Perhaps the three most notable aspects of population and urban change in the region since 1980 are the slower population growth (underpinned by lower fertility and in some instances by less immigration and more emigration), smaller increases in the levels of urbanization and the much slower rates of growth for many of the region's largest cities. In some countries, there were significant changes in the cities and regions where the most rapid urban growth took place. These represent important changes in long established trends since Latin America and the Caribbean have had among the world's most rapid growth rates in total and in urban populations for many decades. Prior to 1980, the region has also had among the world's fastest growing large cities. It had the most rapid population growth rate of any of the world's regions for the period 1920-1970 and during the 1970s, its population growth rate was second only to Africa. By the mid 1990s, its total population was roughly the same as that of Europe whereas only 70 years ago, it was less than a third that of Europe's.

The changes in the rate at which countries were urbanizing and the slower rate of growth for many of the largest cities became evident in the returns from censuses held between 1989 and 1991. For most nations, these are a response to abrupt economic changes. For the region as a whole, the period 1950 to 1980 can be characterized as one of economic expansion, rapid urbanization and increasing life expectancy. In most nations, it was also characterized by strong government support for industrialization largely based on import substitution - priorities that had been initiated in the 1930s and 1940s. By contrast, during the 1980s, most countries experienced a serious and prolonged economic recession. The 1980s was also a period of considerable political change as many countries in the region returned to democratic forms of government and as the scale and nature of government expenditure, investment and intervention changed, partly as a result of democratization, partly as a result of measures linked to the economic crisis and structural adjustment. In most nations, barriers protecting national industries were lowered or removed, many state enterprises were privatized and a greater priority was given to exports.

Care must be taken not to exaggerate the scale of the changes in demographic trends. For instance, many of the largest cities or metropolitan areas had had significantly slower rates of population growth prior to the 1980s. In addition, despite the slower population growth rates, most major cities still had larger annual increases in population during the 1980s than during the 1970s. The economic and political changes of the 1980s are also likely to bring more fundamental changes to the spatial distribution of population during the 1990s but the form that these will take is not yet evident - rather as the spatial implications of the earlier concentration on import-substitution and industrialization also took time to become apparent, especially its concentration of urban and industrial development in a few major cities - most of them national capitals.

By 1990, the region's population totalled 440 million and had doubled in size since 1960. 71.4 percent living in urban areas, a level of urbanization similar to that of Europe. Demographic statistics for the region are much influenced by Brazil and Mexico that between them have more than half the region's total and urban population. Just eight countries have more than four fifths of the region's total population and around 87 percent of its urban population.

Perhaps the two most significant demographic changes in recent decades have been the much increased life expectancy (from 56 to 68 years, 1960-1992) and the declining rates of natural increase evident in most countries. By 1992, Cuba and Costa Rica had among the world's highest life expectancies (75.6 and 76.0) while in many other countries, life expectancies were above 70 years of age (Uruguay, Chile, Argentina and Venezuela) while figures for Mexico and Colombia were very close to 70. Several of these countries also had among the largest increases in life expectancy between 1960 and 1992 (see Chapter 3).

Most countries also underwent what is often termed the demographic transition from high mortality and fertility to low mortality and fertility between 1950 and 1990, although for some, it had begun earlier. There is considerable diversity between countries in population growth rates, that can be illustrated by contrasting three groups of nations:

< Nations with relatively low population growth rates - for instance the three nations in the Southern Cone, Cuba and two of the wealthier Caribbean nations, Jamaica and Trinidad and Tobago. But for all but Trinidad and Tobago, these are countries whose population growth rates were already below 2.0 a year in the 1950s with Jamaica and Uruguay having rates of 1.5 percent a year or less.

< Nations with population growth rates of between 1.8 and 2.5 percent a year during the 1980s that includes many of the largest and wealthiest countries such as Venezuela, Mexico, Brazil and Colombia where population growth rates have fallen steadily from the 1950s or 1960s. The 1991 census in Brazil found that the average population growth rate for the country as a whole had fallen to 1.9 percent a year between 1980 and 1991; for Mexico, it was down to 2.0 percent a year between 1980 and 1990.

< Nations where population growth rates remained high during the 1980s - for instance Honduras and Paraguay with rates above 3.0 percent a year and Guatemala with 2.9 percent a year.

At least since the 1960s, total fertility rates in large cities have been lower than the national averages although fertility rates are still above replacement levels in all but Havana. In general, the larger the city, the lower the fertility rate although there are some medium size cities where fertility levels are even lower than those of the largest cities in their own countries. The lower fertility rates in major cities is partly explained by the fact that desired fertility is lower there and health and contraceptive advice is more readily available; it is also linked to greater participation by women in city labour markets and the fact that the education of children is more valued.

