If resources are not fixed but created, then the nature of the scarcity problem changes dramatically. For the technological means involved in the use of resources determines their creation and therefore the extent of their scarcity. The nature of the scarcity is not outside the process (that is natural), but a condition of it. –Tom DeGregori (1987). “Resources Are Not; They Become: An Institutional Theory.” Journal of Economic Issues.
This is one of the two lead quotes from “Resourceship: Expanding ‘Depletable’ Resources” by Robert L. Bradley. It’s the Econlib Feature Article for May.
Bradley points out that this insight did not originate with the late Julian Simon. He quotes institutional economist Erich Zimmermann laying out the insight in 1933.
The article ends as follows:
Should we dismiss the term “exhaustible” to describe mineral resources, as energy economist M. A. Adelman has suggested? Doing so would be in the tradition of Ludwig von Mises, who stopped short of assigning any special economic meaning to mineral resources.
A methodological lesson emerges from this story. Sixty years ago, Friedrich A. Hayek expressed the following thought about the advancement of good economics: “It is probably no exaggeration to say that every important advance in economic theory during the last hundred years was a further step in the consistent application of subjectivism.” This insight can be particularly appreciated in reference to mineral economics. The “fixed” character of minerals has been a siren song to economists who saw less when there was really more. The mineral economist should never forget that what resources come from the ground ultimately depend on the resources in the mind.
READER COMMENTS
Ryan P
May 7 2012 at 11:12am
These are reasonable points, but I’ve never really been clear how they fundamentally unwind Hotelling’s rule, as opposed to complicating it. Yes, we can discover more resources, and yes, we can discover new ideas that turn previously useless dreck into valuable commodities. But we can also expect those things to happen, and I don’t see why our expectations about future potential are more likely to undershoot than overshoot. So isn’t that more or less the same as saying “the quantity of resources is very large, very uncertain, and clearly finite”? At which point you’d think there should be a positive (if sometimes small) “exhaustible resource rent” and that rent should grow over time for no-arbitrage reasons.
Rob Bradley
May 7 2012 at 11:38am
Ryan P:
Hotelling’s analysis is not ‘wrong’ in a technical sense. What is ‘wrong’ is that Hotelling frames his seminal article in real-world terms with his conclusion being that increased scarcity of minerals has a special depletion value (‘user cost’). (Mises disagreed as I mention in my piece.)
If only Hotelling had re-framed his article as a special case. In fact, his analysis could apply to drinkable water in the fixed desert situation and not ‘minerals’ at all ….
I would bet that Hotelling would be amazed at the post-consumption growth of minerals and would quickly say today, ‘Yes, mine is a special case of an economic good that does not have to apply to minerals.’
In an email to me, Milton Friedman said that Hotelling would conclude that oil is not a depletable resource (http://www.masterresource.org/2010/07/milton-friedman-on-mineral-resources-2010/)
I quote Friedman in defense of Hotelling’s 1932 piece:
“If you use a tool that is designed well for one purpose for a purpose for which it is not suited, that does not detract from the goodness of the tool for its purpose. The same thing goes with economics. Hotelling’s analysis is good economics. If it is applied properly to the case of oil it produces the right result, namely that oil is not as an economic matter an exhaustible resource. That result follows from the Hotelling demonstration that an exhaustible resource will have a price that is rising over time.
Since the price of oil was not declining over time, it is not economically an exhaustible resource.”
Tim Worstall
May 7 2012 at 11:55am
My current favourite example of this resources are limited thing is tellurium.
It is one of the rarest elements on the planet. We currently get it as a by-product of mining certain copper ores. Supply is limited and we are using more of it (for thin film solar cells). We’ve most certainly had a European Union report demanding that we must do much more to recycle it.
The thing is, there really is a limit to the number of Te atoms on the planet. Substitution aside (I don’t want to go there for this argument) that number is indeed the hard limit to the amount of Te that we can use.
However, that limit is some 120 million tonnes in the crust of the Earth. Current annual global usage is around 125 tonnes a year.
OK, so there’s a hard limit to the amount of Te we can use. But it’s an entirely irrelevant one: the real limit we face now is economic and technological. What processes do we have to extract it from the crust at what price?
Agreed, I’ve not gone and tested this against all and every one of the metals but I’d still put forward this thought. Even if we do take the environmental limists seriously, the simple availability of the number of atoms, we’re still nowehere near any of those limits. We’re still in the world where “resource” is an economic concept and thus something that is created by advancing technology.
Rob Bradley
May 7 2012 at 12:29pm
Ryan:
Hotelling’s seminal article is technically correct in almost all of its dimensions, but he mistakenly framed it as real-world and applying to petroleum in particular. He made an ‘institutional error’ and did not realize the limits to his analysis out the window, so to speak.
His was a special case and really did not apply to minerals per se. It applies to a fixed supply under a set of assumptions, perfect knowledge in particular. So it really applies to a situation of limited drinkable water in a desert, not global resources.
