ASIA AND THE PACIFIC

Introduction

Asia contains three fifths of the world's population and also a large and increasing share of its economic activities and its urban population. It also contains many of the world's fastest-growing large cities, reflecting the fact that it also contains most of the nations with the highest economic growth rates since 1980. But few generalizations are valid for the region, given the number of countries that range from the richest to the poorest and the largest and most populous to among the smallest and least populous in the world. Although this remains a predominantly rural region, it also contains a high proportion of the world's largest cities - see Box 2.4.

______________________________________________________________________________

Box 2.4: The share of Asia's rural, urban and large city population within the world

In 1990, Asia contained:

< 72 percent of the world's rural population. This included the five largest rural populations within nations: China, India, Indonesia, Bangladesh and Pakistan which between them had close to three fifth's of the world's rural population. China and India alone had close to half the world's rural population.

< 44.5 percent of the world's urban population. This includes the two largest urban populations within nations: China and India that between them had more than 500 million urban inhabitants. China's urban population was nearly as large as the total population of Latin America. Both China and India had urban populations that were larger than all of Africa and of all of North America.

< 42 percent of the world's "million-cities" including half of the world's ten largest urban agglomerations (Tokyo, Shanghai, Beijing, Bombay and Calcutta).

NOTES: One reason for what appears to be a sudden increase in Asia's total, urban and rural population and in the number of "million-cities" around 1990 was the breakup of the former Soviet Union, as several republics that were previously in the Soviet Union became part of Asia. They include Kazakhstan, Kyrgyzstan, Tajikistan, Turkmenistan and Uzbekistan (now part of South-central Asia) and Armenia, Azerbaijan and Georgia (now part of Western Asia).

_____________________________________________________________________________

Demographic change

By 1990, the region's population totalled 3,186 million with 32 percent living in urban areas. The region's total population doubled between 1955 and 1990 while the urban population more than tripled. Most countries had populations that more than doubled between 1950 and 1990 while many had populations that more than tripled, especially in Western Asia. Demographic statistics for the region are much influenced by China and India; in this same year, between them they had more than three fifths of the region's population and just over half its urban population. Table 2.18 gives statistics on population and urban change for Asian nations with 10 million or more inhabitants in 1990.

______________________________________________________

 

Population growth rates for the whole region have been declining since the late 1960s, although with an annual rate of change of 1.9 percent a year during the 1980s, this remained second only to Africa among the world's regions. As in Latin America, there are countries with very low and very high rates of change. Some had growth rates below 1.0 percent in the second half of the 1980s - including three of the wealthiest Asian countries - Japan (0.44 percent a year), Hong Kong and Republic of Korea. Some had rates that were still above 3.0 a year - including Yemen, Pakistan and Iran and many of the oil-rich states in Western Asia, although for some of these, this was likely to be associated with immigration. Most countries fell between these two extremes. For the most populous countries, China's annual rate of change for the period 1985-1990 was 1.5 percent with India having 2.0 percent and Indonesia having 1.8 percent. India's rate of population growth was reported to have declined to 1.9 percent a year in 1991 and it is expected to fall further in coming years; - one demographer suggested that it will fall to 1.6 percent a year by the year 2000 and 1.3 percent by the year 2005.

Asia has had the largest increase in life expectancy of any of the world's regions since 1960 - from around 45 years in the late 1950s to 62.5 years in the late 1980s. As Chapter 3 will discuss in more detail, most of the countries with the largest increase in life expectancy between 1960 and 1990 were in Asia.

There is also great variety in the age structure of national populations in Asia. Many countries still have predominantly young populations; projections for the year 2000 suggest that many countries in West Asia and several others (including Nepal, Lao and Cambodia) will have more than 40 percent of their population of under fifteen years of age. By contrast, Hong Kong and Japan are projected to have less than 20 percent with Singapore and Republic of Korea having only just above 20 percent. Japan is the only Asian country likely to have more than 20 percent of its population over 60 years of age by this point. There is a comparable diversity in the age structures of city populations, although unlike national age structures, these reflect the extent to which the city attracts or fails to keep younger age groups. Industrial cities, long in decline, will often have relatively old age structures, unless a considerable proportion of those who leave the labour force retire to other locations.