Although fertility rates have generally been declining in most of the region's large cities, in some where total fertility was already low in the beginning of the 1980s - under three children per woman - there have been slight increases that are also evident at a national level. For instance, this is the case in Santiago de Chile and the postponement of births during the economic crisis, particularly between 1982 and 1985, may explain the fluctuations observed later in the decade. In addition, the rate of natural increase in major cities may be closer to other urban centres and rural areas than the differences in fertility rates suggest because of higher than average numbers of women of child-bearing age and because mortality rates are also lower. Infant mortality rates are generally lower in major cities compared to national averages - although the differences between the two are often smaller than might be expected, given the concentration of wealth and health services in the major cities. Women generally had lower mortality rates in large cities, compared to those in other urban centres who in turn had lower mortality rates than women in rural areas.

Large cities tend to have a lower proportion of children and a higher proportion of persons of working age than the national population.The population of large cities is also ageing more rapidly than their national populations, especially in countries where fertility declines began early. For instance, in Buenos Aires and Montevideo, continued lower fertility coupled with the arrival of sizeable numbers of international migrants during the 1940s and 1950s has given rise to older age structures than those found in other large cities in the region. Already by 1980, an estimated 13 percent of the population of Buenos Aires was over 60 years of age. There are often particular neighbourhoods in cities with a population that is ageing particularly rapidly, linked both to the concentration of older people (mostly women) and movement out of these neighbourhoods by younger age groups.

In many countries in the region - especially the most urbanized - the number of rural inhabitants has been decreasing while in many more, population growth rates among the rural population is very slow - less than 0.4 percent a year. In some countries, rural populations declined - for instance in Colombia between 1985 and 1993 and in Argentina for both the 1970-80 and the 1980-91 census periods. However, in several countries, notably Paraguay, Haiti and the countries in Central America, rural population growth rates remained at between 1.3 and 2.0 percent a year during the 1980s.

Detailed studies of migration processes for particular individuals, households or those in particular settlements usually reveal a complexity and diversity that is not captured in national statistics. The scale and nature of population movements and of who moves also changes, influenced by broader economic and social change. As Alan Gilbert notes:

Most countries in Latin America have had major changes in the regional distribution of their population. In many, this reflected changes in their economic orientation - for instance the rapid increase in the proportion of Ecuador's population in the coastal provinces as land there was developed fror export agriculture and as Guayaquil, the chief commercial centre and port expanded and eventually grew larger than Quito, the national capital, located within the Sierras. In many countries, both formal and informal land colonization programmes in "frontier" areas have also changed the distribution of populations within many countries, especially those with land in the Amazon that remained relatively unpopulated in the 1950s or 1960s. The proportion of the national population in Peru, Brazil, Colombia, Ecuador and Bolivia that is within the Amazon area has grown very considerably. Such regional shifts in population were generally largest in countries with large numbers of landless people or peasant farmers with very inadequate land holdings.

Rural to urban migration has come to have a much smaller role in urban population growth as countries become more urbanized. Recent studies show a decline in migration to Mexico City, Rio de Janeiro and Santiago de Chile. In Argentina, Chile and Peru, rural-urban migration accounts for a quarter or less of all internal migration. In addition, urban to urban migration has come to have a much greater influence on population trends in many cities and has become increasingly important in reshaping urban systems in many countries. In Mexico, the growing weight of urban-urban migration has been documented along with the greater diversification of the urban system and increased migration from large to intermediate and small cities. In the Dominican Republic, the proportion of migrants to Santo Domingo who originate in other urban centres has been increasing since 1970. However, the removal or lowering of trade barriers that formerly protected small-scale farmers may in some countries lead to an increased importance of rural to urban migration. For instance, there are some 2.4 million small scale maize producers in Mexico whose livelihoods will be threatened if cheap maize can enter from the United States. Cuba is an exception, since 75 percent of its internal migration still consists of rural-urban or rural-rural movements, despite its high level of urbanization. In Cuba, most rural migrants originate in scattered rural areas and move to rural areas with a higher degree of concentration or to towns.

It may be that migration movements, whether by individuals or by households, have become more diverse and more complex - in response to rapid and often unpredictable economic and social change. For instance, Roberts has suggested that in contemporary Latin American cities, many people cope with unstable economic and political development through moving, seeking out new opportunities, at times returning to their original homes, at times shifting more or less permanently their base.

