RESUMO
PIP: The rapid urban growth and increasing number of megacities in Latin America and other developing countries are fundamentally different phenomena than those observed at the time that location theory was developed. To examine whether existing location theory applies to developing countries, an econometric analysis of the relationship between urbanization, city size, and development was first conducted. The relationship between urbanization and development was expressed in the form of a series of regression analyses applied to World Bank data for 96 developing and developed countries. After logarithmic transformation, a simple equation associating total and total urban population of the country and per capita gross national product was able to explain 93% of the variance in total urban population. This result demonstrates that it is not possible to regard urbanization as abnormal. As to the size of particular cities, deviations from the "normal" trajectory of urbanization apparently originate in particular institutional features of some countries. Manifestations of overurbanization in Latin America demonstrate that urbanization is a necessary, but not sufficient, condition of development. A comparative study of spatial location of employment in different economic sectors in Canada and Mexico was next conducted. For this analysis, cities of over 25,000 population in Canada and Mexico were classified into 32 economic activity sectors, which were regrouped into 18 for the analysis. Two matrixes of ten city types and 18 and 32 employment groups were constructed for each country, with the corresponding number of employees noted. The employment information was transformed into quotients of location, with a quotient above 100 signifying concentration of employment. The results did not suggest that factors of localization of employment are different in developing countries. Models of localization of economic activities adopt analogous characteristics imposed by geography and technological conditions. In Canada as in Mexico, economies of scale and distance are the principle variable explaining models of localization.^ieng