1.2. Global Population Change and Urbanization


Changes in population and in households

The world's total population in 1995 was estimated at 5.7 billion. The rate at which it is has been growing was essentially constant between 1975 and 1990 (at around 1.7 percent a year) and it is projected to drop to an average of around 1.5 percent a year during the 1990s. Fertility rates have declined almost everywhere in recent years, although the pace of change differs greatly. For aggregate regional statistics, the exception is Africa, although fertility decline has started in a number of countries in East and Southern Africa. However, the global picture is one of immense diversity. Thus, the discussion of recent demographic changes and their implications for human settlements will be concentrated in Chapter 2, within sections that consider population and urban change in each of the world's regions.

In most societies, there have been rapid changes in recent decades in the size of households and in their structure (including the proportion of households headed by women and the types of person within the household). One of the most dramatic has been the increase in the proportion of female-headed households which are now thought to comprise more than one fifth of all households worldwide - although with large variations between countries. Box 1.2 gives more details and discusses the links between this growth and economic and urban change. It is also likely that in most societies, there is or has been an increase in the proportion of nuclear families and a fall in the average size of households which has accompanied (and is linked to) an increasingly monetized economy and increasing proportions of the labour force working in non-agricultural activities and living in urban areas. For the United States and for much of Europe, the average household size is now less than 3 persons. Another important change, recorded in many countries in the North, is a rapid increase in the proportion of single person households. In much of Europe, over a quarter of all households are one person households; in the USA, over a fifth of all households are one person households. These and other changes in the size and composition of households have important and often under-rated influences on urban changes and housing markets. They bring major changes in the number of households who need accommodation; a rapid growth in the number of households can mean that housing demand rises, even as the overall population of a city falls. They also bring changes in the type of accommodation that is sought, in household income and in households' preferences as to where they want to live.

However, the scale and nature of changes in household size and structure appear too varied from society to society or within a society over time (or perhaps even between income groups) to permit much detailed generalization. International comparisons are also hampered by the differences in definition or interpretation of "household" between countries and the fact that there is no recent data on households in many countries. In addition, the assumption that economic changes always promote a change from a predominance of extended families to a predominance of nuclear families has been challenged both by historical studies which show that household composition has long been highly variable in rural areas and by contemporary studies which find extended households common in urban areas. The relative proportion of nuclear or extended families depends on a great variety of macro and micro factors - for instance, in rural areas, they include local forms of agricultural production, land availability and inheritance, kinship patterns and demographic variables such as life expectancy. There is also evidence of a rising proportion of extended families in many urban areas, largely as a response by households to falling real incomes. Chapter 6 will describe one aspect of this - the number of young adults (single and married) who share the house of parents or other kin as they lack the income to afford their own accommodation. Chapter 3 will also give examples of the growth of extended families in urban areas as relatives come to live with nuclear families because they have lost their own livelihoods or as relatives are welcomed for the extra income and household work they can contribute - although there are also examples of extended families dividing, as the nuclear household can no longer support relatives. Thus, "while there is little doubt that household composition usually undergoes some change in the course of urban development, these changes take different forms at different times and in different places and they can by no means be generalized."





Box 1.2: Women Headed Households and Urbanization

Women-headed households form a rising proportion of households in most parts of the world. Worldwide, it is estimated that one fifth of all households are women-headed households. This is not a homogenous category as it includes not only single parent women-headed households (headed by unmarried, divorced, separated or widowed women) but also households headed by grandmothers and women who live alone or with other women. Although such households are not a new phenomenon and in many societies they have a long history, what is new is their importance relative to other forms of household and the fact that they have become common in many more societies.

There are large variations between regions and nations in the proportion of women-headed households and in the extent to which this proportion is changing. For instance, only an estimated 13 percent of household in Asia and the Pacific are women headed compared to 19.1 percent in Africa and 18.2 in Latin America. In some countries, the proportion of women headed households can rise to more than a third of all households. In some countries such as India and the Philippines, female household headship does not seem to have risen much or at all since the 1960s while in others, especially in Latin America and the Caribbean it has grown quite substantially.

Although women-headed households are not unique to urban areas and in some regions such as South Asia and sub-Saharan Africa they are more characteristic of rural districts, many of the factors responsible for female-headed household formation arise through urbanization. Urbanization and its outcomes brings changes in gender roles and relations and in gender inequalities (although with great variety in the form and intensity from place to place). This can be seen through the transformation of household structure, the shifts in household survival strategies, and changing patterns of employment.

In addition, the urbanization process is itself frequently shaped by gender roles and relations - for instance through the scale and nature of female migration into urban areas (which is much influenced by decisions in rural households about who should migrate and for what reason) and the influence on the urban labour market arising from constraints placed on women's right to work outside the home by households and societies and by the extent of the demand for female labour. The degree and nature of gender selective movement to urban areas is often a major influence on both the frequency and the spatial distribution of women headed households within countries. In general, where men dominate rural-urban migration streams as in South Asia, North Africa, the Middle East and many parts of sub-Saharan Africa, urban sex ratios show more men than women and women headed households are usually more characteristic of rural than urban areas. In the towns and cities of East and Southeast Asia, and Latin America and the Caribbean, rural out-migration is female selective, urban sex ratios usually show more women than men and levels of female household headship are higher in urban areas.

There is a tendency to equate the growth in female households with the growth in poor or disadvantaged households - but this growth has positive aspects in many societies. For instance, women headed households are likely to be free of patriarchy at the domestic level - while other benefits may be greater self esteem, more personal freedom, higher degrees of flexibility in terms of taking paid work, enhanced control over finances, and a reduction or absence of physical and/or emotional abuse. Women-headed households are not always over-represented among the lowest income groups. In addition, when women decide to set up their own houses, it may be a positive and empowering step, especially if in doing so, they are able to further their personal interests and the well-being of their dependents. Various studies have shown that the expenditure patterns of female headed households are more biased towards nutrition and education than those of male headed households, with less spent on items that contribute little or nothing to the household's basic needs. Where this is the case, women and children within a low income female headed household will usually have better diets than those in male-headed households with the same income, and with less tendency for children or youth to be withdrawn prematurely from school. However, while female headed households may be better off than they had previously been when male-headed, in most societies, they are still disadvantaged by the widespread discrimination against women and against single parent households. Women headed households face greater difficulties than male headed households because of the discrimination women face in, for instance, labour markets and in access to credit, housing and basic services. And single parent households, most of which are female headed, also face the difficulties of one adult having to combine income-earning with household management and child rearing and this generally means that the parent can only take on part time, informal jobs with low-earnings and few if any fringe benefits.

SOURCE: Chant, Sylvia, "Gender aspects of urban economic growth and development", Paper prepared for the UNU/WIDER Conference on "Human Settlements in the Changing Global Political and Economic Processes", Helsinki, 1995, 47 pages.


An urbanizing world

The last few decades have brought enormous changes to the world's settlements - cities, smaller urban centres and villages. They include new forms of city and metropolitan areas, some of unprecedented size. The average population of the world's 100 largest cities was over 5 million inhabitants by 1990 compared to 2.1 million in 1950 and less than 200,000 in 1800. The last few decades have also brought a world that is far more urbanized, and with a much higher proportion living in large cities and metropolitan areas. Soon after the year 2000, there will be more urban dwellers than rural dwellers worldwide. There are also tens of millions of "rural-dwellers" who may live in settlements designated by censuses as "rural" but who derive their livelihood from work in urban areas or who work in industries or service enterprises located in "greenfield sites." In the North, most of these "rural dwellers" also have homes that enjoy the quality of infrastructure and service normally associated with urban locations - for instance piped water and water-borne sanitation and the regular collection of garbage.

Table 1.3 shows how the world's total, urban and rural populations were distributed among the regions in 1990. The extent to which Asia dominates each of the categories is particularly notable - with three fifths of the world's population and just under three quarters of its rural population. Despite being predominantly rural, it still had more than two fifths of the world's urban population and more than two fifths of the world's population in cities with 1 or more million inhabitants (hereafter called "million cities"). Table 1.3 also highlights the scale of Africa's rural population - which in this year was larger than the rural population of Europe, North America, Oceania, Latin America and the Caribbean put together.


Table 1.3: The distribution of the world's total, rural and urban populations and its largest cities, 1990


Total population

(millions of inhabitants)

Percent of the world's:

Number of the world's

Total population

Rural population

Urban population

Population in "million cities"

"million cities"

"mega cities"

































Latin America & the Caribbean








North America
















NOTES: The figures for the number of "million cities" and "mega-cities" and the population in "million cities" should be treated with caution, as the criteria used in setting boundaries for cities or metropolitan areas varies greatly between nations. Some cities with more than a million inhabitants within their urban agglomeration or metropolitan area do not become "million-cities" as the population for their "city" as reported by governments to the United Nations does not contain a million inhabitants. There are also cities whose population may have exceeded 1 million inhabitants by 1990 but which have not been recorded by the United Nations Population Division in their 1994 revisions.

SOURCE: United Nations, World Urbanization Prospects: the 1994 Revision, Population Division, New York, 1995. The number of "million cities" and their populations have been adjusted, where new data was available.


What underlies urbanization

Urbanization trends during the 1980s have to be understood within the longer term changes of a world population that has been urbanizing rapidly for several decades. Urban populations did grow rapidly in most parts of the world during the 1980s - but they actually grew faster during the 1950s. Population growth rates for most major cities in both the North and the South were slower during the 1980s, compared to the 1970s and 1960s. And for many of the world's largest cities, including many in the South, more people moved out of the city than in during the last inter-census period.

Recent trends in urbanization reflect economic and political changes, some long-rooted, some of more recent origin. For instance, the steady increase in the level of urbanization worldwide since 1950 reflects the fact that the size of the world's economy has grown many times since then and has also changed from one dominated by relatively closed national economies or trading blocs to one where most countries have more open economies and where production and the services it needs (including financial services) are increasingly integrated internationally. In 1950, most of the world's workforce worked in agriculture; by 1990, most worked in services. The period since 1950 has brought not only enormous changes in the scale and nature of economic activity but also in the size and nature of households, in the scale and distribution of incomes within and between nations and in the scale and nature of government. All, inevitably, influence settlements patterns.