There is often a preponderance of women in net migration flows to large cities - although another reason for imbalanced sex ratios in some cities is the increasing proportion of people in older age groups in which women predominate. Men move in much greater numbers to mining towns and major new industrial cities where most available work is only open to men but women dominate migration flows to most large cities because of job opportunities there in domestic service, office cleaning, shop work, street selling and prostitution. For instance, in Bogota, there has been an increasing number of female migrants and by 1993, there were 93 males for each 100 females living in the city. Social change and widespread poverty in rural areas have meant a disproportionate number of young women to seek employment in large cities, notably in domestic service but also in other services as well as manufacturing; since the late 1960s, the cut-flower export industry has been growing considerably in importance as a source of employment for young female labour, particularly in the municipalities close to Bogota. Many aspects of women's migration are distinct from men's migration but only recently has much attention been given to these aspects; as a recent review noted, the patterns of selectivity and rationality underlying gender-differentiated migration require more attention.

For cities that attract fewer migrants, so natural increase comes to have a greater importance in city population growth, even if the rate of natural increase is relatively low compared to recent decades. For instance, Buenos Aires grew during most of this century mainly from internal and international migration but during the last two decades, natural increase has accounted for two-thirds of population growth. Like many other cities in the region, Buenos Aires attracts fewer internal and international migrants and former migrants are moving from Buenos Aires to their places of origin or to other urban destinations and a growing number of native city residents have been moving to other parts of Argentina.

International migration has always had an important role in population and urban change within many countries in Latin America and the Caribbean. The large flows of international migrants that converged on the region during the 19th and 20th centuries are an important influence on population concentration between countries - and within countries since immigrants tended to settle in the larger cities. Immigrants attracted to the booming economies of Argentina, Uruguay and Southern Brazil in the late 19th and early 20th century help explain the very rapid growth of many cities there, although immigration became less important than rural to urban migration from around 1940. Only eight countries received 95 percent of the overseas migration that reached Latin America since the early 19th century and among them, seven are among the most highly urbanized. Among the largest countries, only Mexico and Peru received relatively few immigrants and in these countries is was the rural-urban migration of the indigenous population that underpinned urbanization.

In recent decades, emigration has been on a scale to have significantly lowered population growth rates for particular countries for particular periods - often coinciding with military repression, civil strife and/or economic slump. El Salvador and Haiti had very low rates of population growth during the 1980s as a result of net international emigration and high mortality. International emigration has also lowered population growth rates in Montevideo and Buenos Aires.

The economic and spatial changes during the 1980s represent a considerable break from prevailing trends in previous decades. Most countries in Latin America sustained an impressive economic performance for most of the period between 1945 and 1980. For the region, per capita income grew at an annual average of 2.7 percent during these 35 years, far above the historic rate. Manufacturing increased its share of GDP from 14 percent in 1930 to 25 percent in 1980 and imports share of GDP fell from 20 to 15 percent in this same period. Table 2.4 shows the dramatic contrast between economic performance in the 1970s compared to the 1980s. During the 1970s, only in Jamaica and in Nicaragua was there a decline in per capita income and in the case of Nicaragua, this was linked to the civil war that raged there for most of the decade. In many countries, the growth in per capita income was very rapid - especially in Brazil but also, among the other large population countries, in Mexico, Colombia and Ecuador. For most of these countries, this economic expansion built on a comparable or even higher economic expansion during the 1960s.

During the 1980s, only three countries listed within Table 2.4 (Jamaica, Colombia and Chile) avoided a decline in per capita income and their growth rates were very low in comparison to those achieved by most countries in the 1970s. For many countries, there was a rapid decline and for several countries, their per capita income was 20 to 30 percent lower in 1989 than it had been a decade earlier. The economic recession brought declines in income for those working in formal and informal sector enterprises (see Box 2.1) and a rapid growth in levels of unemployment. Social spending was also greatly reduced in most countries.

The core of the industrial growth for most of the region since the 1930s had been import substitution industries that had usually been helped by large scale government support and protective barriers. There had been a strong economic rationale for import substitution from the 1930s, with the depression in Europe and North America both before and after World War II cutting export markets and with the war itself disrupting the possibilities of importing manufactured goods. But from the 1960s onwards, a concentration on import-substitution meant that the region missed the opportunity to take part in the rapid expansion in international trade. High levels of inequality within most Latin American societies limited the expansion of the internal consumer market. During the 1980s, many industries developed on the basis of import-substitution declined or closed as local consumer markets shrunk with the recession and as protective barriers were removed and imports grew.

