The impact of institutional variables on measures of resource scarcity.

摘要

This dissertation examines annual price data of two resources, copper and zinc, to test the hypothesis that these resources have become more scarce during the years 1870 to 1985. A theoretical model is developed which incorporates the effects of endogenous technological change and monopolistic market structure on resource prices over time. The empirical results make three important contributions toward measuring resource scarcity over a long time period. First, more appropriate statistical models are employed, than have been developed in the past, to obtain more robust conclusions regarding the trend of resource scarcity. Second, important non-cost institutional variables--cartels, wars, and depressions--are modeled as intervention variables to test whether they have exerted a significant impact on the trend of resource prices. Lastly, this research tests whether major technological advancements introduced in the early 20th century have altered the trend of resource prices.;The empirical results supported the conclusion that the error structures in the time trend models cannot be assumed to be first order autoregressive. The estimated time trend models using alternative error structures generally yielded lower errors--and hence more robust conclusions regarding the scarcity of the two resources. However, in most instances these models did not lead to significantly different time trend parameter estimates.;In most models tested, intervention variables were the expected sign and significant. Empirical results also showed that the inclusion of non-cost intervention variables altered the conclusion about resource scarcity. In data subsets for both copper (1920-85) and zinc (1925-85), models which contained intervention variables caused the positive time trend coefficient to change from insignificant to significant. That is, the results of models which incorporate the impact of significant non-cost events suggest that copper and zinc have become increasingly scarce over the course of the twentieth century.;Lastly, the Chow test, which tests for structural change, was used. The results of this test suggest that technological improvements introduced in the early 20th century resulted in structural changes in the industry; for both resources, better parameter estimates were obtained with models from the two data subsets rather than from the entire 1870-1985 series.