NATURAL RESOURCE EXTRACTION AND EXPLORATION UNDER UNCERTAINT

There are at least two reasons why the market for natural resources may differ from that of a standard commodity in a perfectly competitive economy. First, natural resources are nonrenewable: extracting a barrel of oil today leaves one less barrel for future generations. Second, the global endowment of resources--the size, quality and location of resource deposits--is for the most part unknown. In this dissertation, we first present a model which captures these two aspects of natural resources, but does so in a manner somewhat different from the standard literature. We view the resource as inexhaustible, but occurring in a continuum of grades. As reserves get depleted, producers resort to lower grade ores. The burden imposed by current extraction, in this framework, is the higher extraction costs borne by future generations. This leads quite naturally to suggesting a motivation for exploration: to find high grade ores to replenish those lost by depletion. However, because of the uncertainty in resource supply, exploration is a risky process and this is captured by a random production function in our model.;Having spelled out this two-period, stochastic dynamic optimization model, we use it to answer three questions: First, if extraction today places a burden on future users, what is the social cost of extracting a nonrenewable resource and how may it be estimated? This has implications for the debate over appropriate measures of resource scarcity since it has been said that rent--the social cost of resource extraction--comes closest to meeting the requirements of a scarcity indicator. In evaluating suggestions for estimating this (unobservable) rent, we show how the use of marginal discovery cost as a proxy could lead to biased estimates. The reason has to do with the uncertainty surrounding the exploration process. Because only exploratory inputs can be chosen ex ante, but marginal discovery cost depends on output, the familiar price-marginal cost equality fails to hold under uncertainty. Methods of identifying the direction of the bias are discussed. Some empirical evidence is offered, however, to suggest that discovery costs for U.S. oil and gas were rising during a period of increasing scarcity (when other indicators were falling), so that the enthusiasm for this scarcity measure may not be misplaced.;The second question addressed in this dissertation is, how does risky exploration affect the optimal strategies of resource firms? Since extraction depletes reserves while exploration augments them, the uncertainty can affect both the extractive and exploratory decisions. We show that the result depends inter alia on how uncertainty affects resource rent. Under plausible conditions, the net effect will be to increase current extraction and lower exploratory effort--a result which is at variance with those derived from simpler views of the uncertainty surrounding natural resource supply.;Finally, we ask the normative question, does exploratory risk cause the competitive equilibrium to diverge from the social optimum? Since the answer depends as much on the structure of the rest of the economy as on the resource sector, we analyze the question for a variety of models of the underlying economy. Conditions under which social and private optima diverge center around imperfections in the capital market (in particular the absence of a stock market) and the possibility of re-opening a market after the output of exploratory activity is known. If we are talking about the efficiency of the world economy, with resources owned by different countries, the former condition may apply; for the U.S. economy, with resources managed by firms, the latter condition is relevant. In either case, therefore, there is some reason to believe that risky exploration causes the competitive equilibrium not to be socially optimal.