Survey of price elasticities from economic exploration models of US oil and gas supply

Abstract

Exploration for oil or gas reserves consists of searching for and finding new reserves. It begins with the study of the geology of an area followed by exploratory or wildcat drilling in promising areas. How much oil or gas is found or the supply of new reserves is a function of exploration, the geology of the area drilled along with a random component measuring the fickleness of mother nature in yielding up her treasures to mere mortals. Knowing how oil and gas prices affect this process (price elasticities) is valuable to those who are involved with this strategic industry. Economic theory suggests that the search for petroleum is affected by prices and can be characterized by functions representing geophysical activities and drilling while the finding of reserves can be characterized by some sort of discovery process. The earliest econometric approach to modeling oil and gas exploration, which yielded price elasticities, was to estimate the search process using a function for wildcat wells drilled (Ww). For example, the discovery process for oil was represented using two equations: a success rate equation % of wells that are commercially successful (W/Ww) and an average oil reserve per successful well equation (O/W). Each of these three dependent variables were dependent on price and other relevant variables. The product of these three variables (Ww^∗W/Ww^∗O/W) yields the supply of oil reserves (O). In this paper we survey all U.S. exploration models for oil and gas that include wells drilled, share of successful wells, average reserves per well, or total reserve equations. We focus our survey on price elasticities to capture the effect of market price on the exploration process. Drilling equation results tend to be good with drilling strongly influenced by oil price and we suspect the long run drilling oil price elasticity to be greated than 1. It is much less clear how natural gas prices have affected drilling but their effect seems to be increasing. The most important issue in drilling equations besides including the correct economic variables is more work to clarify what geological variables provide the best forecasts. The econometric estimates for successful wells and average reserves per wells are much poorer and we would recommend more systematic work on discovery models to determine which perform the best.