Data on sex ratios in large Asian cities for 1990 show that most had more men than women - although not in Japanese or Australian cities. In many countries, especially the higher-income and more industrialized and urbanized countries, there has been an increase in the proportion of women in the workforce. The same is true for Dhaka, although Bangladesh is among the lowest income countries in Asia. In Dhaka, in the past, much of the migration was single males but in recent years, there has been more family migration. There has also been a large in-migration of single females, linked to the rapid expansion of the export-oriented ready garment industries there during the 1980s, and these employ over 700,000 women.

Economic and spatial change

Table 2.19 shows the high economic growth rates that many Asian nations achieved during the period 1980-1992 - which is very much in contrast to most nations in Latin America and Africa. Perhaps the most remarkable is that of China with a fifth of the world's population and the second fastest growth in per capita income of any country in the world with 1 or more million inhabitants - the fastest growing being Republic of Korea.

The nations with the most rapid economic growth include what are often called the "Newly Industrialized Countries" of Hong Kong, Singapore, Taiwan (province of China) and the Republic of Korea - although it is many years since they could be called "newly industrialized" and more than half of their GDP comes from services. Japan also sustained a relatively high growth rate per person, despite already being one of the world's most wealthy nations at the beginning of this period. Three of the "ASEAN 4" - Indonesia, Malaysia and Thailand - also had among the world's fastest growths in per capita income in this period. - while the growth in per capita income in this period in nations such as India, Pakistan and Malaysia compares favourably with most of the rest of the world.

Most of the Asian nations with among the fastest growths in their per capita GNP between 1980 and 1992 also had among the world's most rapid increases in per capita income between 1960 and 1980 - especially China, Japan, Singapore, Hong Kong and Republic of Korea. Indonesia, Malaysia and Thailand also sustained high growth rates per person in this earlier period - as did the oil-rich economies of Iraq and Saudi Arabia whose per capita incomes subsequently fell during the 1980s and early 1990s.

_____________________________________________________

Urban change

Although with two thirds of its population still living in rural areas in 1990, it is clear that Asia remains a predominantly rural continent, the extent of this "ruralness" compared to other regions is partly explained by the criteria used by governments in defining their urban populations. For instance, it would only need India and China to change their definitions of "urban centres" to definitions commonly used in many European or Latin American nations for Asia to become 50 to 60 percent urban - as hundreds of millions of what are now classified as "rural" dwellers become urban. This is not to suggest that the definitions used in India and China are at fault but to demonstrate how the proportion of a country's or region's population that lives "in urban areas" can be moved upwards or downwards, depending on the criteria used. Box 2.5 gives more details about the increase in urban population in China and notes how China's urban population might have been 169 million larger in 1990, had not new criteria been introduced for defining urban centres in the 1990 census. The proportion of Japan's population living in urban areas in 1990 can vary from 63.2 percent to 77.4 percent, depending on whether it is the population in city municipalities, densely inhabited districts or standard metropolitan employment areas that is being considered.

_____________________________________________________

In 1990, China had 302 million urban inhabitants representing just over a quarter of its total population. The urban population had increased by almost 130 million between 1978 and 1990; estimates suggest that 35 percent of this was due to net rural to urban migration, 25 percent to natural increase and 40 percent to various alterations in the urban administrative system. These alterations included a large increase in the number of settlements designated as towns (zhen) and municipalities (shi); in both instances, this large increase was associated with changes in the criteria by which a settlement became one of these. However, the scale of China's urban population would have been much larger if the new, more rigorous criteria had not been applied to its definition. China's urban population in 1990 based on the 1982 census criteria would have been 571 million, encompassing more than half the total population. The many different figures given for China's urban population at different points during the 1980s reflect the difficulties of the (then) current urban definitions in accommodating rapid economic change in both rural and urban areas.

Population counts for metropolitan centres underestimate the true numbers having de facto residence, as the economic reforms have created a growing pool of mobile labour. It is generally agreed that the larger the city, the larger this mobile or temporary population. Since 1984, Beijing has consistently reported over 1 million temporary residents or transients; in Shanghai, estimates for late 1988 suggest almost 2 million.

