Running Out of Oil Was Forecast for a Century, So Why Hasn't It Happened

Imagine a weatherman who has predicted rain every day for a hundred years and has been wrong most of the time. Now imagine he is still on the air, still grim, still looking at the camera like a man about to deliver the news of a death in the family. We have a name for that weatherman in the energy business. He is called an expert on peak oil.

I do not say that to be unkind. Many of the people who made these forecasts were serious, and some of them were right about parts of it. But the history is long, it is documented, and it ought to make all of us humble about the next prediction. Let me walk through it, then explain why the doom keeps failing, because the reason is not luck. It is a mistake about what a barrel of oil is.

The first warnings

The fear of running out is almost as old as the oil business. Not long after the first American wells in the nineteenth century, people worried that the supply would be exhausted. By the years around the First World War, the concern had become a matter of government. As the automobile spread and navies converted from coal to oil, some officials and geologists warned that American reserves were limited and might not last long. In the 1920s the concern led President Coolidge to create a Federal Oil Conservation Board, in December 1924, to study the nation's oil.

I am being general because I do not want to hand you a number that I cannot source, and you should be skeptical of anyone who quotes the old estimates with great confidence. The pattern is what matters, and it is well documented: official estimates of the remaining supply, made in the early twentieth century, were repeatedly overtaken by new discoveries. In 1930 the East Texas field came in, then the largest field ever found, and the problem in the early 1930s was not scarcity but a glut that crushed prices and led states to regulate production. The Texas Railroad Commission's proration system, which limited how much each well could pump, grew out of that surplus.

So the early history is the first lesson in miniature. The year people declared the end of oil is often the year before somebody finds more.

Hubbert's curve

The most serious version of the claim arrived in 1956. M. King Hubbert, a geophysicist who worked for Shell, presented a paper to the American Petroleum Institute meeting in San Antonio. He argued that oil production in any region follows a bell-shaped curve: it rises, peaks and declines. Applied to the lower forty-eight states, his analysis put the American peak in 1965 in his most likely case and in 1970 in his upper-bound case.

And here I will give him his due. He was close. American crude oil production reached its high point in 1970, and it then declined for decades. Hubbert's prediction came true in the narrow sense that the country's output peaked at the late end of the range he gave. That was a real achievement, and it is why the peak-oil movement treated him as a prophet. For a generation after 1970 it seemed that the curve was an iron law.

Notice what that proves and what it does not. It proves that production from a given set of fields, using a given set of technologies, at a given price, rises and falls. It does not prove that production from the whole country is bound by a fixed stock. Hubbert's model assumed a fixed ultimate amount of recoverable oil. The word "recoverable" is carrying more weight than anyone noticed at the time. What is recoverable depends on what it costs to get it out, and what it costs depends on technology.

The seventies and the president's address

The 1970s turned the theory into politics. The Arab oil embargo of 1973 and the price shocks that followed made energy a daily worry for American families, with gas lines and rationing and a sense that the country was at the mercy of foreign suppliers. In April 1977, President Jimmy Carter addressed the nation on energy. He called the effort the moral equivalent of war and told Americans that the era of cheap, abundant energy was ending. He said, in effect, that we were running out of oil and gas and that the country had to change its habits and accept a greater role for government in managing supply and demand.

I do not want to be unfair to a man who was trying to deal with a real problem. The embargo was real, the lines were real, and the dependence on unstable suppliers was real. But the diagnosis in 1977 was that the supply of oil was finite and about to tighten, and the prescription was a program of conservation mandates and federal planning. What followed was not a world without oil. It was a long period of falling prices in the 1980s, as high prices drew out new supply, as consumption responded and as non-OPEC production grew. The price signal did the work the planners said they would do, and it did it faster.

The peak-oil books of the 2000s

The idea came back with force in the early 2000s. Books with titles that included words like "peak" and "twilight" argued that the world was at or near its maximum oil production and that decline was imminent. Kenneth Deffeyes wrote Hubbert's Peak in 2001. Matthew Simmons, an energy investment banker, published Twilight in the Desert in 2005, arguing that Saudi Arabia's fields were in worse shape than advertised. Oil prices had been rising through the decade, and the books fit the mood.

Then something remarkable happened, and it happened in the United States, the country that Hubbert had said was past its peak. A wildcatter named George Mitchell had spent years working to make gas flow from the Barnett Shale in Texas. His persistence, and the combination of horizontal drilling and hydraulic fracturing, opened up rock formations that were known for decades but were considered uneconomic. Producers applied the same techniques to oil in places such as the Bakken formation in North Dakota and Montana, and later in the Permian Basin and elsewhere. American crude production, which had been declining for most of four decades, climbed. By 2018, according to the Energy Information Administration, it passed the 1970 record. Within a few years the United States was the largest oil producer in the world.

Let me put that in the order of a numbered argument, because that is where the real lesson lives.

1. In 1970, American production peaked, and the forecasters said it would never again approach that level.
2. For about forty years the forecast looked right.
3. Then technology changed what was economically recoverable, and the production record was broken in 2018.
4. The oil had been there all along. What changed was our ability to reach it at a price someone would pay.