SOURCE: Gilbert, Alan, The Latin American City, Latin American Bureau, London, 1994, 190 pages.

______________________________________________________

In some countries, there was also a rapid shift in the orientation of the economies from industrialization based on import substitution to outward looking growth and from broad ranging state intervention to much greater reliance on market forces. These changes brought major spatial changes. For instance, in Mexico City, the country's dominant centre for import substitution industries, 250,000 industrial jobs were lost as 6,000 companies closed down while several of Mexico's cities on or close to the border with the United States grew rapidly because of the increase in employment linked to export production. Many other important centres of industrial production such as Buenos Aires (Argentina), Medellin (Colombia) and Monterrey (Mexico) also lost large numbers of manufacturing jobs. Meanwhile, the concentration on exports stimulated population growth in many urban centres that served export processing zones, zones with agricultural exports (especially those such as fruit with high value) and areas where fishing and forestry production were increased; in many countries, there was also rapid population growth among the more successful tourist centres. There was a rapid growth in export processing zones in many countries, most of them located away from national capitals.

The economic crisis during the 1980s brought several important changes in the scale and nature of urban employment. One was the loss of industrial jobs noted above. Another was a cut in public employees or a cut in their real wages. In many countries, government budgets fell rapidly; for instance, in Mexico, the federal budget fell from 34 percent of GDP to 20 percent. This also meant a large loss of jobs, given that around a fifth of all jobs in the region were in the public sector. The recession during the 1980s also meant lower investment and dwindling resources in public works which also meant fewer jobs.

Another notable change was the increasing proportion of the labour force working in services, although it is difficult to separate the influence of the economic crisis from long term economic trends that were already increasing the proportion of people working in services. For the region as a whole, by 1990, 48 percent of the economically active population worked in services (including transport and commerce) compared to 26 percent in 1950. In 1990, 26 percent of the region's population worked in industry - the same proportion as in 1980. In most cities, commerce and services provide most jobs with industry rarely contributing more than 30 percent. In Argentina, industrial jobs had represented 37 percent of all urban employment in 1974 but this fell to 24 percent in 1991 with the proportion working in services rising from 44 to 54 percent. By 1990, only 26 percent worked in agriculture compared to 55 percent in 1950.

Two other changes were of particular importance to urban economies because they brought decreases in consumer demand. The first was the scale of open unemployment that grew steadily and in countries where it stabilized, it was at much higher levels than in previous decades. The second was the increasing proportion of the workforce working in what is termed the "informal sector" in most countries and this probably represents a change in a long-established historic trend towards increased proportions of the workforce working in formal sector enterprises. Although many informal sector enterprises can generate incomes for those working in them that are comparable or higher than in the formal sector, work in the informal economy is generally less stable and also seems as if most of the growth in the informal economy was in what might be termed the 'survival informal sector' where incomes are very low.

There was also evidence of an increase in the proportion of people who are working less hours than they wished - what is sometimes calles visible underemployment. There was also a rapid growth in the number of people with incomes below national poverty lines (see Chapter 3 for more details). It is important to recall that a large proportion of those with below poverty line incomes are fully employed.

There is also evidence of a trend towards a loss of security and continuity in many jobs and also a decline in the proportion of the workforce with social security. This was not only through an increasing proportion of the labour force working in the informal economy but also through ways in which formal sector enterprises can employ people on a temporary basis or in other ways to keep down wages or social security costs - for instance through part time jobs, the use of job-agencies for temporary workers, the use of home-workers or simply employing but not registering workers. Although there are no precise figures as to its scale, it is linked to the restructuring of enterprises that seek to remain competitive as national markets are opened up to international competituion and to the weaker position of trade unions.

There is also a long term trend towards increased participation by women and decreased participation by men in the labour force. Between 1960 and 1985, female participation in the labour force increased in 21 out of 25 countries considered with male participation rates falling in all countries; among the reasons for lower male participation rates are expanding schools and an increased number of people with pensions. In Argentina, the proportion of the labour force that was female grew from 32 to 36 percent between 1980 and 1990 while the proportion in services grew from 55 to 60 percent. The contribution by women to family's cash income also rose from 31 percent in 1980 to 36 percent in 1991.

The region as a whole went from being predominantly rural to predominantly urban between 1950 and 1990 and also with a relatively high concentration of national and urban population in "million-cities". One measure of the scale of urbanization and its concentration in large cities is that by 1990, there were more people living in "million cities" in the region than living in rural areas. By 1990, most countries with more than a million inhabitants had more than half their population in urban areas.

