An Economic Theory Shortage. (in Review)
Kleit, Andrew N., Regulation
HUBBERT'S PEAK: THE IMPENDING WORLD OIL SHORTAGE
By Kenneth S. Deffeyes
285 pp., Princeton, N.J.: Princeton university Press, 2001
Kenneth Deffeyes' book Hubbert's Peak has gained a wide audience in the oil production community. One wonders why. As a work of literature, it is highly tendentious; the reading is not easy, and the chapters appear to be connected only loosely. Much of the book deals with the basics of petroleum exploration, and not in a manner that is likely to be pleasing to the reader. The tie to any type of economic conclusion from Deffeyes' text is tenuous, at best.
The book's core (which is hidden in the appendix and beyond) deals with predicting world oil production and when it will slow because of scarcity. While the math can be a little complicated, the analysis is based upon the assumption that production per year is shaped like the standard normal distribution function. Given that assumption, the author seeks to find when the peak of that normal distribution will occur, and when we will start "running out" of oil.
Deffeyes' approach is based on the work of M. King Hubbert, who predicted in 1956 that U.S. production would peak in the early 1970s. Deffeyes' prediction for world production, not surprisingly, is that we will soon near the peak -- perhaps as soon as 2003. The author thus concludes, "There is nothing plausible that could postpone the peak until 2009. Get used to it" (p. 158). After the peak, of course, the price of oil will rise, economies will slow drastically, and we will all have to change our lifestyles, somewhat for the worse. Yes, you have heard it before.
Information without theory is useless. The exercise Deffeyes undertakes is wholly without theory. It simply assumes that oil production over time takes the approximate shape of a normal distribution. So not only will we eventually reach a peak, but the "backside" of the distribution will look just like the front side. Deffeyes does not spell out why that should be the case.
Hubbert may (or may not, the data are unclear) have gotten the peak (though not the backside of the distribution) right for the United States, but even a stopped clock is right twice a day. Without an appropriate theory, the mathematical models Deffeyes presents simply do not allow for a substantive conclusion.
Scarcity and equilibrium To see the difficulties with what Deffeyes presents, think about what it would mean if the Hubbert-Deffeyes theory is correct. It implies that very soon the price of oil will be sky-high. If you knew this, what would you do? As for me, I would buy oil, in all its relevant forms, as soon as I could. …