David R. Henderson  

The Jevons Fallacy

The CBO's forecast, with and w... Are Competition and Cooperatio...

Noah Smith writes:

And while our use of natural gas and coal doesn't feed the coffers of unsavory regimes like Russia and Saudi Arabia the way our use of oil does, it's still the case that these energy sources are limited. They run out.

Of course, all resources are limited. Even solar energy, which Smith is optimistic about, is limited. That's not because the sun is, in any reasonable sense, limited, but because all the expensive resources used to convert the sun's rays into something useful are limited. In his recent book, The Moral Case for Fossil Fuels, which Bryan Caplan has reviewed here and which I'm reviewing for another publication, Alex Epstein makes that case beautifully.

But I want to make another point: natural gas, coal, and oil are unlikely to run out any time soon, where soon means any time in the next 60 years.

Last century, William Stanley Jevons, the famous economist who was in part responsible for the marginal revolution, reasoned that because the easy-to-get coal was already gotten, what was left was coal that was harder and harder to reach. Therefore, he argued, we would reach a point where coal would be much scarcer and much more expensive.

As I wrote in my bio of Jevons, here was his error:

Jevons failed to appreciate the fact that as the price of an energy source rises, entrepreneurs have a strong incentive to invent, develop, and produce alternate sources. In particular, he did not anticipate oil or natural gas. Also, he did not take account of the incentive, as the price of coal rose, to use it more efficiently or to develop technology that brought down the cost of discovering and mining (see natural resources).

And from the "Natural Resources" article cited in the Jevons bio, by the late Sue Anne Batey Blackman and William J. Baumol:
Innovation has increased the productivity of natural resources (e.g., increasing the gasoline mileage of cars). Innovation also increases the recycling of resources and reduces waste in their extraction and processing. And innovation affects the prospective output of natural resources (e.g., the coal still underneath the ground). If a scientific breakthrough in a given year increases the prospective output of the unused stocks of a resource by an amount greater than the reduction (via resources actually used up) in that year, then, in terms of human economic welfare, the stock of that resource will be larger at the end of the year than at the beginning. Of course, the remaining physical amount of the resource must continually decline, but it need never be exhausted completely, and its effective quantity can rise for the indefinite future. The exhaustion of a particular resource, though not impossible, is also not inevitable.

With respect to oil, here's how I put it in a Sidebar to the Natural Resources article in the first edition of The Concise Encyclopedia of Economics:
Are We Running Out of Oil?
No resource has inspired so great a fear of running out as oil has. This fear is not new. And every time in the last hundred years that an expert predicted we would run out, that prediction has been wrong--and not just wrong, but wrong by a huge margin. In 1891, for example, the U.S. Geological Survey stated that there was little chance that oil would be found in Kansas or Texas. Since then, 14 billion barrels of oil have been produced from just those two states. In 1914 an official of the U.S. Bureau of Mines claimed that the total future U.S. production would be 5.7 billion barrels. In fact, production has already been six times that figure. In 1920 the director of the Geological Survey said that peak annual crude production had almost been reached. By 1948 annual U.S. production was four times its 1920 level. Finally, in one of the most astoundingly wrong predictions, the Interior Department stated in 1939 that U.S. oil supplies would run out in thirteen years. Of course, over sixty years later, the United States is still producing oil.

Source: Baumol, Blackman, and Wolff, p. 214.

Or as I have put it in various talks I've given over the years:
Every single time someone has predicted that they we will run out of oil, that person has been wrong. What reason is there to think that prediction today will be right?

Comments and Sharing

COMMENTS (13 to date)
Rick Hull writes:

Can't these claims can be "weakened" to a defensible ceteris paribus analysis of the one good in question? This would not be a solid basis for making real world predictions of course. But doesn't this cleave rightness from wrongness pretty well? I think it highlights that while violations of ceteris paribus have "saved" us in the past, it's a wildly unpredictable horse to hitch our wagons to, going forward.

AS writes:

Anyone who claims resources will run out and preach government intervention to "conserve" those resources, reveals their own deep ignorance of how markets work and how prices efficiently ration scarce resources.

LD Bottorff writes:

I'll hitch my wagon to the petroleum fueled engine for now. Even when the costs of solar cells comes down, as it inevitably will, the issues of power distribution still need to be addressed. Then, we will need to address the issue of electric transportation. All of these issues may be resolvable, but it sure is difficult to see why when the cost of oil is low and the global temperature has only gone up .04 degrees since 2010.

I don't think Jevons makes the errors you attribute to him, or at least if he does sometimes fall into the error he doesn't always fall into that error.

He does, for example, consider alternatives to coal in The Coal Question--including brief discussions of sun, wind, water, wood, peat, and petroleum--but finds they are inadequate for the task of replacing coal. But recall that his focus was on British economic advantage. It wasn't so much that he dismissed these alternatives altogether as much as just didn't see them as likely to continue Britain's advantage (the U.S. had better wood resources, many places had better sun, and so on).

