| Sumario: | When the increase in oil prices came, users of input-output models had no difficulty in translating them into consumer prices. But what effects would consumer prices have on consumption? What would happen to the demand for tires or to the demand for railroad travel? To give a correct answer, the consumer demand functions had to be able to express complementarity between tires and gasoline and substitution between private transportation and public transportation. Despite a vast literature on demand functions, no one had at that time, so far as I am aware, an operating system of demand equations with such capacities. Since that time, we have developed such systems for several countries. They are used in input-output models designed for long-range forecasting with substantial increases in income over the period of the forecast. The form of the equations we use must give careful attention, therefore, to the effects of prices on the marginal propensity to consume each product as income increases. <BR> If, then, a derivation by utility maximization is neither necessary nor sufficient for the usefulness of a system, we must develop other criteria. In the course of examining the properties of various systems, there gradually emerged a number of features that a system of demand functions needs in order to portray the world as I expect it to be.
|
|---|