Hicks, Hayek, Hotelling, Hubbert, and Hysteria or Energy, Exhaustion, Environmentalism, and Etatism in the 21st Century
1. INTRODUCTION
Having the rare honor of a second IAEE award, I chose to respond by reacting to the present deplorable state of political discourse on energy by revisiting the issues that I treated in my paper (Gordon 1993) for the prior award. The theme here is that while the underlying theory and reality are unchanged since 1993, the politics have degenerated drastically. 1993 in retrospect was the middle of an un- usually favorable political climate. At least since September 11, 2001, politicians have resumed their usual ineptitude about intervention, a problem that extends far beyond energy. Several efforts from purportedly dispassionate study groups have supported these errors. To make matters worse, some have argued with total disregard of economic reality that President Garter's disastrous policies should have continued to prevent the present problems. These errors need response, and this article tries to provide such a critique.
In energy, error has involved reviving and ridiculously extending Nixon's idea of energy independence. To Nixon's earlier error of adopting independence to avoid energy shocks, twenty-first-century politicians have added the idea that independence will also solve problems of depletion, high energy prices, and global warming. A nice coincidence here is that the man who killed the first energy-independence idea. Bill (William W.) Hogan, recruited me into what has been an extended term as book-review editor for The Energy Journal. Bill's team of modelers (U.S. Federal Energy Administration 1974) well quantified the reality that energy independence then made no sense; nothing has changed this reality. A central element of the 1974 work was uncovering the difficulties of rapidly changing energy-supply patterns. The key remains that oil is mainly a transportation fuel and oil-based transportation greatly improves welfare. This is increasingly true despite the political postures that major changes in transportation and its fuel use are easily and quickly implementable.
The extensions suffer from a similar unrealism and the inappropriateness of energy independence in dealing with these further problems. Glearly, renouncing foreign supply sources will not alleviate the alleged peak-oil problem. At best, it is a coincidence that some responses to global warming involve purely domestic energy sources. Others such as nuclear and natural gas do not.
As in my earlier effort, I concentrate on the economics of exhaustible resources. This stress arises from both the depletion fears that poison political debates and the persistence of concerns among economists about the complete- ness and relevance of exhaustion theory. I reiterate and try to strengthen my earlier arguments that we have all the theory we need and that then the question is which model best fits the facts. The problem then is not the theory, but the version considered empirically relevant. In particular, peak energy is no more valid today than in the past; the bizarre National Petroleum Gouncil (2007) effort to develop an alternative rationale for intervention-problems of financing energy development- for current interventionist tendencies is a worse foundation.
I begin with general observations about the role of economic analysis, turn to exhaustion theory and its implications, restate my views on energy in- dependence as traditionally defined and justified, add comments on energy-consumption mandates, and make skeptical remarks about environmentalism.
The current hysteria about all economic problems strengthens belief in the practical value of economics. Economics has clarified many important issues, indeed undoubtedly every one to which it has been applied. Economics has even developed sound economics analyses of why the good advice is ignored. A vast rent-seeking literature explores why vote seeking leads to implementing policies that transfer money from large groups to small ones. The former then have lower unit costs that they ignore while the latter have high unit benefits that win their votes. Usually, a deadweight loss also arises.
What is critical here, however, is that economic observations about en- ergy are among the greatest triumphs. The insights then are what are stressed and applied here. Hicks-Samuelson general-equilibrium theory properly applied leads to strongly free-market conclusions. Energy is a prime example. As I argued in 1993, whatever their views on ideology, energy economists have necessarily stressed the defects of intervention because failure is all that arises.
2, EXHAUSTIBLE-RESOURCE ECONOMICS RECONSIDERED
My initial thoughts about why questions continue to arise about exhaust- ible-resource theory were that both what constitutes a good theory and what is included in exhaustible resource theory are at issue. The first question seems, on reflection on the second, much less interesting. Exhaustible-resource theory, properly delineated, is actually too rich; it covers all the relevant possibilities. This is normal in economics. Relentless theorizing leads to covering all the logical possibilities and thus to the conclusion that the outcome depends on the circumstances. Informed examination of reality is needed to determine which possibility applies. Thus, the theory has largely done all it can. Application must take over and has done so. The biggest problem is the failure by too many to choose the model most reflective of reality.
By now, it should be, but is not, recognized that the core of modem exhaustible-resource economics is the Hotelling-Grey-Cummings generalized model of depletion. In this model, we have a downward sloping demand curve, increasing costs of the rate of production, and increasing costs of cumulative production. Current output is curtailed to benefit from both (1) a rise in the gap between price and marginal cost (marginal profits) to induce hoarding and (2) the cumulative present value of the cost savings from reducing cumulative production. Hotelling (1931) presented but did little with this general model, I tried to extend (1967) but only managed to deduce that the response to depletion of lower-cost resources supplemented the pure hoarding effect and made the latter's required contribution smaller than in cases ignoring heterogeneity.
