"Risk" and "Probability": MIA in the Peak Oil Debate

People desperately want to know the one thing which they cannot know: The future. All sides in the peak oil debate are only too happy to oblige with confident predictions about the future which flatly contradict one another. Strangely, it is not only the charlatans who offer such cocksure pronouncements, but also many thoughtful, accomplished people who are competent in every other way, yet who feel compelled to share their opinions as if they were channelling Nostradamus.

The result can be a game of journalistic ping-pong in which unequivocal pronouncements from various sides of the peak oil debate leave the public confused. Many in the public simply dismiss this debate; after all, if the experts can't agree, let them sort things out before they bother me.

While it is true that the public is looking for definitive answers about the complex issue of oil depletion, it is also true that there are none. We are all groping forward in a twilight of partial and often uncertain knowledge. What the public desperately needs to know are the risks we run as fossil fuels deplete. In short, they need a brief course in probability and risk that can prepare them to think clearly about the path we should all take individually and collectively.

When I speak before groups I often ask who has fire insurance on their homes or apartments. Usually, nearly every hand is raised. Then, I ask how many have ever actually collected on a claim for a fire. Only very rarely does even one person raise his or her hand. "Why do you have the insurance then?" I ask rhetorically. The answer, of course, is that even though house fires are quite rare, their consequences can be quite severe. "And so," I continue, "we routinely insure against events which are rare because they have severe consequences." We do this with life insurance. We do it implicitly with many health insurance plans which have coverage into the millions of dollars--even unlimited coverage--that in all likelihood we will never need.

Another useful illustration I use is the stock market. "If I could prove to you that you could be certain the stock market will go up nine years out of the the next 10," I begin, "you would be at ease with your investments in stocks and you might even increase your investment in them." Everyone seems comfortable with that thought.

"Now, let me add one more piece of information to this scenario," I continue. "In one of those years the stock market is almost certain to go down 80 percent." This completely changes the calculus because the element of probability is combined with the element of severity.

We must combine these things in our discussion of peak oil. No one knows when oil will peak for certain. The range of opinion is such that one could feel justified in remaining entirely unconcerned about oil for another 30 even 40 years by which time, we are told, a new energy infrastructure will be in place. But should we really speak in terms of whether a prediction for the peak of 2008 versus 2037 versus 2050 is correct? No one will know which is correct until after the fact. While the actual date of the peak has huge implications for what we should do now, it is the severity of the consequences of a peak which should be our focus. Are we adequately prepared for a peak even if we believe there is a low probability that it will occur in the next few years?

We should talk in terms of probability rather than absolutes. For example, can we rule out a peak entirely in the next few years? Can we reasonably say that the consequences of a peak are something we don't really need to worry about? In the world of probability, it matters whether you are worrying about a minor risk such as getting a hangnail or a major risk such as getting your arm chopped off. When it comes to the consequences of peak oil, "getting your arm chopped off" is a closer analogy.

Even the oil optimists agree that predicting a precise date or even year for peak oil is problematic. That is part of the reason for their skepticism concerning predictions of a nearby peak. But, oil optimists such as Daniel Yergin and Michael Lynch should be challenged in public to say whether they believe there is a zero chance of an oil peak before 2010. Do they also believe there is a zero chance before 2015 or 2020 or 2030? On what do they base their 100 percent confidence? Can they show us that they have perfect knowledge of the world's oil resources and the path of consumption and technological innovation through these dates?

Of course, both these optimists and their followers are merely assigning a high probability to a later peak. They cannot rule out an earlier one, but can only say that they regard it as unlikely. This means they must agree that with each year the possibility of a peak grows. Since oil is a finite resource, we know that we are depleting it once and for all with every fill-up. And, that means that even in the face of uncertain knowledge we know we are drawing ever closer to the peak as each day passes.

A nearby peak--and by this I mean within the next 10 to 15 years--would be, at the very least, highly disruptive and, at worst, civilization-destroying. Doesn't the possibility of such severe consequences demand our attention even if we believe that the probability of a nearby peak is small? This is the question we must pose to move the argument away from mere point-counterpoint where the validity of precise predictions on both sides is always questionable.

Most of us are not so foolish as to leave all our worldly goods exposed to the risk of a house fire without insurance. Unlike house fires, however, the risk of a peak in world oil production grows every day. Isn't it time to take out some insurance against that day based on the risks we already know about?

[For my suggestions on what form that insurance might take, see my previous post, Peak Oil 'To Do' List: Why We Should Do These Things Anyway.]