I asked Milton Friedman about this. Here is what he said (http://www.masterresource.org/2010/07/milton-friedman-on-mineral-resources-2010/):
“If you use a tool that is designed well for one purpose for a purpose for which it is not suited, that does not detract from the goodness of the tool for its purpose. The same thing goes with economics. Hotelling’s analysis is good economics. If it is applied properly to the case of oil it produces the right result, namely that oil is not as an economic matter an exhaustible resource. That result follows from the Hotelling demonstration that an exhaustible resource will have a price that is rising over time.
Since the price of oil was not declining over time, it is not economically an exhaustible resource.”
David R. Henderson
May 7 2012 at 1:12pm
Here’s the hot link to the Friedman correspondence that Rob Bradley referenced in his comment above.
Becky Hargrove
May 7 2012 at 3:23pm
Wonderful post David. I’ve already got a full page of notes and have yet to read the main essay.
Ryan P
May 7 2012 at 4:44pm
Rob Bradley,
I agree that the original model assumed perfect information about the size of the stock. But I’m still unclear how the result completely reverses when you add uncertainty. I say this because uncertainty very much has been added to the theoretical model (see Pindyk, say) and that alone didn’t overturn the qualitative result. Similarly, it’s quite common to assume in the model that if the resource price hits a certain threshold, we can find infinite quantities of a perfect substitute. This too doesn’t change the Hotelling rule fundamentally.
So what is supposed to do this? Expectations not being rational (i.e., the market consistently underestimates how much will be discovered or invented in the future)? Literally unlimited resources? Failure of no-arbitrage? Something else?
Ryan P
May 7 2012 at 4:53pm
I want to make clear that I’m not here claiming that we should expect the Hotelling price path to assert itself any day now as resource rent finally trumps the secular decline in extraction costs of everythin. If you like, you can see my question as more in the tradition of “that may be true in fact, but how is it in theory?” 🙂
David R. Henderson
May 7 2012 at 6:40pm
@Ryan P,
If you like, you can see my question as more in the tradition of “that may be true in fact, but how is it in theory?” 🙂
Interesting question. And the next thing I’m about to say is something I’m not so sure of: I think the theoretical answer is that there’s such uncertainty that the price path is random in the long run.
Mark Bahner
May 7 2012 at 10:55pm
Well, let’s look at oil. Fossil oil can be used to power cars. That’s finite.
But oil can be made from algae:
http://articles.cnn.com/2008-04-01/tech/algae.oil_1_algae-research-fossil-fuels-nrel?_s=PM:TECH
Oil from algae is essentially infinite (i.e., renewable and capable of being produced at much greater quantities than oil is even being consumed today). That’s because the sun produces algae, and it can be produced in the ocean, or in any desert with some imported ocean water.
So “oil” is essentially infinite. Not fossil oil. But “oil.”
Now, you could say, “Oh, well, there are tremendous technical problems, and oil from algae is much more expensive than fossil oil.”
And you’d be correct about the year 2012. But you wouldn’t necessarily be correct about the year 2022 or 2032.
And that doesn’t even take into account the possibility of battery-powered cars essentially eliminating the need for oil.
Mark Bahner
May 8 2012 at 1:04pm
Hi Tim,
But there isn’t even that “hard limit,” because your limit is based only on tellurium in Earth’s crust.
Just recently, the company “Planetary Resources” was formed with the goal of mining asteroids, using robots.
http://www.planetaryresources.com/category/faq/
Just to give a flavor for the potential, a single 1 km diameter asteroid has precious metals worth an estimated $5 trillion at today’s market prices. (Note that, if brought to earth, such vast amounts of precious metals are expected to collapse precious metals prices, except gold, so that the collapsed value would be “only” $320 billion.)
http://www.agu.org/pubs/crossref/1994/94JE02141.shtml
P.S. Harold Hotelling, in 1931, could certainly be forgiven for ignoring such potentialities. In fact, if he’d considered them, he wouldn’t have been doing science, he would have been doing science fiction.
Jim Glass
May 8 2012 at 4:05pm
Perhaps relevant, regarding the recent “Physicist v Economist” thing that got some attention about the limits of energy consumption.
Ryan P
May 8 2012 at 5:25pm
Mark Bahner,
Isn’t that equivalent to the “backstop technology” version of the Hotelling/Pindyk model? Or alternately, if you think the tech kicks in at a certain year rather than a certain price (and moreover, no one knew that until today), you’d expect that would show up as a one-time shock to oil prices today, followed by the no-arbitrage-induced increase in fossil fuel rents at the rate of interest (which would asymptotically be identical to prices increasing at the same rate).
In other words, that just tells us that we should use up all the stocks of cheap fossil fuels rather more quickly than the market otherwise intended.
Mark Bahner
May 9 2012 at 9:54pm
“Mark Bahner,
Isn’t that equivalent to the “backstop technology” version of the Hotelling/Pindyk model?”
I didn’t even know there was a Harold Hotelling until I saw this post and the comments. But I do know that, while *fossil* oil may be exhaustible, oil from sources such as algae certainly isn’t. And metals aren’t either.
Comments are closed.