The Two Worlds of Oil

in 1980 most experts agreed that oil prices could only go up. Following the panic of the Iranian revolution, the price spiked to more than $80 a barrel adjusted for inflation. I gained some notoriety at the time by publishing an article with William Brown, a Hudson Institute colleague, in the Wall Street Journal predicting that oil prices would fall in 1980 and that the 1980s would be a decade of decreasing, not increasing, oil prices. Indeed, the price fell sharply in 1980 and by the late 1980s the price had fallen to around $30 a barrel, and it dipped to around $20 in the late 90s.

Today there is a great chorus, in which New York Times columnist Thomas Friedman's voice stands out, calling upon the United States and other nations to radically reduce their oil consumption because, as Friedman and others contend, the world will soon run out. In the mean time, they say, our continued dependence on (or "addiction to") oil means a continued dependence on oil exporting countries--so many of which are run by less than democratic governments--and high energy prices of $70 a barrel, or more.

But a deeper understanding of the supply side and a longer term perspective of demand produces a different view. Between now and the middle of the century $30 is likely to be more typical of the price of a barrel of oil than $60. Most of the time sellers will be competing for buyers, not pushing them around. And the Arabs are likely to have a smaller share of the market in the future, not a larger one. Before long the fear of Arab oil power is likely to seem unimaginably dated.

Two factors influence oil prices. First is the amount of oil in the ground. Second is the capacity of oil production and transportation facilities. Too few wells and pipelines create oil shortages, and therefore high prices, regardless of how much oil there is in the ground. For our purposes, "oil in the ground" refers to oil that investors think they can bring to market for less than $20 a barrel if things go near enough to plan. Capacity, on the other hand, refers to every element of the process from extraction to delivery, including the production of equipment associated with each element of that process.

So how much oil is out there waiting to be discovered? Chevron Corporation has been buying advertisements claiming that, "The world consumes two barrels of oil for every barrel discovered." Fortunately Chevron is only speaking the truth if you use an artificial definition of how much oil is being discovered. For example, Canada is now estimated to have 150 billion barrels of recoverable oil in their tarsands. Twenty years ago we couldn't produce that oil at competitive costs. Now we are producing a million barrels/day at a cost of about $15 each. In effect we have "discovered" 150 billion barrels of oil in Canada--more than the entire world used in the last five years--which Chevron doesn't count.

There are other potentially large additions to world oil resources. According to Leigh Price, an expert on oil deposits who died in 2000, there are over 200 billion barrels in the Bakken "oil play" in Montana and North Dakota. It is not yet clear whether Price's theory is correct, but recently a number of drillers have started producing oil from that deposit, which is not even included in standard estimates of oil reserves. Bakken might not turn out to be as significant as Price expected, but the larger point remains: A number of potential sources exist. Recently an Israeli claimed to have discovered how to economically produce oil from oil shale. Since such claims have been made for more than twenty years, we can't put too much faith in this announcement. But there is no reason to assume that we won't solve this problem in the future. There are hundreds of billions of barrels of shale oil in Colorado and probably as much in other parts of the world.

There are two ways to discover oil. One is the conventional way--drilling to find new deposits of the same kind as we produced before. The other is to develop new technologies that either allow us to retrieve oil from formerly out of reach places, or economically recover oil from deposits where extraction is currently too expensive.

The net result is that the known available oil supply in the ground today is larger than it was 20 years ago, or 50 years ago, despite all the oil we have taken out of the ground and burned. Despite this experience, the oil supply in the ground 20 years from now may not be larger than it is today. No one can say either way with any certainty. Still, today, as in the past, oil investors find myriad opportunities to seek oil that is expected to be delivered at a cost of $20 a barrel or less. Therefore we can be relatively certain that present high prices are not the result of a lack of oil in the ground.

Since lack of oil in the ground isn't to blame for current high prices, these prices must be a product of insufficient production equipment. There is no shortage of iron or other raw materials used to build drilling rigs, pipelines, and other necessary equipment, so we can produce as much oil production equipment as we want. Why don't we have enough of it? The answer is easy: It's expensive and the manufacturing process is long.

Operators don't buy oil production equipment if they don't think they'll need it. If producers buy more production capacity than they need, supply becomes too high, the price goes down, and producers aren't able to sell as much as they anticipated. The problem is predicting demand 5 or 10 years ahead of time. High prices are the result of producers underestimating the need for oil; low prices are the result of producers overestimating the need for oil. And the difficulty of estimating the future demand is shown by the wild fluctuation of prices.

This is where all the talk about excess demand from China and the "failure to conserve" comes into the picture. Some critics say we don't have enough oil because demand has been rising too rapidly. Global demand has been rising, but only about 2.5 percent per year including China, which uses only 8 percent of the world's oil. What's more, nothing prevented producers from increasing oil production facilities faster than the demand for oil increased.

Even if demand doesn't increase at all, new wells must be drilled to replace older wells with decreasing production. Therefore an oil shortage can occur with no increase in demand. The world now has high oil prices using about 84 million barrels per day. We could reduce our use to 60 million barrels a day and again have high prices if a lack of investment failed to maintain production capacity. Or, as I predict, a few years from now we will have ample production capacity and lower prices at a consumption rate of 90 million barrels a day.

No matter how much oil is consumed--whether the amount is more or less than today--we can have high prices and shortages if there is not enough production equipment. Or, if enough equipment has been produced, we can have a favorable balance of supply and demand and moderate or low oil prices. This will be true until there is not enough oil in the ground--but there is no hard evidence of such a shortage, only theories.

Let's put all this together in a scenario for the future. Here is my prediction of what might very well happen, although many other scenarios are also possible:

The basic price of oil for the next 50 years will be about $30 a barrel. Some of the time it will be higher, but I would bet that a lot of the time it will be lower. The key point is that any investments made in oil or oil substitutes that depend on oil prices staying well above $30 a barrel stand a good chance of losing money. They are imprudent, risky investments--although nobody can say for sure that they won't pay off.

In the long run price depends on the balance between demand and the amount of oil in the ground and the cost of getting it out. My guess about flat average oil prices--lower than now, but higher than average prices over the past 70 years--is based on another guess: that consumption will grow by a couple of percentage points a year until about 2025, when autos that burn hydrogen made from natural gas will begin to significantly reduce oil consumption for transportation. Demand growth will then slow, and gradually stop. Beginning around the middle of this century, world population will begin slowly declining. Even though gross world product will keep on rising--because we will get richer faster than population declines--oil consumption will be close to flat. In brief, my guess is that mid-century oil production will be between 120 and 150 million barrels per day with a price of approximately $30 per barrel.

To some extent, this conclusion is based on the fact that the price of almost all minerals taken from the ground has consistently decreased--despite increased demand. The best bet is that at the end of the 21st century we will pay much less for minerals and for raw agricultural products like wheat, corn, and rice than we do today. But oil may prove the exception.

On the basis of historical experience and the broadest perspective of geology and technology, my projections represent a plausible scenario. There are other plausible scenarios, too. But there is no reason to believe those who say we are doomed to oil shortages and high prices; or that oil will certainly become more and more scarce; or that, in the long run, it is foolish and dangerous to rely on oil; and that we must take drastic measures to change our use of energy. Any measures we take to reduce oil consumption in the next few years should be based on the assumption that such measures may not be needed over the long term. With that in mind, our first priority should be finding better solutions to the short-term problems, such as increased storage.

This article originally appeared September 27 in the Weekly Standard online.