Friday, March 10, 2006

Peak Oil?

There has been a good deal of talk recently about "peak oil," the idea that world oil production has reached its peak and is going to be declining in the near future, resulting in shortages, skyrocketing prices, and similar unfortunate consequences. The phrase may be new but the idea is not. We have been being told that the world is about to run out of oil for some decades now, and those predictions, along with more general predictions about running out of depletable resources—some going back more than a century—have so far consistently proved false.

That is a reason for scepticism, but not a proof that current claims are false. Most of the arguments depend either on estimates of how much oil there is and what it costs to get at it or on estimates of the cost of alternatives, such as tar sands, liquified coal, solar or nuclear power. Since I am neither a geologist nor an engineer, I prefer to look at what economics can tell us about the situation.

The economics of depletable resources was worked out by Harold Hotelling more than seventy years ago, although outside of the economics profession almost nobody seems familiar with it. The argument is straightforward. Owners of oil underground can choose when to pump and sell it. If the price of oil is rising fast enough so that oil in the ground pays a higher return than money above ground, it pays to leave the oil in the ground—postpone production in order to get a higher price in the future. That reduces present supply, shifting the present price up, increases future supply, shifting the future price down. In a world of secure property rights and perfect information, the process continues until the projected price of oil, net of pumping costs, is rising at exactly the market interest rate, forever. Any faster than that and people shift production to later dates, any slower and they shift it to earlier dates. Unless the people who control the oil and decide when to pump it are wildly off in their predictions of future prices—the theoretical analysis assumed perfect information—the usual crisis scenarios can't happen.

There is, however, a second critical assumption—secure property rights. Suppose I own underground oil, but I believe there is a substantial chance, say ten percent each year, that someone else will seize control over it. I will only leave the oil in the ground if the expected rise in oil prices is enough to compensate me not only for the interest I could have earned on the money I would get by selling the oil now but also for the risk of losing the oil. So insecure property rights result in producing more oil now, less later, and a price pattern that rises faster than in the Hotelling model.

Essentially all property rights in underground oil are insecure. It has surely occurred to the current rulers of Kuwait and Saudi Arabia that money in a Swiss bank account is a safer asset than oil under the desert. The government of Norway is unlikely to fall to a coup or an invasion—but the politicians who control it today cannot be confident of controlling it ten years from now, so have an incentive to pump now and use the money to maintain their political power.

In some countries, such as the U.S., much of the oil is owned by private firms, not governments. But their property rights too are insecure. As we have seen in the past, a rising price of oil results in political pressure for price controls, "excess profits" taxes, and other forms of more or less disguised partial expropriation.

The implication is straightforward. The arguments about oil geology and the cost of alternatives may or may not be correct—on the basis of past evidence, the claim that we will shortly run out of oil should be viewed with considerable scepticism. But the economic argument implies that owners of underground oil will tend to pump and sell earlier than they would in a perfectly functioning market, and hence that oil prices will raise faster than the simple version of the economic argument predicts. How much faster depends on how insecure the relevant property rights are.

There is one further complication whose analysis I leave as an exercise for the readers. Insecure property rights have a second effect—they make finding oil less profitable, since after you find it someone else may steal it.


Anonymous said...

I'll take a guess: if insecure property rights make exploration and discovery less profitable, it will tend to be undertaken at a lower rate than it otherwise would have been. Thus, we will know about fewer oil fields than we otherwise would have, making it seem as though there are fewer remaining resources than *might* (though not certainly) be there.

Anonymous said...

"But the economic argument implies that owners of underground oil will tend to pump and sell earlier than they would in a perfectly functioning market, and hence that oil prices will raise faster than the simple version of the economic argument predicts".

Shouldn't increase in production result in lower prices?

Anonymous said...

This is an interesting perspective. But I would say it is, on the balance, immaterial, because large energy companies have enough clout with government to capture at least as much oil property as they lose to distortion of property rights.

In fact, I'd say the net effect is likely the transfer of oil wealth from small companies to large ones, with no net change in the consumer market or depletion figures.

Glen Whitman said...