Table 2.5 includes the proportion of the population living in urban centres in 1950 and 1990 for all nations that had 1 million or more inhabitants in 1990. Many of these countries went from being predominantly rural to predominantly urban in these four decades. The scale of this transformation is particularly notable in Dominican Republic, Puerto Rico, Mexico, Brazil, Colombia, Peru and Venezuela. The scale of economic growth explains some of this; Brazil, Mexico, Dominican Republic and Colombia had among the region's most rapid economic growth and growth in manufacturing output.

Brazil's rapid economic growth and rapid urbanization also meant that its urban population came to have increasing weight within the region's urban population. In 1950, Brazil had 28 percent of the region's urban population - little more than the three Southern Cone countries of Argentina, Chile and Uruguay with 24 percent. Of the million cities in the region in 1950, the Southern Cone countries had three (one each, in their capitals) and Brazil 2 (Sao Paulo and Rio de Janeiro). By 1990, Brazil had 35 percent of the region's urban population and 12 of its 36 "million-cities"; the Southern Cone had just 13 percent of the region's urban population and 5 of its million cities.

______________________________________________________

Table 2.5 also highlights the variety in the levels of urbanization between countries. Although the accuracy of comparisons between countries in their levels of urbanization are always limited by the differences in the criteria used to define urban centres, it is possible to identify three groups of nations. The first, the most urbanized with more than 80 percent of their population in urban areas includes the three nations in the Southern Cone and Venezuela. The second with between 50 and 80 percent in urban areas includes most of the countries that had rapid urban and industrial development during the period 1950-90 - Dominican Republic, Mexico, Brazil, Ecuador and Colombia - and also Cuba (that was already one of the most urbanized nations in the region in 1950), Bolivia, Peru and Nicaragua and Jamaica and Trinidad and Tobago. The third with less than 50 percent of the population in urban areas includes only one in South America (Paraguay) and one in the Caribbean (Haiti) along with a group of countries in Central America (Costa Rica, El Salvador, Guatemala and Honduras); all are among the less populous countries in the region (all had less than 10 million inhabitants in 1990).

Care must be taken in making inter-country comparisons since what appears as a comparable level of urbanization may have behind it very different demographic dynamics. For instance, the level of urbanization in Uruguay and Venezuela has converged but Venezuela has a much higher rate of population growth and of urban growth during the 1980s. As Lattes has noted, "the exact path that a country follows towards high urbanization depends on the historical and structural processes that condition its demographic evolution." This can be illustrated by the factors that explain why the Southern Cone of Latin America has long had among the highest levels of urbanization. The three countries in this region (Argentina, Chile and Uruguay) have among the highest proportion of their populations living in large cities of any of the world's regions; as a region, it has a higher proportion of its inhabitants in cities of 100,000 plus and one million plus inhabitants than East Europe, West Europe or North America. The three nations in the southern cone did not have rapid increases in their level of urbanization between 1950 and 1990; the proportion of Uruguay's population living in urban areas only grew from 78 to 89 percent in these 40 years. But these three nations are unusual in that they have long been among the most urbanized nations in the world. In addition, they had among the slowest growing economies and slowest growth in manufacturing output in Latin America, at least since the Sixties. Argentina and Chile also had a decrease in the proportion of their labour forces working in industry. The reasons for the high concentration of population in cities with one million plus inhabitants are rooted in their economic and demographic histories. These countries' populations were largely built up from rapid immigration from Europe in the late 19th and early 20th centuries - but where landowning structures ensured little opportunity for the immigrants to acquire farmland. Most immigrants settled in the more prosperous cities.

MAJOR CITIES: The scale of this region's urban population and its population in large cities has grown very rapidly. In 1900, it had less than 15 million urban inhabitants and no "million cities" although Rio de Janeiro had close to a million. By 1990, it had more than 300 million urban inhabitants, and 36 "million cities", including three with more than 10 million (and one other, Rio de Janeiro, with close to 10 million). The relative importance of its major cities within the world had also increased very considerably. In 1900, only 2 cities in the region were within the world's largest 20 cities (Rio de Janeiro and Buenos Aires) while only another two were within the largest 50 cities. By 1990, the region had two of the world's five largest cities (Sao Paulo and Mexico City), two more within the largest twenty cities (Rio de Janeiro and Buenos Aires) and, in total, eight of the world's fifty largest cities.