Cummings (1969), as far as I am aware, was the first to recognize that an extension of Hotelling's analysis produced a more meaningful expression of the optimality condition, Both derivations make profits a function of cumulative output and its derivative, current output. However, Cummings transforms the optimality condition into the form previously indicated. Baumol and Oates (1975) and later Modiano and Shapiro (1980) showed that a shift to discrete-time analysis vastly simplifies the derivation. Despite all these contributions, too much of the literature still emphasizes less general cases.
The general model, as should be expected, is less tractable than the more popular models. As I have often put it, the r percent rule degenerates from an observation about a potentially observable measure to the standard general proposition that something is not an asset unless it earns the market rate of interest. This contrasts to the limiting case of an already produced product being traded by speculators. (This is related to the hardtack hoarding in Fisher's Theory of Interest but was most formally developed by Samuelson (1957).) Here the price rises at r percent.
As we depart from this case, what rises from what cause becomes more complex. In Hotelling's initial case of constant costs, it is the gap between price and average cost; simply adding Grey's earlier (1914) assumption of increasing marginal costs means that it is marginal profits that rise; falling marginal costs contribute part of the r percent rise. Then adding heterogeneity means that the gains from retarding depletion of high-quality resources lessen the need to raise prices and lower the marginal costs of the rate of production. However, far too many writings on exhaustible resources ignore resource heterogeneity or even increasing costs to the rate of output.
A critical further failing of the literature is to neglect the many possible subcases that might apply and thus to fail to choose sensibly among them. Among the possible differences to consider are the shape of the cost curves, the time shifts in costs and demand, and the degree of substitutability among resources.
Time shifts wreak further havoc on simple r percent rules. Neither continual output falls nor price rises are inevitable. Thus, with homogeneous resources, if demand increases so much that the marginal profitability of prior-period output rises by more than r percent, output must be raised to lessen the price rise and marginal cost decline so marginal profits only rise r percent. (With heterogeneity, the demand increase leading to higher output would be less than r percent.) If the more-than-r-percent rise of marginal profitability of a constant output is due entirely to cost decreases, prices will fall to limit the gain to r percent. This then is why that the pattern of rising output and falling prices stressed in standard models can be avoidable.
However, perpetual greater-than-r-percent rises in profitability is inconsistent with any extraction. The infamously paradoxical concept of a commodity that is always traded and never utilized would apply. Thus, for a produced re- source, growth declines, price rises, and output falls will emerge eventually. Re- source optimists then cannot base their conclusions solely on the current absence of output declines and price rises. Fortunately, this error has not arisen.
The theory has ignored substitutability among goods. However, Chakravorty, Moreaux, and Tidball (2008) dealt with a special, deliberately oversimplified case of substitution. The paper treats two resources, coal and natural gas, that in the model differ only in their pollution effects. Both are perfect substitutes and have zero costs of production. The simplifications help to unveil new issues raised by even the very limited aspects of heterogeneity treated.
The complete theory suggests two or three basic types of possible out- comes. Most attention is given to the case in which the stock is depleted. This case is beloved by theorists because they can do more with it. This and the grip held by resource pessimism has caused too many to resist alternative views that may be more realistic. Levhari and Liviatan (1977) suggest the alternative that all that occurs is an effort to delay depletion. A long effort that starts with Bamett and Morse (1963) goes further and suggests that, as is true for many commodities such as the numerous ones displaced by microcomputers, production ends because a substitute arises and takes over the market. Bamett and Morse raised the narrower proposition that technical advance had caused lessening scarcity of supposedly exhaustible resources. Herfindahl (1967) suggested depletion was too distant to affect decisions. It is a simple next step to deduce that the rise of substitutes will doom exhaustible resources long before depletion (see Gordon 1967)." Koop- mans (1974), Adelman (1995), and Julian Simon (1996) all have well developed this argument.
Curiously, the current problem is fending off efforts prematurely to shift to substitutes. Among the related ironies is that the peak-energy faction of the re- source-pessimism approach is pushing a transition on traditional exhaustion-fear grounds while another group bases its case on the argument that the environmental effects of fossil fuel are so dire that a shift to renewables is imperative on an environmental basis. Thus, John Holdren's contribution to the hysterical response in Scientific American (2002) to Lomborg's The Skeptical Environmentalist (2001) rebuked Lomborg for attacking claims of impending depletion. Holdren's case was that this view was outmoded, but this neglects the flood of writings on peak oil. The view here is that neither contention is acceptable.
Another often-mishandled issue is the concept of the stock of minerals and our knowledge about it. This neglect reflects the more basic problem of the practical relevance of the theory. A critical point about mineral stocks is that the economic stock is very different from the physical stock. The latter is in most cases except for fossil fuels limited by the laws of physics. Fossil fuels are produced but in geological time. However, the economic stock changes steadily with shifts in costs and demand. As Adelman (1995) stresses, we will not know what the stock was until extraction ends. He adds that we need not care. What we do know is whether development is successful in maintaining the current ability to produce. This is sufficient to indicate whether exhaustion pressures are mounting. Thus, elaborate surveys of physical endowment lack economic value.