Resources for the Future
Krautkraemer and Toman
less than in a situation with certainty about the resource stock since producers would put weight on the possibility of more scarce and therefore more valuable carry-over stocks. If, on the other hand, there is a renewable “backstop” energy form that can be supplied if the price of depletable energy climbs high enough, the Hotelling Rule will still hold, but the net price path will rise at the rate of discount until it meets the unit cost of the backstop just at the point of full depletion.
Despite its seeming intuitive appeal, however, the basic Hotelling model is built upon very shaky assumptions. It assumes a known stock of a resource of homogeneous quality and that the extraction technology does not change over time. In fact, nonrenewable resource stocks are not a given: exploration for new deposits, as well as the further development of existing deposits, is an important feature of fossil fuel supply. For many nonrenewable resources, including fossil fuels, the discovery of additional deposits has exceeded consumption so that known, or proved reserves have actually increased over many time periods. For example, U.S. oil reserves increased from 13 billion barrels in 1930 to 33.8 billion barrels at the end of 1990, while production in that period was 124 billion barrels (Adelman 1993). More than 30 billion barrels were produced from 1991 to 2000, but reserves at the end of 2000 were 29.7 billion barrels. The increase in world oil reserves is even more dramatic—proved reserves increased from 660 billion barrels at the end of 1980 to 1,009 billion barrels at the end of 1990. This increased slightly to 1,046 billion barrels at the end of 2000 even though 250 billion barrels were produced from 1991–2000 (British Petroleum 2001).
Nonrenewable resources, including energy resources, also vary in quality and in their cost of extraction. Moreover, cost conditions for extraction in a specific petroleum reservoir change over the economic life of the reservoir: as natural drive is depleted, cost rises and yield per unit extraction effort falls, unless costly secondary and other recovery enhancement measures (e.g., injection of water or carbon dioxide) are undertaken to stimulate reservoir pressure. In short, resource stocks are neither fixed nor homogeneous, contrary to the assumptions underlying the simple Hotelling Rule. It is not surprising, therefore, that the simple rule has done poorly in describing actual supply behavior (see below). We must look to a richer theory if we want to better understand observed behavior.