Microeconometric evaluation of natural resource scarcity under restricted cost minimization theory: Sensitivity, theoretical consistency, and structural dynamics
Rent can be derived from the variable cost function by invoking Hotelling's lemma. Determining rent endogenously controls for data limitations; but also brings forth important methodological issues, intrinsic to using an economic foundation for measuring resource scarcity, which have not been comprehensively addressed in the literature. It is surprising, for example, that no one has investigated the sensitivity of rent, as a measure of resource scarcity, to assumptions involved in specifying the econometric model used to describe production technology. This dissertation has two objectives. The first consists of evaluating the sensitivity of rent, as a resource scarcity measure, to econometric specifications of the variable cost function to assess whether assumptions on model characteristics translate into statistically different rent estimates. The second objective is to quantitatively determine whether research decisions and theoretical consistency affect information content of a resource scarcity measure based on rent. Restricted cost minimization theory serves as organizing principle. The translog is used to detail six econometric specifications for the variable cost function based on research decisions that define model characteristics before and after estimation. Labor, energy, and materials are treated as variable inputs, and capital and reserves as quasifixed. Temporary equilibrium with respect to capital is assumed. These models are estimated using iterative SUR and nonlinear MLE. Canadian asbestos industry 1953-1985 time-series data are used. The sensitivity analysis consists of two parts. The first examines whether rent is statistically different across models. The second part fits a response surface to rent in order to evaluate its robustness to research decisions. Rent movement and direction are examined with a correlation matrix and trend estimates. The broad conclusion to emerge from this dissertation is that research decisions make a difference but not to the extent that they affect the informational consistency of rent as a scarcity measure. Particularly, rent is sensitive to assumptions on scale, technical change, and equilibrium; it is also sensitive to regularity properties. Rent is informationally consistent because it moves unidirectionally across models suggesting increasing scarcity over the sample period. These findings lend support to a microeconomic approach to measuring resource scarcity based on classical econometrics.