But despite the fact that the region's urban population increased more than 20fold between 1900 and 1990, most of the major urban centres today were founded in the 16th century by the Spanish and Portuguese with some such as Mexico City, Cusco and Quito being much older Pre-columbian cities. In addition, the most urbanized countries in 1950 are generally still the most urbanized countries while the least urbanized countries have remained as such. What has changed is not so much the list of the region's most important cities but the relative size of these cities and the concentration of population in large cities. In 1990, some 29 percent of the region's population lived in cities with one or more million inhabitants.

In reviewing changes in cities' relative size (in terms of number of inhabitants) and economic or political importance, not unsurprisingly, Sao Paulo has come to be the largest, reflecting its dominant economic role within the region's largest economy - although Sao Paulo was already one of the region's most important urban centres in 1900. Mexico City is the second largest, reflecting its dominant economic and political role in the region's second largest economy. Buenos Aires and Rio de Janeiro remain among the region's largest cities, although both has experienced some loss in their relative importance. Buenos Aires was the economic and political centre of Latin America's most prosperous economy in 1900. In this same year, Rio de Janeiro had been the economic and political centre of Brazil but it lost its role as political capital to Brazilia in the 1950s and Sao Paulo became the dominant economic centre. Most of the other major cities that rose up the list of the region's cities, if ordered by their population in 1990, were cities in the Southeast of Brazil where most of the rapid industrial development was concentrated: Belo Horizonte, Porto Alegre, Curitiba and Campinas although Goiania and Brasilia (both new cities) in the centre and Manaus in Amazonia also grew to become among the region's largest cities. Brazil was also unusual in having so many of its largest cities that had not been important under colonial rule - for instance, Brazilia was only created in 1958 while Porto Alegre was only a small unimportant town in 1800. Belo Horizonte was created as a new city at the end of the 19th century while neither Fortaleza nor Curitiba were important towns before becoming state capitals in the 19th century. In some ways, this re-ordering of the urban system and the fact that many major cities today were relatively small and unimportant 100 years ago is more comparable to the United States than to other countries within the region.

In Mexico, the second and third largest cities, Guadalajara and Monterrey, also grew in importance - again reflecting the fact that Mexico is the region's second most populous country and second largest economy. But in Mexico too, there has been a re-ordering of the urban system that is more recent than in Brazil but this too is creating major new cities away from the cities that are long-established centres of economic and political power - see Box 2.2.

______________________________________________________

Virtually all the cities in Latin America with a million or more inhabitants had much slower population growth rates during the 1980s than the average for the period 1950-1990 and compared to the 1970s. Many had annual average population growth rates below 2 percent a year during the 1980s including Mexico City, Buenos Aires, Sao Paulo and Rio de Janeiro; also, among the smaller metropolitan areas, Recife, Havana, Medellin, Cordoba, Rosario and Montevideo. The major cities in the Southern Cone had among the slowest rates of population increase during these four decades - a combination of low rates of natural increase for their populations and, generally, of slower economic growth. Montevideo had been the fifth largest city in the region in 1950; by 1990, it was no longer among the largest 20. Buenos Aires, Cordoba and Rosario in Argentina also grew relatively slowly in comparison to most other major cities.

Table 2.7 shows the proportion of the national population concentrated in 4 countries' capital cities (or, for Brazil, its two largest metropolitan areas) over time. It shows how Buenos Aires already had a fifth of Argentina's national population in 1895 and an increasing concentration up to 1970 with a decline since 1980. For Mexico City, the concentration of national population there was never on the same scale as in Argentina, but up to 1980, this concentration increased and has decreased since then. In Brazil, the concentration of national population within its wealthiest and largest metropolitan area, Sao Paulo, grew rapidly up to 1980 and declined between 1980 and 1991. For Rio de Janeiro, the decline began in the 1970s - although this was much influenced by the movement of the Federal Capital from Rio de Janeiro to Brasilia in 1960.

______________________________________________________

In this table, Bogota presents a completely different picture from the others since the concentration of Colombia's population in Bogota increased very considerably between 1985 and 1993. It is also interesting in that Colombia has long been noted as a country in which no one city is dominant with Medellin, Cali and Baranquilla all being cities who, although smaller than Bogota, were of a comparable scale, at least up until 1973. However, since 1964, Bogota's inter-census population growth rate has been higher than all three of these other cities with the differential particularly high for the period 1985-1993 when Bogota's population grew at 5.1 percent a year while the other three had average population growth rates of between 2.0 and 2.8 percent a year. Among other countries, the primacy of Santo Domingo within the Dominican Republic was also reported to increase during the 1980s.

