Road network investment model for developing countries

The fact that transportation plays an important role in the achievement of the socio-economic development goals is well recognized by most developing nations. The scarce resources, as well as the competition among all sectors of the economy for the available resources, limit the number of transport projects that can be implemented in any specific period of time. Moreover, the sequence of implementation of these projects within each period is also affected. Thus, in order for the transport improvements to contribute to the achievement of the national socio-economic goals, their sequence of implementation, and hence their contribution, should be carefully considered. The aim of this research, therefore, was to devise a methodology for a road investment program in the context of a developing country. Two objectives were considered: the maximization of relative distribution of regional accessibility, and the minimization of the total system per unit operating cost. Cost minimization has been the most frequently used objective in network improvement literature. However, consideration of cost minimization as the sole objective, especially in the case of network improvement in developing countries, tends to concentrate road development in those regions which already have an extensive infrastructure at the expense of less-accessible and less-developed regions. Another important objective to be achieved by improvements is the provision of accessibility, especially at the early stages of the network evolution. More important is the achievement of an equitable accessibility improvement among the regions--equitable in the sense of relative distribution of accessibility. An entropy statistic was, thus, used to measure the distribution of relative accessibility. The above two objectives were used as the basic criteria for the road network investment model (RNIM) developed in this research. The equity measure was found to be a monotonically decreasing function of the total per unit operating cost. This resulted in a unique set of link improvements for an investment in which both objectives converge and any deviation from that amount resulted in the presence of more than one solution for that particular investment. In this case, a measure, (epsilon), defined as the rate of substitution per unit of investment, was used to choose the best solution. It is of interest to note that at the early stages of network improvement the use of the measure emphasizes improvements which will provide greater equity in terms of the distribution of relative accessibility between the regions of the study area rather than the improvements which minimize the total per unit operating cost. It is also interesting to note that once each region is connected to the rest of the road network with at least one improved road link, then the emphasis shifts to those improvements that will minimize the total per unit operating cost. RNIM is devised as a screening process for the early stages of determining a road investment program, in which further analysis of chosen roads can be carried out in detail. Thus, the data requirement of the model is small and in aggregate form. However, if more accurate data are available, the model can be used to determine the road investment program with no need for further project analysis.