SOURCE: Kirkby, Richard (1994), "Dilemmas of urbanization: review and prospects", in Denis Dwyer (Editor), China: The Next Decades, Longman Scientific and Technical, Harlow, pp. 128-155.

______________________________________________________

The nations in Asia can be classified in three different groups, according to their level of urbanization, with the countries in such groups also having some economic characteristics in common. The first, most urbanized group has Australia and New Zealand, Japan, Hong Kong, Singapore and the Republic of Korea. All are predominantly urban with agriculture have a minor role in their economy; agriculture represented between 0 and 4 percent of GNP in 1990, except for New Zealand where it represented 8 percent. In all these countries, more than half of their GDP in 1990 comes from services. A more recent estimate suggested that services account for close to 80 percent of Hong Kong's GDP and around 60 percent of that of Singapore and Taiwan.

The second group is the "ASEAN 4" (Thailand, Indonesia, Malaysia, Philippines) and Fiji and Pakistan where agriculture contributed less than a third to GDP and where, apart from in Thailand, between 30 and 50 percent of their population is in urban areas. In most of these countries, the contribution of agriculture to GDP declined dramatically between 1960 and 1990 while that of industry and services tended to increase. Thailand appears very unurbanized within this group; it is reported to have only 19 percent of its population in urban areas in 1990 and a total urban population of just 10.4 million but this may be due to the urban definition used in Thailand excluding many settlements that would be classified as urban centres in other countries.

The third group contains China and all Southern Asian countries except Pakistan. These remain predominantly rural with agriculture having greater importance within their GDP and within their employment structure. However, China may soon be in the second group, given the speed of its economic growth, the declining proportion of its GDP in agriculture and its rapid urbanization. But again, China, like India, has such diversity in economic and urban trends within its boundaries that the use of aggregate national statistics is misleading. Certain parts of China and India have the economic and urban characteristics of Group 2 nations. There is also considerable diversity in the economic and urban characteristics of regions in countries such as Indonesia, Thailand and Bangladesh. It is also difficult to generalize about urban change in Asia when the region contains tens of thousands of urban centres. India alone had more than 4,000 urban centres in the 1991 census while there were nearly 500 urban centres in Bangladesh by 1990.

The major cities

Table 1.5in Chapter 1 listed the 30 largest urban centres in the world in 1990; half of them were Asian cities. But the list does not include two of the most dynamic cities - Hong Kong and Singapore. For Singapore, the reason is simply the size of this city-state with just 600 square kilometres. Hong Kong's population is also restricted by its small size but when Hong Kong returns to China in 1997, it would only need a reclassification of its boundaries to reflect the new urban developments closeby for it to become one of the world's largest cities. During the 1970s and 1980s, the transformation of the city-landscape of Singapore and Hong Kong reflected the increasingly important role they came to have within the Asian (and world) economy.

Comparable transformations are also underway in many other Asian cities that are key centres in Asia's economic transformation - for instance Seoul, Kuala Lumpur, Taipei and Bangkok. In China too, there are cities that have undergone a remarkable transformation over the last ten to fifteen years - for instance Guangzhou - and cities that developed almost from scratch - for instance Shenzhen which by 1990 had an official population of 700,000 but whose real population was estimated at more than 2 million. More recently, Shanghai has begun such a transformation, especially in a new development to its east that is being developed as a high technology metropolis - see Box 2.6

______________________________________________________

Box 2.6: Shanghai

In 1949, Shanghai was the pre-eminent city with over half of China's modern industry. But after 1949, throughout the first phase of heavy industrialization, it was deprived of significant investment funds because of its location that was vulnerable to attack from the sea. Its population remained remarkably low, given that it had been the largest industrial centre in the world's most populous nation. In 1982, the city's urban core districts contained only 6.27 million inhabitants with the rest of its 11.8 million inhabitants distributed over what was then over 5000 square kilometres of mainly farming and small towns communities in neighbouring counties. It was not until the early 1980s that it was again given priority - when it was designated one of the 14 "open cities" and with the establishment of the Shanghai Economic Zone in 1983 - but it was still compared unfavourably with Guangzhou. In the early 1990s, a combination of strong local pressure and a concentration of ex-Shanghai leaders in the higher echelons of national politics led to a major programme of redevelopment, especially in its eastern hinterland - a large area of agricultural and marginal land generally referred to as Pudong (literally "east of the Huangpu river"). Here, a new high technology metropolis is being developed, modelled on Singapore.

