; research and development does not come cheaply these days, nor does the construction and installation of more efficient equipment, and the budget cuts currently sweeping through companies and universities worldwide – themselves the harbingers of much greater cuts to come – do not exactly support the act of faith that claims infinite technological improvement as the answer to this and all other problems.

Nor is it valid to put the possibility of increased efficiency for the internet on one side of the balance and ignore the equivalent possibilities on the other side. After all, other technologies – some of which are already simpler and more efficient than the internet – are just as liable to see gains in efficiency as the internet. Even a more efficient internet is unlikely to be the most economical way to use the sharply constrained energy and resource flows of the deindustrializing future; if another technology or suite of technologies can provide something like the same services at a lower cost, that technology or suite of technologies will outcompete the internet. Thus if it costs less, all things considered, to send messages over shortwave radio, order products by mail from a catalog, and get pornography from a local adult bookstore, than to do the same things over the internet, then the internet will fall by the wayside, or at best will be propped up for noneconomic reasons as long as economic realities make it possible to do so.

It's crucial to remember that the entire supply chain that keeps the internet and its potential competitors running has to be factored into these calculations. It's easy to see the internet as uniquely efficient if all you take into account is the energy going into your home computer, or even if you consider the gigawatts used by server farms. Putting those gigawatts to work, however, requires an electrical grid spanning most of a continent, backed up by the immense inputs of coal and natural gas burnt to put electricity into the wires, and a network of supply chains that stretches from coal mines to power plants to the oil wells that provide diesel fuel for trains and excavation machines; the server farms draw on a vast array of supporting services and manufactures, from the overseas mines that produce rare earths for semiconductor doping through the factories that turn out components to the colleges that turn out trained technicians, and the list goes on.

All told, a fair fraction of the world's industrial economy helps support the internet in one way or another, and many of those support functions can't be done at all in a less centralized way or at a lower level of technology. Most of the potential replacements for the internet don't suffer from that limitation. It's entirely possible to build a shortwave radio by hand, for example, using components that can be built by hand from readily available materials; there are radio amateurs alive today who did precisely that before the postwar electronics boom made manufactured components cheap and easily accessible. In a world where the cost of energy is a major economic burden, these differences will matter, and give a massive economic advantage to less energy-intensive ways of accomplishing things.

One useful way to assess the vulnerability of any current technology in a world on the far side of Hubbert's peak, in fact, is to note the difference between the direct and indirect energy inputs needed to keep it working and the inputs needed for other, potentially competing technologies that can provide some form of the same goods or services. All other factors being equal, a technology that depends on large inputs of energy will be more vulnerable and less economically viable in an age of energy scarcity than a technology that depends on less, and the bigger the disparity in energy use, the greater the economic difference. In turn, communities, businesses, and nations that choose less vulnerable and more economical options will prosper at the expense of those that do not, leading to a generalization of the more economical technology. It really is as simple as that.

You might think that this sort of economic analysis would be an obvious and uncontroversial part of peak oil planning. Of course it's nothing of the kind. Most discussion and planning around the subject of peak oil these days pays no more than lip service to economics, if it deals with that dimension at all, and a great many of the plans being circulated these days look very appealing until you do the math and discover that the most basic questions about resource inputs and economic outputs haven't been addressed.

Now part of this blindness to the economic dimension is hardwired into contemporary culture. It hardly needed the mass exodus into delusion that drove the recent real estate bubble to prove that most people in the industrial world nowadays think that getting something for nothing is a perfectly reasonable expectation. We have lived with such abundance for so long that a great many of us seem to have lost any sense that there are limits we can't borrow or bluster our way around. To a very great extent, indeed, the last three hundred years of economic expansion have been driven by a borrowing binge even more colossal, and ultimately more catastrophic, than the one imploding around us right now. Instead of borrowing from banks, we borrowed from the Earth's stockpile of fossil carbon, and squandered most of our borrowings on vaster equivalents of the salad shooters and granite countertops that absorbed so much fictitious value during the late boom. By the time Nature's collection agencies get through with us, in turn, they may just have repossessed everything we bought with our borrowings – which is to say nearly everything we've built over the last three centuries.

Yet there's another source feeding into this blindness, because the theories of economics that have been used to try to make sense of the flows of natural and manufactured wealth in our societies are hopelessly inadequate to the task. It's difficult to construct a meaningful economic analysis of the future within a paradigm that insists that resources magically appear whenever there's money to pay for them, for example, or claims that damage inflicted by human economic activities on the natural systems that allow our economy to function in the first place are "externalities" that need not be considered in cost-benefit analyses. Current economic theory commits both these howlers, and others as well.

With next week's post, we'll begin a more detailed exploration of what an economic vision relevant to a deindustrializing future might look like. That exploration will start from the work of E.F. Schumacher, who was one of the most thoughtful (and heretical) economists of the last century, as well as an early (and rarely remembered) peak oil theorist. Using his ideas as a springboard, I hope to take today's discourse about the future of industrial society into unexplored territory, and – not incidentally – provide some unexpected but practical tools for coping with the arrival of the deindustrial age.