One important urban trend which has recently become better understood is the heavy concentration of productive activities and of urban populations in a few 'core regions' which contain the largest cities or metropolitan areas but which cover a larger area than these. One of the most prominent examples of this was described in Chapter 1: the region in Brazil which centres on Sao Paulo metropolitan area and its surrounding belt but which also stretches to Belo Horizonte and Rio de Janeiro to its north and to Curitiba and Porto Alegre to the South. Two other important core regions are the La Plata-Buenos Aires-Campana Zarate-San Nicholas-Rosario-San Lorenzo region in Argentina and Mexico City-Toluca- Cuernavaca- Puebla-Queretaro in Mexico. While historically, the trend in most cities has been for much of the new (or expanding) industry to be within or close to the city's core area, in recent years or decades, industrial and commercial employment in many large cities has grown more rapidly outside the inner cities. There are examples both of central cities growing more slowly than suburban rings (or even losing population), of outer suburbs and "commuting towns" growing more rapidly than inner suburbs and central cities - and finally of cities beyond the commuting range of the largest centres sustaining population growth rates higher than the metropolitan areas, a process termed polarization reversal. For instance, in Buenos Aires, the central city (the Federal District) lost population between 1970 and 1980 while the population in the counties within the Greater Buenos Aires Metropolitan Area but outside the Federal District had a total population increase of 30 percent. Perhaps more significantly in the long term, Greater Buenos Aires did not increase its share of the national population during the 1970s and lost 1.5 percent of its share during the 1980s.

Box 2.3 describes the spatial expansion of Mexico City. This illustrates the development of a metropolitan area and then of a wider core region with more than one metropolitan centre. The 1990 census figures certain showed that the population in the metropolitan area was much smaller than that predicted by most people. With only 15.1 million people in 1990, it cannot possibly grow to 31 million inhabitants by the year 2000 as predicted by many in the 1970s or even to 25 million as predicted in the 1980s; it will probably be substantially less than 20 million, and may only be 17 million or so. Although still one of the world's most populous metropolitan areas, nonetheless this is still little more half the 31 million inhabitant projection that has been so widely used in the literature in discussions about "exploding cities". However, as Box 2.3 shows, the population within the wider urban agglomeration is significantly larger and this may come to have 25-30 million inhabitants within the next few decades, although over a larger territory than the metropolitan area itself.

_______________________________________________________

SOURCE: Garza, Gustavo, "Dynamics of Mexican Urbanization", Background paper for the UN Global Report on Human Settlements 1996.

______________________________________________________

Bogota reveals another pattern. A growing number of manufacturing establishments have developed around Bogota since the 1950s, particularly in a region within 250 square kilometres of the city, but the growth of such establishments has generally been much slower than in the capital city, Bogota's core has not experienced the regional dispersal of manufacturing industries; indeed, there has been a trend for manufacturing employment within the larger region to concentrate in an area of some 50 km around Bogota - although the dispersal of manufacturing jobs within the Bogota metropolitan area away from the core observed in the 1970s continued throughout the 80s although at a slower pace.

Cuba's pattern of urban development does not bear much relation to that of other nations which experienced comparable rates of rapid economic growth during the 1960s and 1970s. Since the mid Sixties, a declining proportion of its urban population has lived in Havana, the capital and much the largest city. The agrarian reform implemented shortly after the revolution in 1959 removed one of the main causes of rural to urban migration. Since then, a combination of economic and social development outside Havana (in rural and selected urban areas), the rationing system and a postponement of new housing and infrastructure investments in Havana reduced its dominance of the national urban system.

In most Latin American countries, a considerable proportion of their urban population live in urban centres other than the large cities. For instance, the 1991 census in Argentina showed that 46.5 percent of the national population lived in urban centres with less than 1 million inhabitants, including 18 percent in urban centres with less than 100,000 inhabitants. In Mexico, the 1990 census showed that a third of the population live in urban centres with between 15,000 and 1 million inhabitants compared to 27 percent in 1980 - while the proportion of the national population in cities with one million plus had declined in this period. In Colombia, more than half the urban population still lived outside the country's four "million cities" in 1993.

