|
Article Excerpt
"Peak oil" refers to the future decline in world production of crude oil and to the accompanying potentially calamitous effects. The majority of the literature on peak oil is non-economic and ignores price effects even when analyzing policies. Unfortunately, most economic models of depletable resources do not generate production peaks. I present four models which generate production peaks in equilibrium. Production increases in the models are driven by: demand increases, cost reductions through advancing technology, cost reductions through reserve additions, and production capacity increases through site development. Production decreases are driven by scarcity. The models do not rely on market failures and indicate that a peak in production may arise from efficient intertemporal optimization. The models show that prices are a better indicator of impending scarcity than peaking is and that peak production can occur when any percentage from 0-100% of the original deposit remains.
1. INTRODUCTION
The term "peak oil" has come to be synonymous with a host of concerns about future energy supplies. (1) A vast non-economic literature addresses whether global oil production has peaked or will soon peak; what consequences that could have for fossil fuel dependent societies; and what should be done about it. (2) Unfortunately, most of this literature fails to recognize the role that prices could play in allocating scarce oil. For example, Hirsch et al. (2005) notes a wedge between projections of oil production, which peaks, and oil consumption, which does not. (3) The authors analyze a variety of mitigation policies and conclude that prompt action is required to prevent future shortfalls and economic disruptions. However, they do not mention the effects of these policies on prices or the effects of prices on these policies. Similarly, Lovins et al. (2005) proposes a variety of demand reduction policies for "getting the United States completely, attractively, and profitably off oil." However, the analysis ignores the fact that these policies would decrease demand for oil, presumably decreasing the price of oil and the profitability of the policies.
Most economic models of depletable resources do not seem to offer additional insight because they do not explicitly generate a peak in production. (4) For example, the seminal model of Hotelling (1931) predicts that (net) prices should grow at the rate of interest and that production should steadily decline over time. (5) Extensions of this model for uncertainty, limited capacity, set-up costs, different grades of ore, and increasing costs with cumulative extraction, also do not generate peaks in production. (6) This raises the question as to whether the observed production peak could have arisen from an economic model. Is the production peak itself evidence of some market failure or disequilibrium? Is oil peaking just a series of happy (or unhappy) accidents that is not amenable to economic analysis? (7)
This paper answers these questions by presenting four economic models that generate peaks in production without resorting to market failures. The models are solidly based on the classic Hotelling theoretical framework of optimizing producers and consumers. (8) Thus the models show that peaking is consistent with dynamic efficiency and is not evidence of some market failure.
The peak in each model is generated by opposing forces tending to increase or decrease equilibrium production. In the models, increasing demand, improvements in technology, additional reserves, and new site development tend to increase production while scarcity tends to decrease production. Given the fundamental nature of these forces, it would be more surprising if production did not peak than if it did peak!
Indeed, oil production in many regions has peaked. After increasing for over 100 years, U.S. annual crude production peaked in 1970 at 3.5 billion barrels of oil and has generally declined since. Brandt (2006) analyzes 139 (potentially overlapping) oil producing regions throughout the world and argues that production in 123 regions can be reasonably modeled as single peaked and that production in 74 of these regions has already peaked. (9) Furthermore, production of other resources has also peaked. (10) This widespread empirical evidence of production peaking highlights the importance of understanding why production of an exhaustible resource might peak.
Although, the peak-oil literature focuses attention on peak production, the underlying concern seems to be that the transition from cheap oil to expensive substitutes will be sudden, chaotic, and costly. This attention to peak production is misplaced for several reasons. First, peak production is irrelevant to concerns about the transition to substitute resources. The transition to renewables should occur when oil resources are depleted such that their price rises to the production cost of renewables. The models show that the transition should occur after production peaks since oil should be used to smooth the transition to the renewable resource. Second, the price path is a better indicator of impending resource scarcity than the production path is. As the models show, prices will begin to increase before production peaks and thus are an earlier indicator of future scarcity. (11)
While the transition to renewables will surely have some surprises for everyone, energy use is likely to be smooth across the transition. There are two reasons the equilibrium price path (of a barrel of oil equivalent) cannot jump during the transition. First, a forward-looking firm, or government, could profitably save some (or all) of its oil for production after any price jump. Since oil is virtually costless to store in its natural reservoir, such intertemporal arbitrage would eliminate any jumps in the price path. Second, even if there were no forward-looking firms or governments with secure property rights to oil, the increasing cost of oil extraction from different deposits would prevent jumps in the price path. For example, if the marginal extraction cost of the highest cost deposit were $200 per barrel and the cost of the renewable substitute were $260, then completely myopic firms would exhaust the oil at a price of $200, and the price would then jump to $260. (12) However, since oil production costs are smoothly increasing, there are oil deposits with production costs between $200 and $260. Even completely myopic firms without secure property rights would wait to produce from these deposits until the price were high enough to cover the extraction costs. Production from these high-cost deposits will ensure a smooth (although possibly inefficient) transition from oil to renewable resources.
The four models are presented and discussed in Section 2. Each model incorporates fundamental economic forces into the basic Hotelling model and generates an endogenous peak in equilibrium production.
Although the first three models have clear antecedents in the literature, the fourth model is novel. The main motivation for this model comes from casual observation of the oil industry. The original oilfields in the U.S. centered on Pennsylvania....
|