In addition, as you note for Jevons, he pointed out increasing cost rather than physical exhaustion as the issue of concern. This observation was a useful addition to the mid-19th century debate on the future of coal. Noah Smith, overlooking 150+ year old economic insights, thinks the problem with oil is that it will run out.

Jevons did see innovation as an important issue in the debates over coal, but most famously in looking at the relationship between steam engine efficiency and the demand for coal. I don't recall whether he turns to consider the relation between efficiency in production and the relevant size of the resource stock, but I'd be surprised if he didn't make at least some remarks on that issue as well.

Ray Lopez writes:

Good article, but the assumption is that science will find an alternative supply and the AS curve will shift or be replaced by a new supply curve. This is not necessarily true in the intermediate term if you believe in the Great Stagnation thesis. Hence, as a precaution perhaps conservation is wise? That said, solar energy panels for generating electricity have indeed become more competitive, but likely the recent drop in oil prices, if they continue indefinitely (as the Saudis et al want) will kill off green energy. So perhaps an energy surtax for conservation purposes is warranted sometime down the road?

AS writes:

@Ray: The incentive to conserve resources is already internalized by the buyer and seller. Prices clear markets to obtain pareto optimality. Additional governmental surtaxes are unnecessary.

Mark Bahner writes:


A very interesting subject to me. :-)

I think I'll try to put up some projections for 30 and 60 years from now sometime in the next week or two, but my general personal thoughts are:

1) Nuclear does seem to be dead in the U.S. Looking 30 years out, most of the reactors currently operating should be shut down (all but about 10 reactors will have been operating 60+ years):

Nuclear reactor license expirations

It's hard to see enough reactors (~100) to match retiring reactors to be built in the next 30 years.

2) Coal use in the U.S. is presently lower than it was in 2007-2008. I think 2007-2008 will end up being "peak coal".

3) Due to items 1 and 2, natural gas looks to expand significantly...at least 30 years out. I'd need to think about 60 years.

4) Oil is going to be significantly displaced by electricity for automobiles within 30 years. (Computer-driven cars greatly favor batteries as a power source.) I'd expect conventional (and fracking) oil consumption to be down by more than 50 percent 30 years out. It's possible some sort of inexpensive bio-oil could be developed.

5) Wind and especially photovoltaics look promising. Photovoltaics are tremendously good fit for the Southwest (Los Angeles, Las Vegas, Phoenix, etc.)

5) Looking out 60 years is extremely difficult. Virtually impossible. (But that probably won't stop me from trying. ;-))

vikingvista writes:

"Run out" is such an odd claim. What marketable item has ever run out? Marketability *prevents* things from running out even in the unlikely case that they might.

One might claim declining use--peak this or peak that. But even then, how often is a seemingly permanent peak associated with mass privation rather than changing preferences in an increasingly prosperous world? Peak oil may occur one day, but it will be about as traumatic as peak horse.

Mark Bahner writes:
What marketable item has ever run out?

Passenger pigeons?**

Seriously...I think 2007-2008 was "peak coal" in the U.S. And I think that in 60 years, we'll be using virtually no coal. But I think you're absolutely right...it won't be because we "run out" of coal. As the saying goes, "The Stone Age didn't end because we ran out of stones."

And the age of coal and oil won't end because we "run out." It will be because natural gas, wind, photovoltaics...and potentially liquid fluoride thorium reactors...become better options.

**P.S. Actually, I predict passenger pigeons will fly again in the 21st century, due to cloning.

vikingvista writes:


I wasn't aware passenger pigeons were marketable. Do you have any information on their price history?

Mark Bahner writes:
I wasn't aware passenger pigeons were marketable.

I wasn't aware anything wasn't marketable. :-)

Pigeons made of gold

At first some farmers would simply take their pigeons to town and sell them for 10 cents a dozen. But others began using dealers to ship dressed pigeons. In Chicago they sold for 50 to 60 cents a dozen. Baby pigeons called "squabs" fetched even more. Pigeons were shipped to the southern U.S. to feed slaves. There was a burgeoning market for live pigeons to be used in shooting contests. But the big market was New York City and Boston, where pigeons fetched two dollars a dozen by the late 1870s.

Two whole dollars?!!! Well, gollllly! Like they was made a gold! :-)

vikingvista writes:

Cool. They really must've been hypersensitive to decreased flock size afterall. I always doubted the claim that they couldn't survive in small numbers.

"I wasn't aware anything wasn't marketable."

I've been trying to sell my pocket lint, fecal popsicles, and air balls for years. It turns out however, a market can't be created in just anything.

Sldyen Donnelly writes:

Interesting post and comments. But when I look at the statistics over the long term, it does not seem so much like oil has displaced coal (or gas is displacing oil). It looks like innovation results in the addition of energy options to, not the subtraction of options from, our supply mix. It's not like we burned less coal to make power, globally, in 2010 than we did in 1910 or 1810, is it?

Does our economic analysis, perhaps, too often confuse "complements" with "substitutes"?

Comments for this entry have been closed
Return to top