Should anything be done to slow down the growth of Latin America's big cities?

Whatever we conclude about the quality of life in giant cities, many observers argue that their growth should be slowed (see Gilbert and Gugler (1992: chap. 8) for a more detailed summary). First, they are draining resources from the rest of the country. As Sachs (1988: 340) puts it: "Large cities absorb above all the young and enterprising labor force. The distribution of costs and gains between the hinterland and the large city is thus once more biased in favor of the city: the countryside bears the social cost of bringing up this labor force. The benefit of their work accrues to the cities, except for the remittances of part of their meager pay to families left behind." Second, whether or not they are too large, the major cities are growing too rapidly. As Teune (1988: 373) puts it: "It may be that most pathologies of large cities are short-term effects of growth rather than inherent in them as a particular form of human organization." Finally, "because of their complexity, the life-supporting systems of large cities are highly vulnerable" (Sachs, 1988: 345). The danger is that this vulnerability may be "transformed into catastrophic realities" (ibid.: 347).

To many writers, the answer to these problems is to stem the tide of growth. According to Max-Neef (1992: 97) "the sensible move would be to revitalize the small cities - victims of a mistaken idea of progress - that are struggling to survive." Similarly, Wilhelm (1992: 199) asserts that "Latin American countries should encourage stronger networks of cities in order to arrest the intense concentration of urban life in one or two megacities. This requires the growth and development of middle-sized cities."

This kind of view dominated thinking in Latin America for many years and was the basic premise on which many regional development programmes were built. The construction of new cities, such as Brasilia and Ciudad Guayana, the establishment of regional development agencies and infrastructure programmes, as in Amazonia, and the popularity of industrial decentralization programmes throughout the region, all reflected both a real and a rhetorical desire to slow metropolitan expansion. Large cities were seemingly both inefficient and inequitable; they were draining the lifeblood of the nation; they were likely to lead to a social explosion; they demonstrated every conceivable form of social pathology. Their growth should be slowed.

Unfortunately, regional planning in Latin America has not been a huge success. With the exception of Cuba, government action has never managed to stem the tide of migrants to the major cities, to stimulate growth in most poor regions, or to help the poor, even in regions where economic growth occurred (Aguilar-Barajas, 1990; Auty, 1990; Boisier, 1987; Gilbert and Goodman, 1976; Goldsmith and Wilson, 1991). As de Mattos (1990: 26) puts it: "after three decades of effort ... the experience of regional planning ... has led to results which, in the best of cases, can only be called modest."

Perhaps the most damning indictment of regional planning and deliberate efforts to slow metropolitan expansion in Latin America is that the pace of metropolitan growth slowed during the 1980s, precisely the time when regional planning and deliberate government intervention was at its weakest (Gilbert, 1993). Chapter 2 shows that there was little growth in any of Latin America's mega-cities during the 1980s. Among the largest seven cities only Bogotá and Lima grew by more than 2 per cent per annum and Buenos Aires, Rio de Janeiro, and Mexico City grew annually by only 1 per cent.

The pace of growth of the major cities slowed because of changes in the wider economic and social environment. As the basis for development changed during the 1980s, the major cities were disadvantaged. The manufacturing plants that were so favoured by the import-substituting industrialization model and which clustered in the major cities in the 1960s and 1970s suffered badly during the 1980s. Structural adjustment and the new export-oriented model of development decimated much of Latin America's domestic industry. Recession and the removal of urban subsidies hit producers and consumers hard. Trade liberalization allowed foreign manufacturers to compete in Latin America's previously protected markets. The result was sometimes terrible for the major cities. Mexico City lost one quarter of its manufacturing jobs between 1980 and 1988, and, in Argentina, Córdoba and Rosario lost large numbers of manufacturing jobs during the early 1990s.

In the past, "many government policies [had] unintended spatial impacts distorting the spatial allocation of resources and in general unnecessarily increasing the degree of urban concentration" (Henderson, 1991: 223). Today, the situation has changed and many major cities find themselves in a far less favourable position as a result of international competition. Some Latin American mega-cities will prosper but others will not. Inefficient companies in the giant cities are in no position to sell goods abroad, and many have already gone out of business.

A second change that has slowed the growth of the major cities is the movement of industrial companies out of the metropolitan areas into smaller cities nearby. This process has been under way around Buenos Aires, São Paulo, and Mexico City for some years but the pace of change accelerated during the 1970s and 1980s (Rofman, 1974; Gilbert, 1974). There are now clear signs of an emerging polycentric urban form (Richardson, 1989). This process is very marked in the vicinity of São Paulo, where "many branch and assembly plants are locating in industrial towns within a 200-km radius of the city of São Paulo such as São José dos Campos, Piraciciba, Americana, Limeira, Rio Claro, and Campinas ... In other words, we are witnessing the extension of the localization economies of existing industrial complexes from a strictly 'urban' to a somewhat broader 'regional' scale" (Storper, 1991: 61-2).

The final change to have affected the mega-cities is the slowing of migration as a result of economic decline. During the 1980s, people stopped moving to Latin America's mega-cities in such massive numbers. Chapter 2 shows that, during the 1980s, more people moved out of Caracas, Santiago, and Mexico City than moved in. This change in the normal pattern may well continue.

Under these circumstances, the argument that Latin America's largest cities are too large is irrelevant. In any case, since government efforts to slow urban growth in the past were unsuccessful, the reintroduction of similar programmes is hardly to be recommended. Even if it were recommended, few governments would accept such advice. The world has changed and most Latin American governments seem determined to reduce the level of state intervention. They believe that the market should determine the location of economic activity, not the state. Even if some governments are still tempted to intervene, savage cuts in their budgets make that difficult. Decentralization was always conducted in a rather expensive way; incentives were always preferred to prohibition. Now that governments have fewer resources they cannot build new cities, industrial estates, or major infrastructural projects without major private-sector backing. As such, explicit spatial policies are no longer in favour. Few attempts will be made to build grandiose new capital cities in Latin America in the near future (Gilbert, 1989).