I wonder about the theory of the second-best here. As you've pointed out in this blog, negative and positive externalities can offset each other. Likewise, insecure property rights and oligopolistic market structure can offset each other. Insecure property rights give oil producers an incentive to pump more barrels now; but collusion (via OPEC) and imperfect competitition point toward pumping fewer barrels now. Hard to say which effect is more important.

Anonymous said...

Your point about oil production balancing against prevailing interest rates is well-taken. However, in our present situation we don't have free-market interest rates. We have manipulative central banks that have been pumping liquidity into the system. This produces an incentive to pump and consume oil reserves faster.

David Friedman said...

Anonymous asks:

"Shouldn't increase in production result in lower prices?"


The point is that insecure property rights mean that prices are increasing faster over time than they otherwise would. Given the amount of oil at any instant, insecure property rights result in more production and lower prices now, less production and higher prices in the future.

Anonymous said...

Oil doesn't have to be stored underground, although that's the cheapest way. The oil can be pumped today and held by speculators with secure property rights for the future (if prices are expected to rise at a rate that makes it profitable).

Anonymous said...

Whilst the economics are important, it seems that most people discussing 'peak oil' are referring to the fact that as an oil field gets depleted, the rate at which oil can be pumped out of it decreases.

Therefore, if the world stops finding new reserves, the old reserves will slow down in production long before they run out.

David Friedman said...

Edgr writes:

"Whilst the economics are important, it seems that most people discussing 'peak oil' are referring to the fact that as an oil field gets depleted, the rate at which oil can be pumped out of it decreases."

On the other hand, new fields get discovered, and improvements in the technology of getting oil out may increase the amount you can get from existing fields. Your point, after all, has always been true--but total world production of oil has been going up, not down.

Given that there are forces leading to both increased and decreased production, how can you or I decide what the net effect will be?

One way is to try to analyze all of those forces, estimate future finds, the effect of improved production technology or higher oil prices on output from existing fields, the effect of substitutes on demand, and try to combine all that to decide the net effect. In practice, very few people are competent to do that, so we end up simply believing whatever expert source produces the result we want to believe.

The alternative is to use other people's behavior as evidence. People whose income depends on estimating future oil prices are much more likely to get the answer right than you or I. Market prices give us information about what such people believe. If, for instance, we were going to run drastically short of oil ten years from now, prices would already be sky high--because people would be holding oil now to sell it then.

Hence the usefulness of the economic analysis to figure out what we can or cannot deduce about the answers to geological and technological questions by the observed behavior of those who know more about the subject than we do.

Kantor said...

"So insecure property rights result in producing more oil now, less later, and a price pattern that rises faster than in the Hotelling model."

Not only insecure property rigths: also agency problems of mutual fund and oil company managers. If they are focused on short run returns, the financial mechanism David describes won´t work

If the market values more quarterly profits over proven reserves, you will have a closer peak oil with an steeper slope down.

Kantor said...

"However, in our present situation we don't have free-market interest rates"

Yes, we ALWAYS have free market interest rates. I mean real interest rates, of course.

That´s something Mr. Milton Friedman and Mr. Robert Lucas proved a long time ago.

"Accelerationist hypothesis and Rational Expectations..."

Rob Hayward said...

This makes a lot of sense, apart from the fact that the Norwegian government has the incentive to use the petroleum fund (the collection of overseas assets that it has accumulated with oil revenues) to finance government while leaving oil in the ground (if it really believes and if Norwegian oil experts believe that oil will be worth more in the ground than in the market today.

Unknown said...
This comment has been removed by a blog administrator.
montestruc said...

I’d like to amplify David’s point about the effects that insecure property rights have on mineral development. Allow me to point out the fact that the extremely insecure (non-existent) property rights of oil fields at sea outside the territorial waters of any nation mean that no such field has ever been developed, as far as I know. All developed offshore oilfields are within the territorial waters of some nation. As a rule the nation in question is paid a substantial cut of the income from the oil which gives that nation a very strong incentive to defend the facilities that produce the oil.

Please to note that to develop an oilfield for production in deep, or even moderately deep water at sea is extremely expensive. Shallow water is less expensive but more than the means of most moderatly rich persons, and more than 200 nautical miles from land almost no shallow water exists.

