In the literature, comparisons between absolute levels of productivity in different industries occur frequently. This paper explores the meaning and possible significance of such measures. Prices must be used to permit the required comparison of the outputs of different industries. It is shown that base-year weights produce a valid measure of productivity growth but not of absolute productivity level. Current price weights do yield a valid index of absolute productivity but one that tends to be equal for all industries because of the general equilibrium mechanism that reallocates resources from low-productivity to high-productivity industries.

Energy-production subsidies are, paradoxically, shown to be likely to increase U.S. dependence on imported oil. Standard studies of net energy yields are shown to be seriously biased upward for two reasons. First, many omit indirect energy inputs, which, our input-output calculation shows, probably causes large errors. Second, those studies all omit the energy opportunity costs of nonenergy inputs (e.g., the fuel substituted elsewhere for the labor used to produce energy). We prove that, absent externalities, any fuel-output subsidy which causes an otherwise unprofitable expansion must yield an incremental fuel output smaller than the increment in energy input plus the energy opportunity costs of other inputs.

I. Introduction, 48. — II. A preliminary model for the valuation of convertible bonds, 49. — III. Statistical implementation of the model, 51. — IV. Altering the assumptions, 54. — V. Summary comment, 59.