Perhaps the most fundamental influences on the world's settlement system in recent decades have come from the unprecedented changes in economic and political conditions. Section 1.1 outlined the main economic changes that meant that many countries in both North and South had the value of their per capita income multiply several times since 1960. In general, as the final section in this chapter describes in more detail, the countries with the most rapidly growing economies since 1950 were generally those with the most rapid increase in their level of urbanization while the world's largest cities are heavily concentrated in the world's largest economies. On the political front, virtually all the former colonies of European powers gained independence since 1950 and the political changes that decolonization also meant major changes to settlement systems - especially through the concentration of economic and political power in national capitals and, in some nations, the removal of migration controls on indigenous populations. The more recent economic, social and political changes in East and Central Europe and the former Soviet Union and in South Africa are also fundamentally changing settlement patterns, although these changes are too recent to have been captured by available census data.

Many countries also experienced an unprecedented growth in their national populations. Over 20 nations with one million or more inhabitants in 1990 had populations that had more than tripled since 1950 - most of them in Eastern Africa, Western Asia and Central America. Although rapid population growth does not, of itself, increase the level of urbanization, in most nations in the South, it is the most important factor in increasing urban populations. Even in regions with much less dramatic population growth and political change such as Western Europe and North America - there have been major demographic changes - for instance in age structures and household sizes and types - and as Chapter 2 will describe, these have contributed to major changes in settlement patterns.

But amidst rapid economic, social, political and demographic change, there are some elements of continuity in settlements. The average size of the world's largest cities may have changed enormously but their location has changed much less. For instance, in most of the world's regions, there is a perhaps surprising continuity in the list of the largest cities and metropolitan areas; more than two thirds of the world's "million-cities" in 1990 (ie those with a million or more inhabitants) were already important cities 200 years ago while around a quarter have been important cities for at least 500 years. In Latin America, most of the region's largest cities today, including virtually all national and most provincial capitals were founded by the 18th century with most of the largest cities founded by the year 1580 AD. In Asia, close to 90 percent of all its cities that had a million or more inhabitants in 1990 had been founded by 1800 AD and around three fifths were already important cities by that date. In countries or regions with long urban histories, there is often a comparable continuity, even for small market towns. For instance, studies of various regions of India have also found that most of the urban centres, even down to small administrative centres and market towns, have long histories. In most countries, settlement patterns have changed much less than the degree of economic, social and demographic change might imply. One reason is that existing cities represent a considerable concentration of human and physical capital and it is difficult for new cities to arise that can compete with the old ones. Another is that in most instances, cities adapt to changing circumstances. A third is that most major cities are also important administrative centres, and this generally provides some economic stability.

A predominantly urban world?

Perhaps too much emphasis is placed on the impending transition of the world to where the number of urban dwellers will exceed the number of rural dwellers. There are two reasons for caution. The first is the extent to which this transition could be hastened or delayed by changes in definitions. It would only take China, India or a few of the other most populous nations to change their definition of urban centres for there to be a significant increase or decrease in the proportion of the world's population living in urban centres. The proportion of the world's population currently living in urban centres is best considered not as a precise percentage (ie 45.2 percent in 1995) but as being between 40 and 55 percent, depending on the criteria used to define what is an "urban centre". What is perhaps more significant than the fact that more than half of the world's population will soon be living in urban centres is the underlying economic and social changes it reveals - that a steadily declining proportion of the world's population make a living from agriculture, forestry, hunting and fishing.

The second reason for caution arises from the confusion between the terms "urban centre" and "city". Many people have commented that more than half the world's population will soon be living in cities - but this is incorrect since a significant proportion of the world's urban population live in small market towns and administrative centres. Certainly, the proportion of the world's population living in "cities", however defined, is substantially smaller than the proportion living in urban centres of all sizes.

The definition of city and metropolitan populations

Caution is also needed when considering the population of individual cities - or comparing populations between different cities - since the size of a city's population depends on the boundaries chosen. For instance, the current population of most of the world's largest urban areas including London, Los Angeles, Shanghai, Beijing, Jakarta, Dhaka and Bombay can vary by many million inhabitants in any year, depending on which boundaries are used to define their populations. Different boundaries also mean different population growth rates - so London, Los Angeles, Tokyo, Buenos Aires or Mexico City can be correctly stated as having populations that are declining and expanding in recent decades, depending on which boundaries are chosen for defining their populations.

"Even the citizens of a city use its name in different senses, depending on context. A Parisian may at times think of Paris as the central département of that name - largely corresponding to the area of 19th-century Paris. At other times (especially if the Parisian in question is an inhabitant of, say, suburban Seine St Denis), the continuously urbanized area, including the surrounding départements, may be included in the implied definition. At yet others, the whole Paris region - the Ile de France - may be the intended sense."

It is also difficult to compare the populations of major cities "outside the central city" because of the different ways in which the areas outside the "central city" can also be subdivided - for instance into inner suburbs and outer suburbs - and often into "exurbs" or population living in the metropolitan area but outside the built up area of the city itself. In many countries, the boundaries of the metropolitan area may be extended further to include urban centres or other settlements from which a considerable proportion of the economically active population commute to work in the metropolitan area or whose work is tied into the economy of the metropolis. Boundaries may also be set for "extended metropolitan regions" which provide a useful regional planning framework for public authorities but which often encompass significant numbers of rural inhabitants and have a significant proportion of the labour force working in agriculture. Comparing the size, population growth rate or density of the central city population of one city with the size, population growth rate or density of another city but for its metropolitan area or "extended metropolitan region" will produce dramatic contrasts but this is not comparing like with like. Even comparisons between two cities in terms of their "central cities" can be invalid as in one city the central city refers to a very small central "historic city" while in the other, it refers to a much larger area.

This is why any international list of "the world's largest cities" risks great inaccuracy as some city populations are for large urbanized regions with thousands of square kilometres while others are for older "city boundaries" with a few hundred square kilometres. It only needs a few cities to change the basis by which their boundaries are defined for the list of the world's largest metropolises to be significantly altered. For instance, London has long dropped off the list of the world's largest cities since it had less than 7 million people in its metropolitan area in 1991 - but it has a population of 12.53 million if considered as a metropolitan region. One of main reasons that Shanghai's population appears so large - at over 13 million - is that this figure is the population in an area of over 6,000 square kilometres which includes large areas of highly productive agriculture and many villages and agricultural workers. The same is true for Beijing and Dhaka - see Table 1.4. "Metropolitan Toronto" in 1991 can have between 2.2 million and 4.8 million inhabitants, depending on the boundaries used. Figures for the population of Katowice in 1991 can vary from 367,000 to nearly 4 million for similar reasons - and Tokyo can have anywhere between 8 and 40 million inhabitants.


Table 1.4: Examples of how the populations of urban areas change with different boundaries

City or metropolitan area




Area (square







c. 5,400,000







4 inner city districts including the historic old city

"Core city"

Inner city and inner suburban districts

Inner city, inner and outer suburban districts and 8 counties











Historic city

Dhaka Metropolitan Area (Dhaka City Corporation and Dhaka Cantonment)

Dhaka Statistical Metropolitan Area

Rajdhani Unnayan Kartripakhya (RAJUK) - the jurisdiction of Dhaka's planning authority

Katowice (Poland)





The city

The metropolitan area (Upper Silesian Industrial Region)

Katowice governorate

Mexico City






c. 18,000,000




The central city

The Federal District

Mexico City Metropolitan Area

Mexico City megalopolis












The central city (23 wards)

Tokyo prefecture (Tokyo-to)

Greater Tokyo Metropolitan Area (including Yokohama)

National Capital Region.














City of Toronto

Metropolitan Toronto

Census Metropolitan Area

Greater Toronto Area

Toronto CMSA equivalent











The original 'city' of London

Inner London

Greater London (32 boroughs and the city of London)

London "metropolitan region"

Los Angeles











Los Angeles City

Los Angeles County

Los Angeles-Long Beach Primary Metropolitan Statistical Area

Los Angeles Consolidated Metropolitan Area

To address the large physical expansion of major cities and the fact that their labour market may be considerably larger than the size of the urban agglomeration (for instance with settlements outside the urban agglomeration having mostly daily commuters to the urban area) new concepts have been developed as the basis for measuring urban populations. These include functional urban regions and statistical metropolitan areas. These definitions seek to include in a city's total population all nearby settlements whose economies or residents can be considered too intricately related to the city core to be considered a separate settlement. The U.S. definition of "Standard Metropolitan Areas" delimits what can be called "functional regions" that include suburbs and other settlements outside the core city's built up area if the proportion of residents commuting to work in the central core exceeds a certain level and if the settlement meets other criteria of metropolitan "character" (eg population density). In Europe, the tendency seems to be more to view metropolitan areas as areas that conveniently encompass all the continuously built up area of a metropolis but that exclude large and small urban centres and rural settlements nearby, even though their housing and labour markets are fully integrated into those of the metropolitan centre. Extending metropolitan definitions in Europe to encompass "functional urban regions" would considerably increase the population of most major cities - and for some would include many small cities, urban centres and villages at distances of 50 or even 100 kilometres from a large city, especially those close to railway stations with rapid services to the city.

No definition of urban areas is necessarily superior; each serves a particular purpose. Without more-or-less standardized definitions, however, international comparisons of urbanization may be highly misleading.

The world's largest cities

Table 1.5 lists the world's 30 largest urban agglomerations in 1990, and it shows that the population of most of them actual grew quite slowly during the 1980s - in terms of the annual rate of growth. Only two among these 30 cities (Dhaka and Lagos) had annual average population growth rates that exceeded 5 percent during the 1980s - and it may be that Lagos actually had a much slower population growth rate than that shown in Table 1.5. Most of the "million-cities" with the highest population growth rates during the 1980s were not among the world's largest cities. Many examples will also be given in Chapter 2 of relatively small cities that had much more rapid population growth rates than the cities listed in Table 1.5 during the 1980s.