By 1992, greater Shanghai's population had risen to 12.87 million but the urban core districts now contained 7.86 million inhabitants. There was also a large floating population that was not included in this figure - for instance in 1993, it had a registered temporary population of over 600,000. Migrants into Shanghai generally find employment in the suburban districts and the 6 counties annexed to the city and in almost every zhen and smaller town within Shanghai's orbit, there are established communities of migrants, Most are engaged in new small plants and off-farm enterprises.

Shanghai's official boundaries encompass more than 6000 square kilometres and much of the remaining pockets of untouched farmland are being developed as housing estates, relocated factories and local enterprises, especially in ribbon developments along each highway. The development of rural industrialization in the city's attached counties has also introduced a new sense of urbanness into much of the peri-urban core.

SOURCE: Kirkby, Richard, "Country Report: China", Background Paper for the Global Report on Human Settlements, 1995

_______________________________________________________

Large flows of foreign investment in industry, tourism, land and business development are largely focused on major metropolitan centres such as Tokyo, Seoul, Beijing, Shanghai, Hong Kong, Taipei, Bangkok, Singapore and Jakarta and are leading to intense competition between the administrations of cities and the country of which they are part to capture a greater share of these investments and of tourism. A report by the United Nations Economic and Social Commission for Asia and the Pacific noted that

"Investment in conventional centres, the rehabilitation of central cities and the growth of hotels are major features of these cities. Since these developments, while heavily reliant upon private capital, also need state investment particularly in ancillary infrastructure, there is intense competition for available capital for urban development" (page 2-27).

When comparing the populations of cities in China with those of other countries, one could use the populations of core cities which usually under-estimate the city populations since the exclude many of those living in suburban areas that are part of the city economy or the populations of city-regions which exaggerate their populations. In China, the process of annexation by large cities of surrounding areas tends to exaggerate the size of their populations and their relative importance. One extreme example is Chongqing that with annexed counties had some 12 million inhabitants by the mid 1980s, yet the urban agglomeration itself had less than 3 million inhabitants. Figures given for Shanghai's population have long been much larger than the population in the central urban agglomeration, although the rapid expansion of Shanghai's economy in recent years have lessened this difference. In China, there has been considerable incentives for local authorities to expand their boundaries since it offers them more direct control over rural resources.

What is probably true for much of Asia is that the largest cities had slower rates of population growth during the 1980s than during the 1970s - although this is partly because as a city gets larger, it takes an ever increasing absolute increment in the population to maintain a city's growth rate. Thus, while most of the large cities in India had slower population growth rates during the 1980s, compared to the 1970s, in fact most had larger annual increments in their population in the 1980s, compared to the 1970s.

Centralized deconcentration within urban systems?

There are too few detailed studies of changes in the spatial distribution of urban population within Asian nations to know whether urban populations are concentrating or deconcentrating within their urban system. In the largest and most populous countries, there are likely to be regions where there is increasing concentration of urban population in the major cities and also regions with increasing deconcentration - for instance, it would be surprising if all states in India and all provinces in Indonesia were experiencing comparable trends in this. A detailed analysis of this has been undertaken in Japan (see Box 2.7) which shows a continuous concentration of population in three metropolitan regions since 1945 and thus no "counterurbanization" of the kind experienced in Europe during the 1970s - although as the Box notes, it may be that the trend to urban concentration in these three regions is ending. However, Japan's urban system is inevitably influenced by the role of Tokyo as one of the pre-eminent world cities - and it may be that an end to the urban concentration was delayed - or will continue to be delayed - by the expansion in economic activity within Tokyo.

