The Internet's role in aggravating and alleviating the energy crises Andrew Odlyzko University of Minnesota http://www.dtc.umn.edu/~odlyzko The Internet has long been touted as a major weapon in reducing road congestion, through enabling and stimulating telecommuting. More recently, it has been promoted as a means to fight global warming. (An excellent source of information on these topics is Bill St. Arnaud's "Green IT/Broadband and Cyber-Infrastructure" blog at http://green-broadband.blogspot.com/.) But these hopes fly in the face of centuries of experience. It is likely that the Internet will play a major role in alleviating problems caused by any of several energy crises that we face, but if so, this will not take place as commonly expected. The expectation that communication and transportation are substitutes for each other has been widespread. J. W. Croker, in arguing against the Penny Post reform in England, wrote that the increase in travel by horse-drawn coaches "militates ... against an increase of correspondence" (pp 524-525 in "Post Office reform," published in the Quarterly Review, Oct. 1839, pp. 513-574). Croker turned out to be wrong, as in the years following his article both passenger travel and postal correspondence grew explosively. But this did not stop others from repeating his error, or similar mistakes, such as expecting that the telegraph or the telephone would kill ordinary mail. (For extensive compilation of evidence on the spread of various communication technologies in synergy with apparent competitors, see my 2000 manuscript, "The history of communications and its implications for the Internet," http://www.dtc.umn.edu/~odlyzko/doc/history.communications0.pdf.) The reasons for the simultaneous growth of supposed substitutes are manifold. There is a direct stimulation effect (higher capacity on fiber cables enables off-shoring jobs to India, but the workers there have to be trained, supervised, and coordinated, which means a jump in air travel to and from India). There are also improvements from one technology that are absorbed into another (computer and communications improvements led to better and much less expensive printers, which satisfied the latent demand for printing on the desktop, and have thus far made a mockery of the "paperless office" concept). And there is the old Jevons Paradox, which says that greater efficiency often increases usage (see, for example, http://green-broadband.blogspot.com/2008/02/greater-energy-efficiency-may-increase.html). All these combine to yield a simple observation, with overwhelming evidence going back centuries, that as society develops economically, both communication and transportation boom. (For data about communication, see "The history of communications ..." referenced above. For a survey of transportation, see the book of W. L. Garrison and D. M. Levinson, "The Transportation Experience.") They are both services whose consumption grows with technological and economic advances. Thus simply deploying the Internet more widely and with increased capacity is likely, in the absence of other developments, to stimulate travel and energy usage. This would be true even without counting the energy usage of the Internet and the computers connected to it. The counterintuitive conclusion of the preceding paragraph is not an argument for holding back on Internet deployment. The reason is that we appear to be subject to "other developments," ones that will radically transform how the economy evolves. The centuries of experience where both communication and transportation grew vigorously were also centuries in which (aside from minor hiccups) energy has been getting steadily less expensive (in inflation-adjusted terms) and more convenient to use. Human society started with wood, and then went on to coal, oil, and natural gas. But there does not seem to be any successor to these, at least in the short term. Instead, we appear to face three serious threats, any one of which can drive up the cost of energy by itself: 1. Rapid growth in the developing world: With China's economy growing at 10%, and India not much less, the traditional rate of technological improvement in economic efficiency and the rate of discovery of new energy sources can't keep up. 2. "Peak oil": The widely predicted, although still controversial, start of a decline in total world oil production. 3. Global warming: Surely no need to explain. "Unsustainable trends do not sustain themselves." And dire predictions do sometimes come true. (The effect of a Category 5 hurricane hitting New Orleans had been well known, it's just that political leaders for decades took the gamble that something like Katrina would not hit on their watch, or at least that they would not be blamed for the damage if it did.) So, unless all three of the energy threats above are avoided, the price of energy will continue to go up (with some potential slips, say from a recession or a disappearance of the war premium on Middle East supplies). And if energy becomes more expensive, the Internet may become a substitute for transportation (as well as an aid towards greater efficiency in a variety of ways), not because that is the usual outcome, but enforced by drastically changed circumstances. So in that sense the advocacy of greater broadband deployment may be very productive. But it will be a very different world than the one that led to our current technological and economic state. We may drive less, but not because we love to telecommute, or because we are tired of traffic jams (that had not stopped people from driving more in the past), but because the costs will get too high. Many of the rules we relied on to judge economic and technological developments will simply not apply.