Recent discoveries have typically been relatively small with high decline rates-and have been exhausted relatively quickly. With respect to the United States, it has been observed: "In the old days, we found elephants-now we find prairie dogs."

A growing consensus accepts that the peak is not that far off. It was a geologist, M. King Hubbert, who outlined the theory of peaking in the middle of the last century, basing it on the experience that as an oilfield passes the halfway point in extracting its reserves, its production goes into decline. Hubbert correctly predicted that production in the United States itself would peak out around 1970. Dissenting from that view are the economists, who have a deep (and touching) faith in the market mechanism-and a belief that over time market forces can adequately cope with any limits on oil supply.5 In the extreme, some economists have regarded oil supplies as almost inexhaustible.

The optimistic view is held by the Energy Information Administration of the Department of Energy, as well as the International Energy Agency. What lies behind it? While it is conceded that we have not been finding many new giants, it is contended that "additions and extensions" of existing fields will sustain growth. There is some truth in that contention-in that new technologies have been the basis of many of the additions to existing fields-and the hope is always there that we can increase overall recovery from the already discovered fields.

Optimists are buttressed in their views and are fond of pointing to the many earlier statements about "running out of oil." Perhaps the most notable example was one by the director of the U.S. Geological Survey, George Otis Smith, who suggested in 1920 that we had already used up 40 percent of the oil to be found here in this country. That was a decade before the discovery in 1930 of the vast East Texas field, a bonanza that made oil supply so available that it drove oil prices below a dollar a barrel during the 1930s. A recent Chevron advertisement makes this substantive point quite dramatically: "It took us 125 years to use the first trillion barrels of oil. We'll use the next trillion in 30."

Such past failed predictions are far less comforting than the journalists who cite them believe. The future may actually be different from the past. The optimists, mostly non-experts, seem unable to think quantitatively. Things are different now. In 1919 the world consumed a modest 386 million barrels of oil. Today the world is consuming some thirty billion barrels of oil each year. Statements like that of Director Smith were made before we had something approaching a billion automobiles worldwide, before we had aircraft and air transportation, before agriculture depended upon oil-powered farm machinery.

Hubbert's peaking theory, based on observation of individual oil fields, was static in that it abstracted from improvements in technology. It also dealt strictly with conventional oil supplies. One notes that today those who are challenging Hubbert's Peak are changing the rules of the game. They rightly point to dramatic improvements in technology, most notably deep-sea drilling. Somewhat less legitimately, they include in their projections all sorts of unconventional oil, like the Canadian tar sands and the prospects for shale oil. For example, of late, estimates of Canadian oil reserves have jumped by 180 billion barrels, now including the tar sands of Alberta. This is not a refutation of Hubbert's theory (though it is frequently treated as such); it is simply a change in the rules that does not gainsay the fear that we will reach a plateau in conventional oil production.

We must bear in mind that earlier estimates suggested that there were some two trillion barrels of conventional oil in the earth's crust. Now the estimate has grown to around three trillion. We have now consumed over a trillion barrels of oil. As indicated, we are consuming oil at the rate of thirty billion barrels a year. If one accepts Department of Energy projections, worldwide we will be consuming forty billion barrels of oil by 2025.

At such rates of consumption, the world will soon have reached the halfway point-with all that implies-of all the conventional oil in the earth's crust. At that point, the plateau or the peak will be near. And such calculations presuppose what cannot be assumed, that all the nations with substantial oil reserves will be willing to develop those reserves and exploit them at the maximum efficient rate. Both the Russian Federation and Saudi Arabia seem to intend to reach a plateau that they can sustain for a long time-the Russians at around ten million BPD, the Saudis up to but no more than 15 million BPD.

In thinking about the problem, we need not more rhetoric but, instead, quantitative reasoning. We also need to add political wisdom. The inability readily to expand the supply of oil, given rising demand, will in the future impose a severe economic shock. Inevitably, such a shock will cause political unrest-and could impact political systems. To be sure, we cannot anticipate with any precision the year or even the decade that we will reach that plateau. Yet, as Justice Potter Stuart suggested, in seeking to define pornography, we shall know it when we see it.

That brings us to the question of the transition away from conventional oil as the principal source of energy for raising the living standards of the world's population. That transition will be the greatest challenge this country and the world will face-outside of war. The longer we delay, the greater will be the subsequent trauma. For this country, with its 4 percent of the world's population, using 25 percent of the world's oil, it will be especially severe.6 The Day of Reckoning is coming, and we need to take measures earlier to cushion the shock. To reduce the shock, measures to ameliorate it should start ten years earlier at a minimum, given the length of time required to adjust the capital stock-and preferably much longer. The longer we delay, the greater the subsequent pain.

Both people and nations find it hard to deal with the inevitable. Even though it was long recognized that a Category 4 or Category 5 hurricane would inevitably strike New Orleans, a city substantially below sea level, Hurricane Katrina reminds us that political systems do not allocate much effort to dealing with distant threats-even when those threats have a probability of 100 percent.

We should heed a lesson from ancient Rome. In the towns of Pompeii and Herculaneum, scant attention was paid to that neighboring volcano, Vesuvius, smoking so near to them. It had always been there. Till then, it had caused little harm. The possibility of more terrible consequences was ignored-until those communities were buried in ten feet of ash.

1 See World Energy Outlook 2005 (International Energy Agency, 2005)

2 See, inter alia, Robert L. Hirsch, "The Inevitable Peaking of World Oil Production" (Atlantic Council of the United States, October 2005), which includes a range of different estimates for the peak year. For a more comprehensive analysis, see Robert L. Hirsch, Roger Bezdek and Robert Wendling, "Peaking of World Oil Production: Impacts, Mitigation and Risk Management" (National Energy Technology Laboratory, February 2005).

3 One exception is a different view of oil's origins developed by Soviet scientists. Contrary to the standard view that oil, like coal, was laid down long ago and there is a finite amount available, the Russians argue that oil is a primordial product continuously produced deep in the earth's mantle. It comes to the surface when it can find a route to do so. Thus, there may be more oil to be found outside sedimentary basins. The theory remains highly conjectural. While this alternative view needs to be explored, it is notable that even the Russian Oil Ministry pays little attention to it in developing projections of Russian production.

4 See Pang Xiongqi, et al., "The Challenge Brought by the Shortage of Oil and Gas in China and their Countermeasures", a presentation at an international seminar in Lisbon in 2004. One may assume that such presentations do not depart significantly from the views of the Chinese government.

5 Many economists take great comfort from the conviction that there is always a price at which markets will clear, and that the outcome determined by supply and demand is not only inevitable, but is also politically workable and acceptable. An outcome in which the price of a crucial commodity like oil rises to a level causing widespread economic disruption, along with the political consequences that flow from such disruption, turns out to be a secondary consideration, if considered at all. One is reminded of the phrase used by Wesley Clair Mitchell and Arthur F. Burns in their classic, Measuring Business Cycles (1946), in which they spoke scornfully of the "Dreamland of Equilibrium."

6 The high percentage of world production consumed in the United States is used by critics to point to our presumed wastefulness. It is, however, misleading in that the United States also produces between 20 and 25 percent of the gross world product. Nonetheless, it does appropriately point to our greater vulnerability to a future period of oil stringency.

James Schlesinger is chairman of the Advisory Council of The National Interest. He has served as secretary of defense, secretary of energy and director of central intelligence. He is currently chairman of the MITRE Corporation.