The curve did not break because Hubbert's arithmetic was wrong. It broke because the curve describes a technology, not the planet.

Resources versus reserves

This is the idea that unlocks the whole story, and it is the one that almost nobody in the news coverage explains.

• Resources are the total amount of oil that is believed to be in the ground, whether or not anyone can get it out profitably.
• Proven reserves are the part that can be recovered with reasonable certainty under existing economic and operating conditions. That phrase is from the reporting standards used by regulators and companies. The words "existing economic conditions" are the key.

Reserves are an accounting number, not a geological one. If the price of oil goes up, or the cost of extracting it goes down, reserves go up without a single new barrel being created. A company does not drill all the oil it knows about. It books reserves it can profitably produce, which means the number tracks the price and the technology. A news story that says "the world has forty years of oil left" is dividing today's proven reserves by today's consumption, which is like telling a family it has three weeks of groceries because that is what is in the refrigerator.

The same thing happens with every resource. The stone age did not end for lack of stone. Whale oil lit American lamps in the nineteenth century, the price rose as whales became harder to catch, and kerosene from petroleum took over, so the whaling business collapsed before the whales ran out. The market did the substituting, the price did the signaling and the problem solved itself. I will not claim that every case ends so neatly. I will say it is the usual pattern.

The best case for the doomsayers

Fairness requires a hearing for the other side. Here is its strongest form. One: oil is finite. There is a limited amount in the ground, and at some point, every field declines. Two: the easy oil is gone. The cheap, shallow, high-pressure fields have been tapped, and what remains costs more, takes more energy to extract and is often in harder places. Three: Hubbert's curve did describe the lower forty-eight production peak correctly, and there is no reason to think the logic does not apply globally in the long run. Four: even if the world has plenty of oil, it may not be enough at an affordable price.

I accept all four of those as matters of fact. The planet is finite. Fields do decline. The best oil is usually found first. But none of those facts supports the conclusion that the doomsayers draw, which is that we are about to hit a wall. The question that matters for a family or a country is not whether oil is finite in some final geological sense. The question is whether energy will be available at a reasonable price over the time that anyone can plan for. Over that time, history says that technology and price have been better guides than the depletion curve. And in the very long run, we will not run out of oil in the sense of a sudden stop. We will run out of reasons to burn it, the way we ran out of reasons to burn whale oil, because someone will find a better way to do the job.

What the failed forecasts cost

The reason I press this point is not to mock people who got it wrong. It is that forecasts of scarcity have consequences. They are used to justify policies: restrictions on drilling, mandates on consumption, subsidies for favored alternatives and the rationing of choice. A policy built on a prediction that is wrong is a policy that costs us in real terms. If the forecast says that supply is about to vanish, the government steps in to manage the decline. If supply instead grows, the government has spent political capital and money solving a problem that did not exist, and meanwhile it has restricted the very activity that would have produced the abundance.

The Strategic Petroleum Reserve is one result of the fear of the 1970s, and its history shows how much policy turns on how we think about supply. See From the 1973 Embargo to the 2022 Sale, a Strategic Petroleum Reserve Timeline for the story. And for how an energy that was supposed to be scarce became an export, see How America Went From Importing Natural Gas to Shipping It Worldwide, where the same shale technology flipped the picture on natural gas.

Also see Four Pieces of Every Gallon and Why the President Controls the Fewest for what a price at the pump consists of and how little of it a president controls, and The 1980 Bet on Five Metals That Settled the Great Scarcity Debate for the wager that tested the scarcity argument with real money.

The Founding angle

There is a deeper principle here, and it is one the Founders understood. They believed that free people, secure in their property and rewarded for their labor, produce more than any plan can anticipate. The Constitution's patent clause, in Article I, Section 8, gives Congress power "To promote the Progress of Science and useful Arts." That is a bet on human ingenuity written into the supreme law of the land. A society that bets on ingenuity, and gives inventors a reason to invent, tends to find a way. A society that bets on rationing tends to get rationing.

The history of oil fits that principle. The drillers who found the field, the engineers who learned to turn a well sideways and the entrepreneurs who risked money on a formation the majors had written off did not wait for a federal plan. They acted on a price, on a hunch and on the chance of a payoff. That is not a flaw in the system. It is the system.

The case in brief

1. Forecasts that America would run out of oil go back to the years around World War I, and each was overtaken by new discoveries, new technology or lower demand.
2. M. King Hubbert's 1956 paper came close in predicting that American production would peak between 1965 and 1970, but his model assumed a fixed recoverable stock, and technology changes what is recoverable.
3. Jimmy Carter's April 1977 address and the peak-oil books of the 2000s repeated the exhaustion thesis, and horizontal drilling and hydraulic fracturing then pushed American output past its 1970 record by 2018.
4. Proven reserves are an economic measure tied to price and technology, not a count of what is physically in the ground, so they grow when oil becomes cheaper to reach.
5. Policy built on scarcity forecasts costs real money, and the safer bet is on free people and prices.

What you can do: Look up the Energy Information Administration's annual crude oil production series on its website and see for yourself where the 1970 peak and the later records fall.