Table 1.5: The World's Largest Urban Agglomerations in 1990




Population (thousands 1990)




A.A. increment in population 1980-1990



A.A. growth rate






New York

Mexico City

Sao Paulo




Los Angeles

Beijing (Peking)



Buenos Aires




Rio de Janeiro












































































The population would be c. 31.6 million if Greater Tokyo Metropolitan Area was taken - see Table 1.5


19.3 million in the CMSA in 1990


Would be several million larger if considerd as a polynucleated metropolitan region


The population is for a large metropolitan region


This is the population within a large metropolitan region


One reason for its relatively rapid growth 1980-90 was a considerable expansion of its boundaries


14.53 million in the CMSA in 1990


Population for a large metropolitan region; the core city has a much smaller population - see Section 2.6


Urban agglomeration


Urban agglomeration




The population would be larger if measured as a Standard Metropolitan Economic Area.


Urban agglomeration


Urban agglomeration


This is the population for a large metropolitan region; the core city has about half this population.


The population of the wider metropolitan region is almost twice this


Urban agglomeration


The population in the "Greater Cairo Region" is several million larger than this.


Urban agglomeration


Urban agglomeration


Some local estimates suggest this is rather low


This appears high in comparison to the 1991 census


The population could be 12.5 million within a metropolitan region boundary


8,240,000 in the wider CMSA


Urban agglomeration


Lima-Callao metropolitan area


Urban agglomeration


Urban agglomeration


This is for Bangkok Metropolitan Area; Greater Bankok Area or the metropolitan region have several million more


Statistical metropolitan area

































- 41










































- 0.5











SOURCE: United Nations, World Urbanization Prospects: the 1994 Revision, Population Division, New York, 1995

But population growth rates can be misleading in that the larger the city population at the beginning of a period, the larger the increment in population has to be to produce a high population growth rate. Table 1.5 also includes figures for the annual average increment in the city's population which reveals that only one urban agglomeration, Bombay, had a population that had an annual increase of more than 400,000 during the 1980s - and a large part of this is the result of much extended boundaries being used in the 1990 census compared to the 1980 census, with many people counted as part of its population in 1990 that had been excluded in 1980. Three urban agglomerations had populations that grew by more than 300,000 a year between 1980 and 1990: Tokyo, Jakarta and Lagos - although this may overstate the scale of population growth in Lagos. Again, some caution is needed in interpreting these figures, as the population growth rate or the annual average increment in population will depend on which city or metropolitan boundary is chosen. For instance, if for Los Angeles it is the Consolidated Metropolitan Area that is taken, the average increment in population during the 1980s was also over 300,000.

If a much larger sample of the world's largest cities are considered - the 281 or so "million cities" that existed by 1990 - Table 1.6 (not available in this document) )shows the much increased proportion of each region's population that lives in million-cities. This Table also highlights some of these regional shifts in the concentration of the world's urban and large city population. The much increased role of Asia within the world economy since 1950 is reflected in the sharp increase in its concentration of the world's urban population, "million city" population and proportion of the world's largest cities between 1950 and 1990. But this is not so much a new trend as a return to what had been the case in previous centuries. Historically, Asia has long had a high concentration of the world's urban population and has always had most of the world's largest cities. As Table 1.6 shows, Asia had more than three fifths of the world's largest 100 cities in 1800 AD. Many of Asia's largest cities in 1990 had long been among the world's largest cities - for instance Tokyo, Beijing (formerly Peking), Guangzhou (formerly Canton), Istanbul and Calcutta. Most of the other that grew to become among the largest cities in the region were also cities with long histories. Most were the major cities either in the most populous countries or in the most successful economies - for instance Kyoto in Japan, Seoul and Pusan in South Korea, Karachi and Lahore in Pakistan, Delhi, Bombay and Bangalore in India, Jakarta and Surabaya in Indonesia and many of China's major cities.

The Americas, as a region, increased its share of the world's urban population, and "million city" population and share of the world's largest cities between 1950 and 1990 but with an important intra-regional shift. In 1950, Northern America had most of the urban and "million city" population; by 1990 this was no longer so. This is also not a new trend but a return to what existed prior to rapid industrialization in Northern America in both pre-Columbian and colonial times when most of the urban population and major cities in the Americas were in Central and South America. There was also a strong contrast within the Americas between the major industrial centres of Northern America that been among the world's largest cities in 1950 and were no longer so in 1990 - and the growing prominence of the major cities in Latin America's two largest economies - Mexico and Brazil - especially their largest industrial concentrations, Mexico City and Sao Paulo. This does not imply a lack of rapid urban change in Northern America; as Chapter 2 describes, many of the relatively new cities in the South and West of the US have been among the world's most rapidly growing cities since 1950.

In Africa, there was a rapid increase in the proportion of the world's urban population and 'million-city' population between 1950 and 1990, although the speed of change appears particularly high in all but Northern Africa because it had such a small base in 1950. One reason why it began from such a small base is that colonial governments in strictly controlled the right of Africans to live in urban centres.

In Europe, the rapid decline in the region's share of the world's urban population, 'million-cities' population and share of the world's 100 largest cities between 1950 and 1990 is particularly striking. Part of the reason is the dramatic decline in the relative importance of what were among the world's largest industrial centres in 1950 - such as Naples in Italy, Hamburg and Dusseldorf in Germany and Birmingham and Manchester in the United Kingdom. But another important reason is the much slower rate of natural increase. Europe was the first of the world's regions to begin a rapid and sustained increase in its population, as birth rates came to regularly exceed death rates but also the first region to undergo a rapid decrease in birth rates to the point where the total population is hardly growing or even declining in many European countries.


But once again, the problem of urban boundaries is evident here. For instance, the number of million cities in Europe is considerably increased - and the share of the total population they concentrate considerably increased - if urban populations are defined by functional regions rather than by city cores. For instance, Glasgow in the United Kingdom is not included in the United Nations list of "million cities" but had more than a million in its functional urban region while Newcastle and Liverpool would both have more than a million in 1990, if their boundaries were for metropolitan regions (see section 2.4 for more details)

Although most of the large cities in the North listed in Table 1.5 had very slow population growth rates during the 1980s, not all the world's most rapidly growing cities are in the South. This is especially the case if the growth of city populations is considered in this century - during which the United States has had many of the world's most rapidly growing large cities. For instance, Nairobi is often held up as one of the world's most rapidly growing cities - but both Miami and Phoenix in the United States had larger populations than Nairobi in 1990, yet all were small settlements in 1900. The population of Los Angeles was around one tenth that of Calcutta in 1900 yet in 1990, it had about the same number of people in its metropolitan area (see Table 1.5). These and other examples of rapidly growing cities in the North do not alter the fact that most of the large cities in the world with the fastest population growth rates are in the South but it does suggest that the rate of growth of the largest cities in the South is not unprecedented.

Over the last few decades, there have also been important rearrangements of population and production within cities, metropolitan areas or wider "city regions." For many of the world's largest cities, part of the slowdown in their population growth is explained by a rapid growth in production and population just outside their boundaries - and with much of this production intimately connected to enterprises still within its boundaries. In general, all major cities or metropolitan centres experience a decentralization of population and of production, as they grow. This generally begins with suburban housing being developed at ever greater distances from the city centre and then a widening commuting field and an increasing concentration of enterprises in suburban locations or in belts around the metropolitan area. But the speed of this decentralization of people and enterprises and its spatial configuration seems to vary greatly from city to city and to change over time. There are also recent examples of city centres attracting new enterprises other than those that concentrate in central business districts and also new residents.

This rearrangement of production within cities is perhaps best understood in terms of three sets of factors with different spatial implications:

- the factors that encourage a movement of enterprises out of major cities, metropolitan areas or even wider metropolitan regions, discourage new ones locating there, and cause a decline in the enterprises that are concentrated in central cities;

- the factors that still concentrate enterprises within or close to metropolitan areas (or urban regions) but outside the central city; and

- the factors that encourage enterprises back into central cities.

The scale and importance of the first of these sets of factors has long been evident and carefully documented. It includes advances in transport and communications and the fact that an increasing proportion of the national territory, including many smaller cities, had a skilled and literate workforce, good quality transport and communications and adequate quality infrastructure and services. In many countries in both the North and the South, large incentives were offered to attract manufacturing to smaller urban centres or poorer regions.

What is less well documented, especially in the South, is the fact that it has become common to have a decentralization of population and of production away from the central city (and even its inner suburbs) but with a continued or increased concentration of population and of production within the metropolitan area or wider region. Judged nationally, this remains an increasing concentration of population and economic activity within what might be termed "core region" but a decreased concentration of population and economic activity within the core region itself. In most major cities in the North and many in the South, there has been a declining proportion of the population living in the central city and the outward sprawl of the urban agglomeration, especially along major roads and highways. This has probably gone furthest in the United States where population densities in outer suburbs are generally much lower than in Europe and where many enterprises have also developed close to major highways outside the central areas. Some of the most innovative and successful concentrations of enterprises are not in cities but concentrated along major highways (for instance the firms along Route 128 in Massachusetts) or in Silicon Valley. This has led to new terms such as the "100 mile city" where there is no obvious "central city." This decentralization of production and of urban population within core regions can also be seen in the rapid growth of smaller cities that are close to the major cities or metropolitan areas as these attract both industrial and service enterprises that previously would tend to concentrate in major cities. Regions with good quality transport and communication networks encourage this, as does the "just-in-time" system which needs some physical proximity.