______________________________________________________

Box 2.7: Urban concentration in Japan

Since 1945, the national population has been continuously concentrating in the three major metropolitan regions centred on Tokyo, Keihanshin (Kyoto-Osaka-Kobe) and Nagoya. By 1990, they contained half of the national population, compared to 38 percent in 1950. As Table 2.20 shows, all three of these metropolitan regions had their most rapid population growth rates in the early 1960s and were growing relatively rapidly until 1975, after which their population growth rate dropped considerably. It is also worth noting how the population growth in the other regions was very small or negative from 1955 to 1970 but more rapid for the 1970s and early 1980s, coming to almost equal that of the three major metropolitan regions in the later 1970s.

 

It may be that the long term trend towards the increasing concentration of Japan's population in its three metropolitan regions - and especially that of Tokyo - has ended. Figure 2.4 (not available)shows the important changes in the number of net inmigrants to the three metropolitan regions and "other". For Tokyo region, there was net inmigration for the whole period, except in early 1993 when the long term trend towards increasing concentration of the national population in Tokyo stopped. For the other two metropolitan regions, net in-migration continued until the early 1970s and then virtually stopped. For "other regions", there was net out-migration until the mid 1970s when it stopped, then an increase in net out-migration, then a second halt in this in the early 1990s.

 

In several of the large Asian countries, many peripheral regions have urban populations that are growing well above the average for all urban centres. For instance in Indonesia, the provinces with the highest urban population growth rates during the 1980s were generally among the least populous and the top seven were all transmigration receiving provinces and all had annual average population growth rates of between 8.7 and 17.7 percent a year. In India, many of the states with the lowest per capita incomes and the lowest levels of urbanization had a more rapid growth in their urban population than wealthier, more urbanized states. In China, the peripheral "border region" had a more rapid increase in its population living in cities of 500,000 or more inhabitants than the much more developed and urbanized coastal and inland regions.

Given the very rapid urban and industrial development in China since the early 1980s and the fact that so much of it is concentrated on the coast, this would be expected to show up in urban statistics. An analysis was made, comparing the growth in large city populations between 1981 and 1990 for the coastal region, the inland region and the border region. However, the regional differences were not so dramatic. The growth in the aggregate population of the 46 cities that had more than 500,000 inhabitants in 1981 between 1981 and 1990 was 26 percent for the coastal region, 24 percent for the inland region and 15 percent for the border region. The difference between the coastal region and the inland region becomes greater - 36 percent compared to 31 percent - if the inhabitants of outer suburban areas are added onto those of the core city populations.

 

Metropolitan areas and extended metropolitan regions

Available data suggests a deconcentration of population within most of the large metropolitan centres with relatively slow population growth rates within the central city (or even population decline) and much higher population growth rates in the outer areas. For instance, the population within the metropolitan region of Jakarta appears to be deconcentrating quite rapidly. Central Jakarta had a relatively modest population growth rate during the 1980s - around 3.1 percent a year - but the growth rates of the urban population in the three neighbouring districts that are part of "Jakarta Greater Metropolitan Area" were 11.7, 20.9 and 19.8 percent. In the main cities within the more successful economies, there is also a major restructuring of central cities - as in office developments, convention centres, hotels and in the diverse range of service enterprises that develop to serve these and in the transport and communication facilities that have to be installed and these will tend to expel population from central cities. As Chapter 7 will describe in more detail, millions of people are evicted from their homes each year to make way for new urban developments, and some of the largest evictions have taken place in the major cities of Asia.

In India, a process of urban and industrial dispersal can be seen within or around the largest metropolitan areas. This can be seen first in the emergence of new towns and rapid growth of small towns within the metropolitan region and the slow population growth within the central city. In the 1991 census, there were 856 new urban centres i.e. settlements that met the criteria for being classified as urban centres - and most of the new urban centres in the more urbanized states were close to metropolitan cities and have strong links with them. Most of the new towns in the State of West Bengal are concentrated around Calcutta - that has long been India's largest city, although its population was surpassed by that of Bombay in the 1991 census. Other new urban centres have emerged around the other metropolitan cities of Bombay, Delhi, Madras, Bangalore, Ahmedabad, Hyderabad, Pune, Nagpur and Visakhapatnam. The expanding physical area of major metropolitan centres such as Calcutta and Bombay has also absorbed what were previously physically separate urban centres.

