The limitations of physical planning to control urban sprawl are evident. Its control would only be possible in a context of profound economic adjustments of the current model, or by a slowdown in economic growth or an adjustment of the role of private ownership in the economy.

At existing rates of economic growth in Latin America – given the relationship between economic growth and urban sprawl – could not expect a sprawling city, which is based on a highly developed infrastructure (car) and only in this way allows the decoupling comparative advantages of urban agglomeration. It is expected that American cities are compact, rather than Europe, where rates of investment per km2 are incomparably higher and finally explaining the suburbanization of urban development. Latin American cities expand as oil stain with edge irregularities and gaps that manifested while market imperfections in land prices are more continuous than those seen in the old continent. Still require several points of GDP growth to experience expansive formats leapfrog older polymerizations hinterland.

Thus the control and minimization of urban fragmentation appears as a priority objective of physical planning and public policy. This implies the ability to predict the expansion space vector, the weight of its components so as to anticipate adverse effects. Also adequate monitoring of its features. Not understanding the combination of factors can lead to land policy to focus on minor issues losing their effectiveness. Moreover, the ability to predict spatial patterns based on algorithms based on probability of urban change is a powerful tool to implement in future development of land policy in Latin American countries, where the absence of empirical models, robust from the scientific point of view so far relies on intuitive judgments.