One of the most remarkable examples of this is the region bounded by Sao Paulo, Belo Horizonte, Rio de Janeiro and Porto Alegre in Brazil which includes a great range of cities of different sizes, many of which have been successful in attracting new enterprises that previously would have concentrated in the major cities. Many such cities are within 200 km of Sao Paulo's central city and have many advantages and few disadvantages when compared to investment in Sao Paulo itself. Private investment in Greater Sao Paulo was discouraged by the well-organized trade union movement, pollution, transport problems and a lack of suitable land sites. Investment in cities nearby was encouraged by a whole range of factors including better roads and telecommunications, the availability of land and fiscal incentives in the smaller cities. The decentralization of production was also boosted by the promotion and support of the government in information technology in Campinas and in the aeronatics industry in Valle del Paraiba. Motor vehicle companies such as General Motors, Volkswagen and Mercedes Benz have chosen different small cities within this region. However, this model of a region with a diverse but highly integrated urban system and a declining importance for any dominant city depends on good transport and communications system and good quality infrastructure and services available in different cities. Many of the largest cities in the South remain more concentrated, because of poorer transport and communications system and as scarce infrastructure and skilled labour tend to be concentrated in large cities so modern economic activity clusters around them.

There are also examples of successful cities once again attracting employment and residential populations back to central areas or to inner suburbs. There are also examples of cities that have transformed their economic prospects, after facing economic decline - for instance Barcelona and Baltimore. One important reason why, as noted in a survey of cities in The Economist in 1995, is that financial markets like traditional cities in which enterprises are concentrated, even if advanced telecommunications allow many routine tasks to be located outside the financial district. For banks and other financial services, "face-to-face contract is irreplaceable, partly because it promotes the trust that is essential to make deals, and partly because the informal exchange of ideas in such businesses is unpredictable.... serendipitous proximity cannot be reproduced by fax machine or video-conferencing." The concentration of banks in turn encourages a concentration of service enterprises associated with banks or their staff. In addition, many other service activities also find advantage in concentration, including design, marketing, advertising, film and television.

The "world cities"

The structural changes within the world economy described earlier have helped to reorder the relative importance of cities around the world and, for many cities, to re-shape their physical form and the spatial distribution of enterprises and residents within them. Regions and cities have proved more flexible than nations in adapting to changing economic conditions - and certain key regions and cities have become successful locales of the new wave of innovation and investment - for instance Silicon Valley and Orange County in California, Arizona, Texas and Colorado in the western United States, Bavaria in Germany, the French Midi - from Sophia-Antipolis via Montpellier to Toulouse, certain cities in southern Europe, and, of course, many cities in the Asian Tigers and in China. Meanwhile, most major port cities and traditional industrial centres in the North (and some in the South) that grew to become among the world's major cities during the second half of the nineteenth and first half of the 20th century lost importance - and most had population declines during the 1980s.

The large and increasing share of the world's economy controlled by multinationals has led to certain cities becoming what are often termed "global" or "world" cities - as they are key command and control points of the world economy. Some are unambiguously world cities as they are the "commanding nodes" of the global system - London, New York and Tokyo. New York and London are not only leading financial markets but also leading producers and exporters in accounting, advertizing, management consulting, international legal services and other business services. Tokyo - developing as an important centre for the international trade in services such as construction and engineering services, has developed beyond its initially restricted role of exporting only the services required by its larger international trading houses. It is interesting to note that Tokyo also remains one of the world's major concentrations of industrial production whereas New York and London do not. It is a reminder of how advanced telecommunications systems and logistical management of production can separate almost completely in spatial terms the production process from those who manage and finance it and the "producer services" that the managers and financiers need.

Another category of world cities is the cities that articulate large national economies into the system (Paris, Madrid, Sao Paulo) or subnational (regional) economies (Chicago) or simply have a commanding multinational role (Miami, Singapore). Global cities are not necessarily the same as large cities - as many cities that are not among the world's largest have major international roles - for instance Singapore and Zurich. Several of the world's largest cities do not owe their size and economic base to their role within global production but to being national capitals in more populous nations with a high concentration of political power there - for instance Delhi and Cairo. Several other cities that are among the world's thirty largest listed in Table 1.5 derive their size more from their national role. It is interesting to note that the former national capitals of the most populous nation in Africa (Lagos in Nigeria) and Latin America (Rio de Janeiro in Brazil) are among the 30 largest cities in the world - and both owe much of their size and economic importance to the period when they were political capitals. Calcutta too grew to be much the largest city in India as the former capital; Delhi, to where the national capital was moved early in the 20th century, has grown much faster than Calcutta in recent decades and may soon have a larger population than Calcutta.

A few of the world cities are capitals (London, Paris, Tokyo) but most are not, for instance New York, Frankfurt, Zurich, Amsterdam, Sydney, Los Angeles, Osaka and Toronto. World cities have to adapt to changing circumstances to stay as world cities. For instance, Singapore's position could have been threatened by the rapid growth of the Thai and Malaysian economies but it chose to keep high wage jobs in dominant financial and producer services for itself while creating labour intensive activities in manufacturing and tourism in regions just outside its boundaries.

Internal and international migration

The scale, direction and nature of migration flows are so fundamental to understanding urbanization and, more generally, changes in the spatial distribution of population that it is surprising that internal migration receives so little attention. In most countries in the South, there has been relatively little recent research on the scale and nature of internal migration flows. There is much more work on the scale and nature of international migration flows, even though there is far more internal population movement than international population movement. The much increased volume of research on international migration is no doubt linked to political worries of governments in wealthy countries - for instance for countries close to areas of conflict or potential conflict that might create large flows of refugees and other emigrants and for wealthy countries that are close to low-income countries and with borders over which it is difficult to control population movements.

In most countries, there is very little idea of the scale and nature of internal migration flows, except that they are almost always diverse, complex, constantly changing and include rural to urban, urban to rural, urban to urban and rural to rural flows. It is often assumed that rural to urban migration is the dominant migration flow but rural to rural flows may be of a larger scale - and in many countries, urban to urban migration flows outnumber rural to urban flows. In virtually all nations, there are also large urban to rural migration flows, even in countries which are urbanizing; the fact that the nation is urbanizing merely reflects the fact that rural to urban migration flows outnumber migration flows in the opposite direction.

There is also considerable diversity between nations and regions in the migrants themselves - for instance in terms of age, level of education, extent to which the move is considered permanent or temporary and the extent to which this move is part of a complex and diverse household survival strategy. Recent studies have highlighted the extent to which migration patterns are also differentiated by gender. Various studies have shown how female migration is of much greater volume and complexity than was previously believed and also how the migration of women differs in many ways from that of men in its form, composition, causes and consequences.

There is also great diversity in the scale and nature of migration. There are around 30,000 urban centres in the South and each has its own unique pattern of in-migration and out-migration that constantly changes, reflecting (among other things) changes in that centre's economic base, labour market and age structure. It also reflects social, economic and political changes within the region and nation and is influenced by such factors as crop prices, land-owning structures and changes in agricultural technologies and crop mixes in surrounding areas and distant regions. Each detailed study of migrants in urban settings and of conditions in areas of out-migration reveals a long list of factors which influence migration, including: those relating to individuals or household structures and gender-relations within households; local social, economic and cultural factors; regional and national social and economic change; and international factors. In each location, the relative importance of the different factors is subject to constant change. This cautions against seeking too many generalizations and general recommendations in regard to rural-urban migration.

One example of this diversity is the differences in the scale and nature of demand for female labour. For instance, in South East Asia, the demand for women workers in multinational industries and in unskilled and semi-skilled service occupations (domestic service, informal commerce, sex) is important in drawing young women to cities in Taiwan, Philippines, Thailand, Malaysia and South Korea.

Although it is often assumed that the largest migration flows are in countries in the South, in fact, there is probably more migration in the United States in any one year in terms of the proportion of the population who move than in virtually all countries in the South. For instance, it is estimated that one person in five in the USA moves each year and as Chapter 2 will describe, the restructuring of the urban system there during the 1980s was more rapid and fundamental than that taking place in most countries in the South. This does not show up as rapid change in the level of urbanization as most migration is urban to urban. Chapter 2 will also describe the very rapid change in the urban system in China and here too, the change in the proportion of the population living in urban areas gives little indication of the scale of the change.

The scale of international migration has certainly increased over the last ten to fifteen years. Estimates for 1992 suggest that over 100 million people lived outside their own country of which 20 million were thought to be refugees and asylum seekers. An estimate for 1990 suggested that around 15-20 million were in western Europe, around 15-20 million in North America, and 2-3 million were in the wealthiest nations in Asia including Japan, Taiwan and Singapore. There are also large populations of foreigners within the wealthier countries in Africa and Latin America, most of them being from lower-income countries closeby. As in internal migration flows, there is great diversity in the types of international migrants and the forms that their movements take - contract labourers, students, professionals and skilled workers, immigrants joining their families through family reunification, people who retire to a foreign country - and asylum seekers and refugees. Their scale and nature is greatly influenced by the attitude of the receiving country to immigrants and the provisions made to allow or control immigration.

Among these different categories of international migrants, the growth in the number of refugees has been most dramatic - by 1994, 23 million people qualified as refugees compared to about 2.5 million 20 years ago. There has also been a considerable growth in international migration flows of highly qualified or skilled labour migrants, that include the professional and managerial staff transferred within the international labour markets of transnational corporations. For instance, in 1988, 83,000 Japanese were assigned to work in overseas branches of Japanese companies while a further 29,000 (not including students) went overseas to engage in scientific study and research.

There is also great diversity in the scale of emigration out of and immigration into each nation - and also in who moves. A distinction can be drawn between countries where there is far more immigration than emigration - generally the wealthiest countries that also have relatively open policies to immigrants - and countries with high levels of emigration. There are also the exceptional cases of certain oil-exporting Middle East countries that have foreign workers making up more than half their workforce - for instance this was the case in 1990 in Bahrain, Kuwait, Saudi Arabia, Qatar and United Arab Emirates. Other OPEC nations have generally been centres for immigration within their own region, although usually less so in the 1980s and early 1990s as the real price of oil fell. Singapore was also reported to have 11 percent of its labour force made up of foreign workers - mainly from Malaysia, Thailand, Indonesia and the Philippines. Although there is little recent data for most African countries, in nations such as Congo, Zaire and South Africa, immigrants made up more than 5 percent of their population in mid 1980s.

Many countries have several percent of their population or labour force living abroad - including Jordan, Russia, Burkina Faso and Egypt with more than 10 percent of their population or labour force living abroad (although for Russia, most are in republics that were formerly part of the Soviet Union). A considerable proportion of the population of Central America (excluding Mexico) live outside their country of origin. In some instances, this is associated with war or civil strife - as in the huge emigration flows out of Afghanistan, Mozambique and most Central American countries, although in recent years, in all these countries, many emigrants (including refugees) have now returned. An increasing number of countries from which there were large emigration flows have changed to become major centres for immigrants - for instance most countries in Southern Europe, and several Central and East European states - particularly Hungary, Poland and the Czech Republic. There are also many countries with large emigrant and immigrant flows - for instance with large emigration movements of skilled and professional emigrants and with unskilled immigrants coming in from lower income countries.

Many of the main migration flows have well-established historical roots. A considerable part of them are also a consequence of an increasingly globalized economy. Between 25 and 30 million migrants are thought to be foreign workers, most of whom will return to their own countries. An estimate for 1991 suggested that the total value of their remittances back to their own countries was $71 billion and if this is an accurate estimate, then it means that total remittance flows are larger than total aid flows and makes remittance flows one of the largest items in international trade.

Below, are outlined some of the important changes in international migration during the last ten to fifteen years. This is not a comprehensive list, but it does illustrate how international migration flows have to be understood within specific national and regional contexts.

The removal of the Iron Curtain, the break-up of the former Soviet Union and the changes in international boundaries in East and Central Europe has changed the scale and nature of international migration in this region. Migration flows have been largest to Germany, although as Chapter 2 will describe, a large part of this is ethnic Germans moving to Germany and these are not considered by the German government as immigrants but citizens. There have also been large migration flows from the former Soviet Union to Israel and North America - and also from Hungary and Poland to North America. Changing boundaries are also changing people's status from "nationals" to "immigrants" as in the splitting up of Czechoslovakia, Yugoslavia and the division of the USSR into many independent republics.

The large scale recruitment of foreign workers in the oil-rich Middle East. Most were admitted on strictly limited contract terms, were not allowed to bring dependents and were expected to return to their home country when their contract finished. In the mid 1980s, with the drop in the real price of oil, hundreds of thousands of Arab, South and East Asian workers lost their jobs and returned home. Similarly, large numbers of foreign workers lost their jobs or were expelled during and after the Gulf War. However, most oil-rich Middle Eastern countries still rely heavily on foreign workers. There are considerable differences in the kinds of migrants coming from different countries - for instance, a higher proportion of skilled and professional migrants come from certain countries, while most women workers come from Philippines, Indonesia, Thailand, Korea or Sri Lanka.

The very large increase in the number of refugees in Africa, associated with the wars and civil strife there - for instance the refugee flows from Rwanda, Burundi, the Sudan and Somalia. Mozambique was the single largest source of refugees in the early 1990s but many have returned in the last few years. During the 1980s, there were also mass expulsions of foreign workers at particular times in countries such as Nigeria, Congo and Mauritania, linked to economic recession when in previously more prosperous periods, they had been tolerated.

The changing nature of international migration in Asia as it grew rapidly and also became more concentrated within Asia, between the wealthier and the poorer nations, whereas previously it had been more oriented to nations outside Asia. Within Asia, it is possible to distinguish between labour importing countries (Japan, Singapore, Taiwan, Brunei), countries which import some types of labour but export others (Hong Kong, Thailand, Malaysia, Republic of Korea) and countries which are predominantly labour exporters (China, Philippines, India, Bangladesh, Pakistan, Sri Lanka, Indonesia).

In Latin America, there is uncertainty as to whether international migration increased during the 1980s and early 1990s. Certain factors led to a decrease - for instance, as the economic crisis lessened economic differences between countries and with the reduction in population displacements from war or civil strife in Central America. The return to democracy in many countries encouraged or allowed the return of many hundreds of thousands of Latin Americans to their own countries.

The changing nature of international migration to the North, with a growing importance within migration flows of high-skill workers, clandestine migrants and asylum seekers. The growth in the number of asylum seekers is perhaps the most dramatic in the last decade with the number of asylum applications to Europe, North America and Australia multiplying around sevenfold between 1983 and 1991 to reach 715,000 in 1991. It has grown considerably since then, reflecting the breakup of the former Yugoslavia; estimates for October 1993 suggest 5 million residents had been displaced from their home areas with 4.3 million moving to other places within the old borders and 700,000 having left for other countries.

The relationship between economic growth and urbanization

Rising levels of urbanization and rapid population growth in large cities has often been considered problematic because governments and international agencies fail to ensure that infrastructure and service provision keeps up with the growth in population and often fail to enforce pollution control and other regulations needed to protect the quality of life in urban areas. Although the debate about the role of cities in development and in environmental problems continues, the key role that cities have in dynamic and competitive economies was increasingly acknowledged during the 1980s and early 1990s. So too was the fact that major cities generally have a significantly higher concentration of the nation's economic output than of its population. In 1995, Singapore's information minister suggested that "In the next century, the most relevant unit of economic production, social organization and knowledge generation will be the city.

One reason for this is that rising levels of urbanization are strongly associated with growing and diversifying economies - and most of the nations in the South whose economic performance over the last two decades is so envied by other nations are also the nations with the most rapid increase in their levels of urbanization. The nature of this relationship between the scale of the economy and the scale of the urban population is also illustrated by the fact that most of the world's largest cities are in the world's largest economies. Taking economic and population statistics for 1990, the world's 25 largest economies also had over 70 percent of the world's 281 "million cities" and all but one of its 12 urban agglomerations with 10 million or more inhabitants. The world's five largest economies on that date (United States of America, China, Japan, Germany and France) had between them half of the world's 10 million plus inhabitant urban agglomerations and a third of its "million-cities".

Some argue that, whilst increases in the level of urbanization may be high, they are not abnormal when compared to other countries at a similar stage of their development. For instance, there are examples of countries in the North which had increases in their levels of urbanization during the second half of the nineteenth century or first half of the 20th century which were as or more rapid than most nations in the South over the last 30-40 years. There are also many examples of cities in the North that had population growth rates for several decades during the 19th or early 20th centuries that were comparable to those of the most rapidly growing cities in the South over the last few decades. There are also instances of countries in the North where the overall growth rate of their urban populations was very rapid, even by contemporary standards - for instance in the United States, between 1820 and 1870, the estimated average annual growth rate for the urban population was 5.5 per cent. The urban population in Japan grew at around 6 percent a year during the 1930s. This can be compared to growth rates for the urban population in Africa of less than 5 percent a year during the 1980s - and growth rates for the urban population of Asia of less than 4 percent and for the urban population of Latin America and the Caribbean of close to 3 percent. What is unprecedented when comparing the last 3-4 decades to earlier periods is the number of countries undergoing rapid urbanization and the number of cities worldwide that are growing rapidly - although as noted earlier, the population growth rate of many cities slowed during the 1980s.

Many factors influence the scale of net rural to urban migration that in turn underlies increases in the level of urbanization. However, there is an evident relationship between changes in economic development in a country and the level of urbanization. The relationship can be seen by plotting countries' level of urbanization against per capita income. The relationship is complex, with many factors at work - and whilst economic development may result in growing levels of urbanization, higher levels of urbanization in turn can stimulate more economic growth.

It has also been argued that urban growth has been detrimental to economic growth and the term "over-urbanization" has been used to describe countries where the level of urbanization relative to national income is considered to be high in comparison to reference countries. One reason put forward for this is that, with limited investment funds, high levels of investment in urban areas will reduce investment in other productive sectors of the economy including agriculture. Another argument is that economic growth or stability may be compromised by high levels of government investment in urban infrastructure and services.

One basis for the "over-urbanization" thesis is that specific urbanization levels in the South have been achieved at lower levels of per capita income than those associated with similar levels of urbanization in the North. For example, the North was less urbanized than Asia or Africa when its per capita income was US$300. In the case of Latin America, average levels of urbanization in 1980 were similar to those in the North but the level of per capita GNP was less than one third. It has been argued that increases in the level of urbanization in the Third World have been achieved without economic growth, industrialization and increases in agricultural productivity. However, there are a number of reasons why the "over-urbanization" thesis might be questioned:

1. There are doubts about the quality and the comparability of the data for both urban populations and for per capita GNP. The margin of error for Northern urban data for 1800 and the South for 1930 is estimated to be about 6 to 8 per cent. Another problem with such international comparisons noted already is that each country uses different criteria for measuring its urban population.

2. The accuracy of the GNP per capita estimates is questionable; many Southern countries have only recorded national income since 1950 and it is recognised that a considerable part of the economy may be unrecorded because of the extent of the informal economy and other unregistered activity. The accuracy of estimates for GNP may be particularly poor when a large proportion of production is for subsistence or traded outside of the money economy.

3. One important aspect of the relationship between level of urbanization and level of per capita income of particular interest is the extent to which levels of urbanization fall (or their rate of increase slows) with a stagnant or declining economy. Historic studies suggest that city populations may not respond as rapidly to economic decline as they do to economic growth and increases in the level of urbanization as income rises may not be matched by an equal reduction in the level of urbanization as income falls. The same can be said for countries in the Southern Cone of Latin America in recent decades where increases in the levels of urbanization have slowed, as have urban population growth rates - but they did not slow as abruptly as the economy. The results of an analysis of how urbanization levels change over time in response to economic change will show a strong correlation between the two - but while economic stagnation or decline may slow increases in the level of urbanization, they do not generally halt it. In some countries in Africa, economic stagnation is reported to have encouraged some movement of city populations back to rural areas but the limited information available on this to date does not suggest a large scale movement. Even when incomes are falling, it seems to be rare that households perceive the advantages in rural areas to be sufficient to encourage them to move back on an equivalent scale. What appears more common, as will be described in Chapter 3, is household livelihood strategies that draw on both rural and urban resources.

4. In addition to an inertia to reverses in the level of urbanization, there may be other reasons to account for the high level of urbanization. One is that some economic activities such as mining increase urbanization by requiring a concentrated workforce (and usually an urban setting for their shelters) but do not necessarily increase national income by as much as had been the case in nineteenth century Europe. National income may increase little because the profits from these activities are invested overseas (and much of the processing also takes place overseas). Another reason why levels of urbanization may be higher in recent decades for any given level of per capita income is that the role of government has expanded considerably during the twentieth century and state employees live mainly in urban centres.

5. Perhaps the main reason for doubting the validity of the "over-urbanization" thesis is that it implies that the model of urban development undergone in the North represents a pattern that should be followed in the South and that any deviation from this model represents "over-urbanization". In this, it has similarities with the normative judgements made as to whether the distribution of population within urban centres in a country corresponds to particular mathematical distributions that are assumed to be "correct" or "balanced". If, in general, Latin American countries are more urbanized today relative to per capita income than countries in the North, what should be sought are the reasons why. It would be surprising if urban trends in the South followed patterns in the North. The much weaker position of most countries in the South within world markets and the fundamental differences between the world in the 19th and the end of the 20th centuries must affect the social, economic and political factors that influence levels of urbanization.

The level of urbanization in any country is influenced by many factors, both economic and social including:

The proportion of the economy that is derived from manufacturing or service industries rather than agricultural activities.

The nature of the economic activity within each sector. For example, the type of agriculture affects the scale of urban settlements. The extent to which agriculture stimulates or supports local urban development depends critically on the value of the crop, the extent to which there are local possibilities for adding value to the crop (for instance fruit juices and alcoholic beverages, jams and sweets) and the nature of land ownership. High value crops that provide good incomes for farmers and agricultural workers within relatively intensive farming systems can support rapid growth of local urban centres to the point where agriculture supports a relatively urbanized population - and can also attract new enterprises from outside the area.

The influence of land-ownership patterns is important not only for its influence on what is produced but also in where the profits generated are spent or invested. In general, the larger the farm, the less likely the value generated by the production will be spent locally. Plantations are an extreme example of agricultural production where it is usual for only a small proportion of the value generated by their production to be spent or invested nearby.

The relative importance of factors that influence levels of urbanization may change as countries become more urbanized. In advanced economies, many rural areas are urbanized as manufacturing and service enterprises can locate on "greenfield" sites with many of the service and infrastructure benefits of urban locations but none of the congestion; and city-workers can live in rural areas and travel each day to work.  In countries with a high population density, these criteria may apply to a large proportion of rural areas, as will be described in Section 2.6. This also helps explain why the relationship between per capita income and level of urbanization becomes much less obvious among the higher income countries - for instance there were large increases in per capita income for the United States and many European countries between 1970 and 1990 but relatively little increase in their level of urbanization.

To conclude, levels of urbanization for each country are therefore likely to reflect not only the level of per capita GDP and the nature of the economy but a number of other factors including definitions used for urban areas, the nature of agriculture, physical factors such as the size and typography of the country, political factors including the relative degree of security in rural locations, and cultural preferences for types of lifestyle. In addition, government policies and state institutions are a major influence on the level of urbanization. This influence is felt in a number of different ways including:

The share of national income spent by the public sector. This has clearly increased considerably in most countries during the last few decades, although much less so in the last decade or so. In part this is due to some areas that were previously provided by the private sector being drawn partially or wholly within the public sector, for example, health or transport services. Most state employees are urban residents (although not all are employed in the major cities). The importance of government employment within the urban labour force has been demonstrated by the impacts of redundancies arising from the implementation of structural adjustment programmes in the South. Assessments of the significance of this factor may take some time to emerge, in particular, because reduced employment opportunities in the public sector are likely to impact more on future rural to urban migration trends rather than result in urban to rural migration.

Government macro-economic and regional policies also impact on the level of urbanization. Governments influence aggregate national income and the distribution between the different sectors of the economy. Subsidises for agriculture or farmers in most countries in the North have helped to maintain farmers' income and investment capacity. Subsidised services in urban areas have encouraged industrial development. They also influence the relative cost of capital and labour in urban and rural areas and thereby affect employment opportunities. For example, countries in Eastern Europe have successfully kept levels of urbanization below what would have been expected relative to levels of income through a strategy of labour intensive agriculture. Investments in infrastructure influence the costs associated with transportation and telecommunications. In the North, for example, improved transport links and homebased working using advanced telecommunications have also permitted a delinking of the location of residence and employment opportunities.

It is often suggested that governments have favoured urban rather than rural areas within their investment strategies and pricing policies and that this has encouraged the migration of people from rural to urban areas and therefore an increase in the level of urbanization, although within the Indian context public sector policies do not seem to have had a significant impact on the level of urbanization. Instead, the evidence suggests that the level of urbanization in India is relatively insensitive to public sector investment strategies.

These influences on the level of urbanization suggest that countries may have very country-specific components to the relations between levels of urbanization and economic variables, although the underlying nature of such relations may be broadly similar.

Recent census information has allowed a reconsideration of the trends and relationships discussed above . The graphs in Figure 1.1(not available) illustrate the relationship between urban and economic change for certain regions. For the countries included in this analysis from South America, the wide range in the levels of urbanization is immediately evident. The eight countries fall into three groups: high levels of urbanization for the Southern Cone countries that experienced economic growth during the first few decades of the twentieth century plus Venezuela (one of the OPEC countries); intermediate levels of urbanization (Peru and Colombia), and low levels of urbanization (Paraguay and Ecuador). Considering the graphical relationship between levels of urbanization and per capita income, two characteristics stand out. The first is the "shock" to incomes during the late 1970s and early 1980s which took place with little reduction in the increase in the level of urbanization and second, the increase in urbanization in Venezuela despite a considerable reduction in per capita income.

Northern Europe (Scandinavia plus the UK and Eire) shows a very similar pattern between the two graphs indicating that most countries have succeeded in achieving small but continually rising per capita incomes (and at a similar level except Eire) during the period under consideration. The countries with high levels of urbanization show little increases in recent decades, despite the fact that incomes have continued to rise. These graphs suggest that urbanization levels will stabilize at different levels in different countries - although the different levels may simply reflect different definitions used for urban centres. There is also the fact noted earlier that in the wealthiest countries, an increasing proportion of the rural population are either urban workers (who commute to work) and work in manufacturing and service enterprises located in greenfield sites.

Southeast and East Asia and Oceania also includes some of the wealthiest nations (Australia, Japan and New Zealand) for whom urbanization levels have shown little change in recent decades. Both South Korea and Indonesia have had increasing urbanization although this trend is less evident for Fiji and the Philippines. Considering incomes, two groups appear to emerge from the second graph: a high income group which has seen very little change in urbanization levels, and second, a low income group with increasing levels of urbanization and rising incomes. While South Korea's level of urbanization at first appears to have increased very rapidly, when considered with changes in per capita income (where the country has had one of the world's most rapid increases over the last 40 years), the trends are consistent with other countries.

In general, there is a close relationship between the level of urbanization and per capita income (measured here by GDP per capita). For the time period and countries considered here, there is also a correlation between the level of urbanization and the date (ie. a time trend) in addition to the relationship with per capita GDP. This suggests that there may be other factors not specifically included in this model that are important in better understanding the process of urbanization and that have, in broad terms, been increasing over time. In this case, the time trend is acting as a proxy for a set of factors that are positively related to the level of urbanization and are correlated to the time trend but which have not been separately identified.


  1. . OECD, OECD Economic Outlook, December 1994.
  2. . World Bank, Annual Report 1994, Washington DC, 1994.
  3. . World Bank, Annual Report 1993, Washington DC, 1993.
  4. . World Bank 1994, op. cit. and World Bank, World Development Report 1994: Infrastructure for Development, Oxford University Press, Oxford, 1994, Table 2.
  5. . World Bank, World Development Report 1991, Oxford University Press, Oxford, 1991.
  6. . World Bank 1994, op. cit.
  7. . Government of Kenya, Economic Survey 1994, Central Bureau of Statistics, Government Press, Kenya
  8. . Per capita income is taken as per capita GDP in this section
  9. . UNDP, Human Development Report 1994, Oxford University Press, Oxford, 1994.
  10. . UNDP 1994, op. cit.
  11. . UNDP 1994, op. cit.
  12. . OECD, OECD Economic Outlook (Various Issues) and World Bank, World Development Report 1991, Oxford University Press, Oxford, 1991.
  13. . OECD, OECD Economic Outlook, December 1988 Issue.
  14. . Hardoy, Jorge E. and David Satterthwaite, Squatter Citizen: Life in the Urban Third World, Earthscan Publications, London, 1989.
  15. . Hardoy and Satterthwaite 1989, op. cit.
  16. . United Nations, Housing and Economic Adjustment, United Nations Publication Sale No. 88.IV.1, DIESA, 1988.
  17. . OECD Economic Outlook, December 1992 Issue.cited above
  18. . World Bank 1991, op. cit.
  19. . See Table 2.5 in The World Bank, Annual Report 1994, Washington DC, 1994, 254 pages.
  20. . This section draws heavily on Sassen, Saskia, Cities in a World Economy, Pine Forge Press, Thousand Oaks, London, New Delhi, 1994, 157 pages. It also draws on Friedmann, John, "Where we stand: a decade of world city research", Paper prepared for the Conference of World Cities in a World System, Center for Innovative Technology, April 1993, 37 pages and Friedmann, John and G. Wolff, "World city formation", International Journal of Urban and Regional Research, Vol. 6, No. 3, 1983, pp. 309-343.
  21. . Parker, John, "Turn up the lights; a survey of cities", The Economist, July 29th, 1995, 18 pages.
  22. . Sassen 1994, op. cit.
  23. . Sassen 1994, op. cit.
  24. . Sassen 1994, op. cit.
  25. . Harris, Nigel, City, Class and Trade, I.B. Tauris and Co, London and New York, 1991 , 262 pages.
  26. . Amin, Ash and Nigel Thrift, "Neo-Marshallian nodes in global networks", International Journal of Urban and Regional Research, Vol.16, No.4, December 1992, pp. 571-587 and Hay, Dennis, "On the development of cities", in David Cadman and Geoffrey Payne (Editors), The Living City - Towards a Sustainable Future, Routledge, London, 1990.
  27. . Ref 28 in globalization
  28. . Sassen 1994, op. cit.
  29. . Preliminary figure, from World Bank, Annual Report 1994, Washington DC, 1994, 254 pages.
  30. . World Bank 1994, op. cit.
  31. . Sassen 1994, op. cit.
  32. . World Bank 1994, op. cit.
  33. . Sassen 1994, op. cit.
  34. . Parker 1995, op. cit.
  35. . United Nations, World Urbanization Prospects: the 1994 Revision, Population Division, New York, 1995
  36. . United Nations 1995, op. cit.; Cleland, John, "Population growth in the 21st century; cause for crisis or celebration", Paper presented to the 21st Century Trust meeting on Population Growth, Health and Development: Problems and Prospects for the 21st Century, Oxford, 1995, 18 pages.
  37. . Cleland 1995, op. cit.
  38. . Cleland 1995, op. cit.
  39. . Varley, Ann, "Women-headed households: some more equal than others?", World Development (forthcoming)
  40. . See Bourne, L.S., Urban Growth and Population Redistribution in North America: A Diverse and Unequal Landscape, Major Report 32, Centre for Urban and Community Studies, University of Toronto, Toronto, 1995, 41 pages.
  41. . Chant, Sylvia, "Gender aspects of urban economic growth and development", Paper prepared for the UNU/WIDER Conference on "Human Settlements in the Changing Global Political and Economic Processes", Helsinki, 1995, 47 pages.
  42. . Chant 1995 op. cit., drawing from: Flandrin, Jean-Louis, Families in Former Times: Kinship, Household and Sexuality, Cambridge University Press, Cambridge, 1979; Laslett, Peter, "Introduction: the history of the family", in Peter Laslett and Richard Wall (editors), Household and Family in Past Time, Cambridge University Press, Cambridge, 1972, pp. 1-89; Mitterauer, Michael and Reinhard Sieder, The European Family, Blackwell, Oxford, 1982; Peil, Margaret, with Pius O. Sada, African Urban Society, John Wiley and Sons, Chichester, 1984; Stivens, Maila, "Family and state in Malaysian industrialization: the case of Rembau, Negeri Sambilan, Malaysia", in Haleh Afshar (editor), Women, State and Ideology: Studies from Africa and Asia, Macmillan, Basingstoke, 1987, pp. 89-110.
  43. . See Chant 1995 op. cit. for a review and a list of empirical studies on this; see also Gugler, Josef, "Social organization in the city" in Alan Gilbert and Josef Gugler (Editors), Cities, Poverty and Development: Urbanization in the Third World (Second Edition), Oxford University Press, Oxford, 1992, pp. 155-176.
  44. . Chant 1995, op. cit. page 7
  45. . Varley, forthcoming, op. cit.
  46. . Chant 1995, op. cit.; Townsend, Janet and Janet Momsen, "Toward a geography of gender in the Third World", in Janet Momsen and Janet Townsend (editors), Geography of Gender in the Third World, Hutchinson, London, 1987, pp. 27-81.
  47. . Varley forthcoming, op. cit.
  48. . Brydon, Lynne and Sylvia Chant, Women in the Third World: Gender Issues in Rural and Urban Areas, Edward Elgar, Aldershot, 1989; Chant, Sylvia and Sarah A. Radcliffe, "Migration and development: the importance of gender", in Sylvia Chant (Editor), Gender and Migration in Developing Countries, Belhaven Press, London, 1992, pp. 1-29.
  49. . Appleton, Simon, "Gender dimensions of structural adjustment: the role of economic theory and quantitative analysis", IDS Bulletin, Vol.22, No.1, 1991, pp. 17-22; Chant, Sylvia, "Single-parent families: choice or constraint? The formation of female-headed households in Mexican shanty towns", Development and Change, Vol.16, No.4, 1985, pp. 635-65; Chant, Sylvia and Cathy McIlwaine, Women of a Lesser Cost; Female Labour, Foreign Exchange and Philippine Development, Pluto, London, 1995, 372 pages (Chapters 3 and 7); Engle, Patrice, "Father's money, mother's money and parental commitment: Guatemala and Nicaragua", in Rae Lesser Blumberg, Cathy Rakowski, Irene Tinker and Michael Montéon (editors), Engendering Wealth and Wellbeing: Empowerment for Global Change, Westview, Boulder, 1995, pp. 155-179; Hoddinot, John and Lawrence Haddad, Household Expenditures, Child Anthropomorphic Status and the Intra-Household Division of Income, Unit for the Study of African Economics, University of Oxford, 1991; Kanji, Nazneen, "Structural adjustment policies: shifting the costs of social reproduction to women", Critical Health, Vol. 34, 1991, pp. 61-67; Kennedy, Eileen, "Development policy, gender of head of household and nutrition", in Eileen Kennedy and Mercedes González de la Rocha, Poverty and Wellbeing in the Household: Case Studies of the Developing World, Centre for Iberian and Latin American Studies, UCLS, San Diego, 1994, pp. 25-42; Weekes-Vagliani, Winifred, "Structural adjustment and gender in the Côte d'Ivoire", in Haleh Afshar and Carolyne Dennis (Editors), Women and Adjustment Policies in the Third World, Macmillan, Basingstoke, 1992, pp. 117-149.
  50. . Drawn from Chant 1995 op. cit., quoting Browner, C.H., "Women, household and health in Latin America", Social Science and Medicine, Vol.28, No.5, 1989, pp. 461-473; Buvinic, Mayra, Juan Pablo Valenzuela, Temistocles Molina and Electra González, "The fortunes of adolescent mothers and their children: the transmission of poverty in Santiago, Chile", Population and Development Review, Vol.18, No.2, 1992, pp. 169-197; Elson, Diane, "The impact of structural adjustment on women: concepts and issues", in Bade Onimode (editor), The IMF, the World Bank and the African Dept Vol 2: The Social and Political Impact, Zed, London, 1989; Haddad, Lawrence,"Gender and poverty in Ghana: a descriptive analysis of selected outcomes and processes", IDS Bulletin Vol. 22, No.1, 1991, pp. 5-16; Safa, Helen and Peggy Antrobus, "Women and economic crisis in the Caribbean", in Lourdes Benería and Shelly Feldman (editors), Unequal Burden: Economic Crises, Persistent Poverty and Women's Work, Westview, Boulder, 1992, pp. 49-82.
  51. . Satterthwaite, David, "Continuity and change in the world's largest cities", Background paper for the Global Report, 1995
  52. . 1990 was chosen as the reference year in this section and also in Chapter 2 since the most recent census data in most nations is from censuses held between 1989 and 1991 and thus virtually all statistics for rural, urban or city populations for 1995 are simply estimates or projections based on earlier data.
  53. . See United Nations, World Urbanization Prospects: the 1994 Revision, Population Division, New York, 1995.
  54. . As Hardoy and Satterthwaite 1989 (op. cit) note, it is still common for people to confuse 'growth in urban population' and 'growth in a nation's or region's level of urbanization'. Virtually all changes in the level of urbanization (ie in the proportion of population living in urban centres) are a result of population movements in or out of urban centres. Natural increase in population (ie the excess of births over deaths) does not contribute to increases in urbanization levels except where the rate of natural increase in urban centres is higher than that in rural areas. If this is the case, this may be the result of high proportions of migrants from rural to urban areas being of child-bearing age and their movement to urban centres changing urban centres' rate of natural increase. But rates of natural increase are generally lower in urban centres, compared to rural centres. Part of a change in a nation's level of urbanization is often due to rural settlements growing to the point where they are reclassified as urban (and thus are added to the urban population) or boundaries of cities or metropolitan areas being extended and including people that were previously classified as rural - and rapid rates of natural increase can increase this contribution. But in general, a nation's level of urbanization is not influenced much by population increases for it is essentially the result of changes in economic structure; increased proportions of national populations in urban centres reflect an increase in the proportion of employment opportunities (or possibilities for survival) concentrated in urban centres.
  55. . This calculation is made from the Cities Database of the Human Settlements Programme of the International Institute for Environment and Development (IIED) and of IIED-América Latina. This combines the data on city populations from 1950 to 1990 from United Nations 1995, op. cit. with recent and historic data drawn from around 250 censuses and from Chandler, Tertius and Gerald Fox, 3000 Years of Urban Growth, Academic Press, New York and London, 1974.
  56. . Hardoy and Satterthwaite 1989, op. cit.