No one in the oil industry is going to put a $100 million dollar plus drill ship to work drilling wells in a field that might be grabbed by the first party with something that passes for a navy that wants it. Oil firms are not going to risk their investments unless such a pirate is going to be risking serious military repercussions by doing so. That takes a government agreeing to protect them, and nations could goto war if one nation claims rights well beyond it's territorial waters.

As an example note that even though the waters bordering China and Vietnam are well known to have significant oil deposits, but because the ownership of the mineral rights is disputed by the two governments major oil companies have refused to get directly involved in drilling operations being done by the Chinese (they do consult, but not invest), as they judge the risk of war too high. I tend to think it might happen, but that if it does, it will be short and Vietnam will lose after having destroyed or inflicted heavy damage to Chinese government owned oil rigs in that area.

The fact that property rights to minerals outside of territorial waters is essentially not defined legally, means no one owns them, and what no one owns, cannot be legally stolen, regardless of the fact that you developed it. So property rights are non-existent.

I also point this out with regard to “Peak Oil” about 7/10ths of the earth’s surface is ocean, perhaps 10% of that is within 200 nautical miles of a nation (my guesstimate), much of even that small portion of the world’s oceans is as yet unexplored for oil. People are still finding oil on land for crying out loud. This leaves something like 60% of the earth’s surface utterly unexplored for oil, and most geologic models now indicate that oil is mostly formed under old sea bottoms. While it is true that it is more expensive to recover oil from a deep well far out at sea, it is a moneymaking proposition at current market rates. Ergo dire predictions of collapse of civilization in the next few years from oil starvation are, shall we say, overblown, as these rights will, I am sure, become defined in the not too distant future.

Unknown said...

Insecure property rights also have another effect: reduced investment. If there's a 10% chance that I'll lose my oil field, I may want to pump all I can, but I'm not going make long-term investments in capacity expansion. This results in less oil today, and leaves more in the ground for later. So while prices may rise faster when property rights are less secure, the actual amount of long-term depletion is lower. If property rights are then strengthened sometime in the future and investment returns, prices would then be lower than if property rights had been secure all along.... [More]

Anonymous said...

Really Professor, I am surprised. You talk about Hotelling’s model for natural resources, but oil is different and in many way more complex than other natural resources. The best economist on oil specifically was M.A. Adelman. Though he never gave a date for ‘peak oil’ per-say, he did discuss at length and very convincingly the role of production cost as the real barometer for gauging the time of peak oil production. From a geological stand point it seems far fetched that the earth is now looking at the top of the oil production curve considering that many sources, sea bed, shale, preserves, have not even been touched.

What is reasonable is that the lowest-cost sources of oil production are entering their peak and decline period. Its known that the second largest oil field in the world, located in Kuwait, has peaked. For the largest consumer the US, most of our domestic fields peaked in the late 1970’s. Although, the fastest growing oil consumer, China, is placed strategically close to Russian and Iranian oil fields, all of them lag behind in refinement capacity. Large questions loom over what exactly the KSA has under the sand, but its recent decision from OPEC meetings would suggest that their production levels are at or very near full trying to save off any further increases in oil prices.

The end point is this, peak oil happens as soon as the price exceeds the next best option. This situation happens very readily when the lowest cost sources of production peak, it does not necessarily have occur as a peak in total oil production.

russ said...

Thank you for bringing this up. The United States has 3% of the world's total oil reserves yet it produces something near 10% of the world's oil. Property rights is the primary reason the US is so efficient compared to other countries, in oil production.

Anonymous said...

It's always a good idea to examine the assumptions, both stated and unstated.

One stated assumption: " a world of perfect information..." We do not live in such a world. We do not even really know to what degree our information is imperfect, so we can't estimate how this departure might affect the model.

Unstated assumption: Actual resources are infinite. By focusing only on the economics, this model fails to even consider the possibility that the resource itself could become rate-limiting.

Peak oil theory is very well-supported by sound science. M. King Hubbert predicted, in 1965, that US production would peak around 1970 - which it did. The same science applies regardless of scale.

David Friedman said...

Tim Bray writes:

"Unstated assumption: Actual resources are infinite."

That assumption was not stated because it is not part of the argument--Hotelling's analysis applies just fine to finite resources. The price rises at the interest rate, and reaches the price at which quantity demanded is zero just as the last drop of oil is pumped.