The allocation of investment to a transportation system cannot be prescribed without accounting for the interdependencies that relate the transportation system to the larger economic system of which it is a part. Like all other sectors of the economy, the transportation sector must function within the operating conditions and constraints determined by the other sectors of the economy. In turn, the transportation system influences the performance and operating structure of the rest of the economy. These interdependencies are of course complicated by the fact they occur dynamically as both the transportation system and the other sectors of the economy evolve through time. Despite the significance of these interactions, very few models have been developed which explicitly account for these interrelations, particularly within a planning context. (See for example Bronzini et. al. [4] and Amano and Fujita [1].) Most models of this sort attempt to effect a rather complex amalgamation among input-output, econometric, and transportation sub-models. By far the best known model is the complex simulation model developed within the Brookings Transport Research Program [11]. In this model, an attempt is made to trace the effects of transportation system improvements through the economic system to a point where they ,can be directly and unambiguously compared in terms of their effect on gross domestic (or regional) product, the rate of inflation, and national and regional employment. However, none of these criteria are directly altered by changes in unit transport costs or capacity. Generally, the effects on such measures represent the outcomes of the interaction of a number of complex interrelated processes involving changes in the derived demand for intermediate goods, adjustments in the levels of production, inventory, and investment, and ultimately in the locational profitability of economic activity. Many of these mechanisms involve behavioral decisions and time lags and, thus, cannot be specified by simple mathematical functions. The particular form of the functions chosen by Kresge and Roberts [11] to simulate the required functions in their empirical work on Colombia has been severely criticized by Holland [9, pp. 21-23]: