Transportation is not merely a derived demand, but a determinant of new production possibilities. In developing economies, where the lack of mobility is self-evident, it is absolutely necessary to consider the catalytic impacts of transport services. Transportation not only directly affects the overall output in an economy through accessibility and costs, but also stimulates and influences the shift in the demographic sector in terms of population movements and unemployment rates. To successfully plan for the development of a region, one must understand the possible causal relationships, feedbacks and interactions between the different sectors of both the investment region and the possible spatially impacted region.

In this study, the impacts of three investment strategies for the Essequibo Coast region, in Guyana, are evaluated through the use of a computer simulation and system dynamics methodology. The model consists of two regions -- the region in which investment is provided and the region that is spatially impacted due to this investment. The hypothesized interrelationships of the main sectors (demographic, economic and transport) and components of each region were first developed as causal submodels. Secondly, the submodels were synthesized to form a compre- hensive system dynamics computer model represented by approximately 280 equations to evaluate the three strategies: (1) Do Nothing; (2) Investments in Roads Only; and (3) Investment in Roads, Drainage and Irrigation.

Sensitivity analyses were performed on the key socio-economic variables (Drainage, Irrigation, Fertilizer, Mechanization and Regional Migration) to determine which variables most significantly influence regional behavior. These tests showed that a tacit acceptance (i.e., not explicitly incorporating the above factors in a model) of the availability of these resources overstates the impacts due to transport investments -- i.e., roads equal development is also a misconception.

The investment strategy in Roads, Drainage and Irrigation provided the greatest net benefits and most favorable socio-economic characteristics in terms of population level, regional income per capita, outmigration and unemployment. So, given its financial feasibility, it is reco1m1ended for implementation. Further, it is also suggested, because of the model's demonstrable flexibility, that it be used for post investment analyses and future model calibration.