Information and energy
Via Felix Salmon, I found this interesting piece by Tim Wu, comparing monopolies in broadband and energy, and looking at ways to make better use of currently idle spectrum. Wu’s starting point is that “Like energy, bandwidth is an essential economic input”. In fact, as he implies, information is more essential than energy to a modern economy. It’s massive amounts of information, rather than massive amounts of energy, that distinguishes our economy from that of 50 or 100 years ago.
But even taking the weaker position that information, like energy, is essential puts a different perspective on a lot of current debates, like those about Peak Oil and climate change. In much of this discussion, it’s assumed, explicitly or otherwise, that energy is uniquely important. In reality, its just one of a number of inputs that are essential (in the sense that we can’t do without them completely) but still subject to the ordinary economics of demand and supply.
If we are going to do anything about climate change, one implication is that energy must get a good deal more expensive, perhaps three or four times as expensive as it is now. That obviously has some economic impacts. But on the other side of the coin, the costs associated with information have been dropping at a rate that is hard to measure accurately, but certainly amounts to a halving every couple of years ago (things seem to have stalled a bit lately in relation to broadband, which is part of Wu’s point, but the underlying trend is obvious).
On any plausible estimate of the properties of demand, the benefits of ever-cheaper and more plentiful information will far outweigh the costs of less carbon-intensive energy. Partly this is a matter of substitution in production (for example, videoconferencing replacing business travel) and partly in consumption (as information based activities become cheap and energy-based activities become dear, people will switch from one to other). The numbers cited by Wu show that this is already happening.
UpdateSome data may help to clarify the point a bit. According to the US Internet Industry Association, the volume of information transmitted over the Internet backbone rose from 1.5 million GB (petabytes) in 1995 to 700 petabytes in 2006, or roughly a factor of 500 in 10 years. In comments, Ikonoklast reports a five-fold increase in global energy use over 50 years. Over the same period, Brad Delong estimates a tenfold increase in global output. That is, the rate of growth of information greatly exceeds (and leads) the rate of economic growth, while energy use has declined relative to output. This is unsurprising, given that no fundamentally new energy technology (except for the so-far unsuccessful nuclear power industry) has emerged in this time, while information technology has been repeatedly revolutionised. But it makes the point that an analysis of the economy based primarily on energy has been obsolete at least since 1950, and was arguably obsolete when Jevons first raised the alarm about the Coal Question back in C19.
On a more minor point, a bit of data on energy use in agriculture (sources to come). Agriculture accounts for about 1 per cent of direct energy use and maybe 1.5 per cent more through fertilisers and pesticide. In total, energy amounts to about 20 per cent of the costs of agricultural production, and a substantially smaller share of the retail cost of food.