There are great contrasts in the size, growth rate and economic base among the urban centres in the region with less than one million inhabitants. They include thousands of small market towns and service centres with a few thousand inhabitants but also some of the region's most prosperous and rapidly growing cities over the last ten to twenty years. For instance, some of the most rapidly growing smaller cities in the region over the last few decades have been cities that served agricultural areas producing high value crops for export or cities that attracted substantial numbers of international tourists. There are probably hundreds of urban centres in Latin America with between 20,000 and 300,000 inhabitants that grew rapidly and became prosperous because of high value agricultural export crops; the development of such centres has rarely been documented in detail. One example is the Upper Valley of Rio Negro and Neuquen in Argentina where, on a 700 square kilometre fertile river valley, the total population grew from around 5,000 inhabitants in 1900 to 400,000 in 1990 with more than 80 percent of this population living in urban centres. Another is the city of Zamora in Michoacan, Mexico whose rapid growth was linked to high value agricultural crops. It is also worth recalling the importance of high value agricultural goods to the early development of some of the region's largest cities; Sao Paulo's early history was as a small frontier town that was well located to serve the production and export of coffee. However, there are also large numbers of small urban centres within agricultural areas that received little economic stimulus from agriculture or the demand generated by those who earned an income from agriculture. This included urban centres within prosperous farming areas where most of the producer and consumer demand for goods and services bypassed the local town and went direct to larger and more distant cities. The pattern of landownership is often a major influence on the extent to which high value crop production stimulate urban development locally; a large number of small but prosperous farmers with intensive production can provide a large stimulus to local urban development whereas agricultural plantations or large landholdings steer most of the economic stimulus to more distant cities, including those overseas.

There are other cities that have grown rapidly on the basis of tourism or where tourism has become increasingly important. For instance, Cuautla in Mexico grew from a small market town with 18,000 inhabitants in 1940 to a city with over 120,000 in 1991, with tourism becoming increasingly important, especially weekend tourism from those living in Mexico City. Bariloche in Argentina had just 6,562 inhabitants in 1960 and 81,001 in 1991 - and it received half a million visitors in 1992. Other small or intermediate urban centres have had their size and importance boosted by major universities developing there - for instance Merida in Venezuela that was originally a small market town that grew to serve the production and export of coffee greatly increased its size to around 250,000; one of the main factors has been the development and expansion of the University of the Andes there that by 1987 had 37,616 students, 2949 teachers and 3268 employees and workers.

There are also hundreds of smaller urban centres within the region that have lost importance, as the goods or services they provided were no longer in demand or because they were a market town and service centre for an agricultural area whose products were also in less demand or faced falling prices or competition from other areas. Latin America has many urban centres that originally grew as mining centres or to serve the production of coffee, bananas or sugar or that were centres of transport that are now bypassed (for instance old river ports with much less demand for their services as most goods move by road).

Many countries have achieved relatively rapid rates of economic growth in the first few years of the 1990s and this may have increased once again the rate of increase in the level of urbanization. However, in some of the most urbanized countries, levels of urbanization are now so high that there is no longer great potential for substantial increases in urbanization levels, although there is considerable scope for changes in the spatial distribution of city and metropolitan centres that, as in the United States, are not reflected in changes in urbanization levels. For instance, profound changes will be brought to Mexico's urban system by a much smaller public sector and a more open economy (that may include millions of small farmers moving off the land because their maize production is undercut by cheap imports) and a continued shift of its modern industrial base to the border region. The drop in employment in manufacturing and in government that became evident in the 1980s may well continue through the 1990s; One estimate suggests that a further 10 million jobs may be lost in manufacturing in the region during the 1990s; another suggests that between 5 and 10 percent of the workforce may lose their jobs because of cutbacks in government and public sector enterprises.

It is difficult to predict the scale and nature of urban change, given that it is so dependent on economic performance. For the countries that sustain rapid economic growth, urbanization is likely to continue. For countries that do not, levels of urbanization may not grow unless civil or political strife becomes important in forcing people off the land - as it did in Colombia during "La Violencia" in the 1950s and 1960s, and in Peru during recent years because of the activities of the "Shining Path" guerillas. But whatever the economic performance, the dominance of the region's largest metropolitan areas both in their concentration of economic activities and their concentration of population is likely to lessen in most countries. In some, it will be in new cities; in others, still concentrated in core regions with the largest metropolitan centres at their core. Part of the reason is the change in economic orientation in which other cities have greater comparative advantage - the important centres for tourism and cities well located to attract new investment in export-oriented manufacturing or to benefit from forward and backward linkages from high value export agriculture. Part of the reason is the shrinking of government, that is likely to affect national capitals most. But there are also the factors that helped decentralize urban development in Europe and North America that are also operating in many of the higher income countries, including good quality inter-regional transport and communications systems.

NOTES AND REFERENCES; SECTIONS 2.1-2.3