  57. . Hardoy, Jorge E. and Satterthwaite, David (Editors), Small and Intermediate Urban Centres; Their Role in National and Regional Development in the Third World, Hodder and Stoughton (UK), 1986 and Westview (USA) 1986.
  58. . Harris 1991, op. cit.
  59. . Hardoy and Satterthwaite 1989, op. cit.
  60. . Ibid.
  61. . Cheshire and Hay 1989, page 14
  62. . See Section 2.4 for more details and the basis for this figure and its source
  63. . Hawkins, J.N., "Shanghai: an Exploratory Report on Food for a City", GeoJournal Supplementary issue, 1982.
  64. . Information supplied by Richard Kirkby based on data from the 1990 Census, in Zhongguo renkou tongji nianjian 1992 (Yearbook of Population Statistics, 1992), Beijing, Jingji guanli chubanshe (Economic Management Press), 1992, p.448 and (for area) Beijing Municipal Statistics Bureau, Beijing Statistics in Brief, Beijing, China Statistical Publishing House, 1988, page 1. Apart from the educational quarter in the Haidian District (northwest) and the steel works and heavy industrial area of Shijingshan (west), prior to the 1980s economic boom the city proper could be broadly defined as that area within the san huan lu - the Third Ringroad. This encircles an area of just 158 km2 in a total municipality spanning almost 17,000 km2. Its population comprises all of the four inner city districts and parts of the 4 inner suburban districts. In total, this 'core city' comprises only around half of the 10.82 million official residents of the capital in 1990.
  65. . Garza, Gustavo "Dynamics of Mexican Urbanization", Background paper for the UN Global Report on Human Settlements 1996.
  66. . This ensures the inclusion within Tokyo of the vast suburban areas and includes Tokyo-to (including the islands) and Chiba, Kanagawa and Saitama Prefectures
  67. . Includes Greater Tokyo Metropolitan Area plus Yamanashi, Gunma, Tochigi and Ibaraki Prefectures.
  68. . This is what Toronto's population might be if it was defined with the methodology used in the United States for defining Consolidated Metropolitan Areas. This would include Toronto Metropolitan Area, the adjacent Hamilton CMA (0.6 million), Oshawa CMA (0.24 million) and the rest of York County.
  69. . Note that these figures for the City of London, Inner London and Greater London are census figures; official estimates for 1991 for Inner London were 2,627,400 and for Greater London were 6,889,900.
  70. . A.G. Champion - see Section 2.4
  71. . See note 10 in Table 1.6; Aina, Tade Akin, "Metropolitan Lagos: population growth and spatial expansion; city study", Background paper for the Global Report on Human Settlements, 1995
  72. . The 1991 census figure for Lagos put its population at around 5 million
  73. . Bairoch, Paul, Cities and Economic Development: From the Dawn of History to the Present, Mansell, London, 1988, 574 pages.
  74. . See Chapter 8 of Hardoy and Satterthwaite 1989, op. cit.
  75. . Satterthwaite, David, "Continuity and change in the world's largest cities", Background paper for the Global Report, 1995.
  76. . Parker 1995, op. cit.
  77. . Caride, Horacio E. and José A. Borello, Industrial Location in Major Latin American Metropolitan Areas, report prepared by IIED-América Latina for the Business Council on Sustainable Development, 1995.
  78. . Caride and Borello 1995, op. cit.
  79. . Harris, Nigel, "Bombay in a global economy; structural adjustment and the role of cities", Cities, Vol.12, No.3, June 1995 , pp. 175-184.
  80. . Parker 1995, op. cit.
  81. . Borja, J. and others, Barcelona; un Modelo de Transformación, Urban Management Programme, Quito, 1995.
  82. . Chapter on Baltimore (reference to be added) in Joanne Fox-Przeworski, John Goddard and Mark de Jong (editors), Urban Regeneration in a Changing Economy - an International Perspective, Clarendon Press, Oxford, 1991.
  83. . Parker 1995, op. cit.
  84. . Parker 1995, op. cit., page 7.
  85. . Parker 1995, op. cit.
  86. . Castells, Manuel and Peter Hall, Technopoles of the World: The Making of 21st Century Industrial Complexes, Routledge, London and New York, 1994, 275 pages.
  87. . Friedmann and Wolff 1983, Friedmann 1993, Sassen 1994, op. cit.
  88. . Friedmann 1993, op. cit.
  89. . Sassen 1994, op. cit.
  90. . Sassen 1994, op. cit.
  91. . Friedmann 1993, op. cit.
  92. . Friedmann 1993, op. cit.
  93. . Parsonage, James, "Southeast Asia's 'Growth Triangle': a subregional response to global transformation", International Journal of Urban and Regional Research, Vol.16, No.2, June 1992, pp. 307-317.
  94. . See for instance Chant, Sylvia and Sarah A. Radcliffe, "Migration and development: the importance of gender", Ratcliffe, Sarah A., "Mountains, maidens and migration: gender and mobility in Peru," Pryer, Jane, "Purdah, patriarchy and population movement; perspectives from Bangladesh", and Hugo, Graeme, "Women on the move: changing patterns of population movement of women in Indonesia" in Sylvia Chant (Editor), Gender and Migration in Developing Countries, Belhaven Press, London, 1992.
  95. . See references in previous footnote; also Hugo, Graeme J., "Migration as a survival strategy: the family dimension of migration" in United Nations, Population Distribution and Migration, ST/ESA/SER.R/133, New York, 1995.
  96. . See for instance references in the previous two footnotes; also Baker, Jonathan and Tade Akin Aina (editors), The Migration Experience in Africa, Nordiska Africainstitutet, Uppsala, 1995, 353 pages; United Nations, Population Distribution and Migration, ST/ESA/SER.R/133, New York, 1995; Douglass, Mike, "Thailand: territorial dissolution and alternative regional development for the Central Plains", in Walter B. Stohr and D.R. Fraser Taylor (Editors), Development From Above or Below, John Wiley and Sons, Chichester, UK, 1981 , pp 183-208 and Saint, William S. and William D. Goldsmith, "Cropping systems, structural change and rural-urban migration in Brazil", World Development, Vol. 8, 1980, pp. 259-272.
  97. . Chant and Ratcliffe 1992, op. cit.
  98. . Chant and Radcliffe 1992, op. cit.
  99. . Castles, Stephen and Mark J. Miller, The Age of Migration: International Population Movements in the Modern World, MacMillan, London and Basingstoke, 1993, 306 pages.
  100. . Martin, P.L., "Labour migration in Asia: conference report", International Migration Review, Vol.25, No.1, 1991.
  101. . Champion, Tony and Russell King, "New trends in international migration in Europe", Geographical Viewpoint, Vol. 21, 1993, pp. 45-56.
  102. . Kane, Hal, The Hour of Departure: Forces that create Refugees and Migrants, Worldwatch Paper 125, Worldwatch Institute, Washington DC, 1995, 56 pages. Note that this source also estimated that there were 27 million "internal" refugees who have fled from persecution but have not crossed international boundaries.
  103. . Skeldon, R., "International migration within and from the East and Southeast Asian regions: a review essay", Asian and Pacific Migration Journal, Vol.1, No.1, 1992.
  104. . ILO, Social and Labour Issues concerning Migrant Workers in the Construction Industry, International Labour Organization, Sectoral Office Programme, TMMWCI, International Labour Office, Geneva, 1995, 114 pages.
  105. . Castles and Miller 1993, op. cit.
  106. . Russell, Sharon Stanton, "Migration between developing countries in the African and Latin American regions and its likely future", in United Nations, Population Distribution and Migration, ST/ESA/SER.R/133, New York, 1995.
  107. . Stalker, Peter, The World Of Strangers, International Labour Organization, Geneva, 1994
  108. . Russell 1995, op. cit.
  109. . Castles and Miller 1993, op. cit.
  110. . Castles and Miller 1993, op. cit.
  111. . Castles and Miller 1993, op. cit.
  112. . Russell 1995, op. cit.
  113. . The information presented about each region is largely drawn from Castles and Miller 1993, op. cit.
  114. . Castles and Miller 1993, op. cit.
  115. . Brubraker, W.R, "Citizenship struggles in Soviet successor states", International Migration Review, Vol. 26, 1992, pp. 269-291; and Buttner, T. and C. Prinz, "Structure and impact of German east-west migration", in W. Lutz (editor), Future Demographic Trends in Europe and North America: What Can we Assume Today?, Academic Press, London, 1991, pp. 379-398.
  116. . Castles and Miller 1993 and ILO 1995, op. cit.
  117. . Castles and Miller 1993, op. cit.
  118. . Russell 1995, op. cit.
  119. . Castles and Miller 1993, op. cit.
  120. . Russell 1995, op. cit.
  121. . Champion, A.G., "International migration and demographic change in the Developed World", Urban Studies, Vol.31, Nos.4/5, 1994, pp. 653-677; White, P., "The social geography of immigrants in European cities: the geography of arrival", in R. King (editor), The New Geography of European Migrations, Belhaven, London, 1993, pp. 47-66.
  122. . Fernhout, R., "Europe 1993 and its refugees", Ethnic and Racial Studies, Vol.16, 1993, pp. 492-506; and Black, R., "Refugees and asylum seekers in Western Europe: new challenges", in R. Black and V. Robinson (editors), Geography and Refugees, Belhaven, London, 1993, pp. 87-103.
  123. . Champion 1994, op. cit.
  124. . UNECE, International Migration Bulletin No.3, United Nations Economic Commission for Europe, Geneva, 1993.
  125. . See for instance Harris 1991 and Parker 1995, op. cit.; also World Bank, Urban Policy and Economic Development: an Agenda for the 1990s, The World Bank, Washington DC, 1991, 87 pages.
  126. . George Yeo, reported in Parker 1995, op. cit.
  127. . These statistics are drawn from the same sources listed in Table 1.6
  128. . Satterthwaite 1995, op. cit.
  129. . Satterthwaite 1995, op. cit.
  130. . Becker, Charles M., Jeffrey G. Williamson and Edwin S. Mills, Indian Urbanization and Economic Growth since 1960, The Johns Hopkins University Press, Baltimore and London, 1992, 328 pages.
  131. . Preston, Samuel H., "Urban growth in developing countries: a demographic reappraisal", Population and Development Review, Vol. 5, No. 2, 1979.
  132. . United Nations 1995, op. cit.
  133. . See Preston 1979, op. cit.and Kelley, Allen C. and Jeffrey G. Williamson, What Drives Third World City Growth: A Dynamic General Equilibrium Approach, Princeton University Press, Princeton, 1984, 273 pages.
  134. . Preston 1979, op. cit.
  135. . Bairoch 1988, op. cit.
  136. . Bairoch 1988, op. cit.
  137. . Manzanal, Mabel and Cesar Vapnarsky, "The Comahue Region, Argentina" in Hardoy, Jorge E. and Satterthwaite, David (Editors), Small and Intermediate Urban Centres; Their Role in National and Regional Development in the Third World, Hodder and Stoughton (UK), 1986 and Westview (USA) 1986; Hardoy and Satterthwaite 1989, op. cit.
  138. . Manzanal and Vapnarsky 1986, op. cit.
  139. . Hardoy and Satterthwaite 1986, op. cit.
  140. . Ofer, Gur, "Industrial structure, urbanization and the growth strategy of socialist countries", Quarterly Journal of Economics, Vol. 10, No. 2, 1976, pp. 219-244.
  141. . Becker, Williamson and Mills 1992, op. cit.
  142. . Mitlin, Diana, The Relationship between Economic Change and Level of Urbanisation, Background paper for the Global Report on Human Settlements, 1995.