ABSTRACT
One of the important findings that emerges from the preceding chapter and its review of historical trends in resource availability is that extrapolations of past trends are not likely to provide reliable forecasts, whether these are projections of price, extraction costs, or user costs. 1 This, of course, is not a big surprise. The validity of such projections requires that trends in the important underlying factors governing the past—along with the complex mechanism by which they interact—remain unchanged during the forecast period. Or alternatively, if there are changes, they must completely offset each other so that their net effect is zero. On occasions, one of these conditions may hold, and such forecasts may turn out to be quite accurate. But this is more a matter of luck than of any true power to discern the future (see Box 5-1).