There is also a tendency for the population in major cities to spread outwards even beyond the boundaries of metropolitan areas - and these are often referred to as Extended Metropolitan Regions. The built-up area and labour market of cities such as Jakarta, Tokyo and Bangkok extend for great distances from their urban cores - and this is often recognized by governments in defining new planning regions - for instance, in Thailand, the government has created an Extended Metropolitan region around Bangkok that stretches for some 100 kilometres from the central core.

It has also been suggested that extended metropolitan regions can be expanded to include corridors that run between them and thus to have interlocking extended metropolitan regions. This would apply to Tianjin and Beijing, Hong Kong and Guangzhou and Jakarta and Bandung and if these were defined as single metropolitan regions, it would bring considerable change to the list of the world's largest cities. There are similarities to the United States Consolidated Metropolitan Regions described in section 2.2 although population densities in both the rural and the urban parts of these regions are typically much higher in Asia than in the United States. However, a certain care is needed in defining such regions, if this is to establish a new base for comparing the population of city regions worldwide. For instance, new extended metropolitan regions could be defined in parts of Europe where major metropolitan areas are as close together as the cities noted above and these would increase the number of the world's largest cities from Europe.

There may be more diversity in this spatial reordering of population within metropolitan areas or extended metropolitan regions than the analyses to data have identified, in terms of the distance from the city centre of the areas with the most rapid population growth and in terms of their location - ie is there rapid population growth in virtually all peripheral areas or is it concentrated in certain peripheral areas or along particular transport axes.

There is also the difficulty of knowing where the metropolitan area or extended metropolitan region ends, both in terms of the urbanized area and in terms of the labour market (ie including settlements which are physically separated from the metropolitan area but have a high proportion of people who commute to work in the metropolitan area). T.G. McGee has highlighted how large areas of the countryside have acquired increasingly urban characteristics such as an increasing proportion of their population working in non-agricultural activities (including increased female participation in such activities) and a diverse mix of agriculture, cottage industries, industrial estates and suburban developments - but they are not part of any major city or metropolitan area. They are also characterised by a fluid and mobile population, including those that commute to larger urban centres, but they are not simply suburban areas. These are often in areas adjacent to the periphery of metropolitan areas or other cities or along main roads that link cities that are reasonably close together. These are also areas where it is common for farmers and agricultural labourers to derive a significant part of their income from non-agricultural activities and for farm households to have household members working in non-agricultural activities.

There are also some interesting examples in Asia of urban agglomerations which cross national boundaries. For instance, Hong Kong can be regarded as the centre of the Hong Kong-Zhujiang Delta region, and also with the common pattern in large metropolitan areas of slow population growth in the central city (Hong Kong) while very rapid population growth in outer areas (Shenzen and Zhuhai). A large proportion of Hong Kong's manufacturing production has been relocated in southern Guangdong and some three million workers in this part of China are employed in factories that were funded, designed and managed by Hong Kong entrepreneurs.

Another example is the urban development around Singapore in Malaysia and Indonesia. In the late 1980s, Singapore sought to use its sophisticated financial, communications and management facilities to exploit relatively inexpensive labour and the rich land and other natural resources in the nearby industrial zones in the Johor state of Malaysia and Indonesia's Riau archipelago - notably Batam Island. There are good possibilities for this "growth triangle" to be developed as a single investment area for industry and tourism.

Another "international" aspect of urbanization has been the large but fluctuating flows of foreign workers into the oil-producing countries in West Asia which obviously affects these nation's urban systems. But it has also had a considerable impact on the countries from where this emigration came as the emigrants channel back remittances to households and enterprises in their own countries. There is also the reverse movement that was of particular importance during the 1990s with a large number of "returnees" from the Gulf states in the 1990s. In 1991 alone, they totalled 2.6 million with three quarters of these returning to countries in West Asia.

What underlies urban change

Urban change in Asia shows how the growth (or decline) of major cities or the rapid growth of small urban centres has to be understood in terms of economic, social or political changes that are specific to that city or wider region. For instance, Karachi's population growth over the last few decades to become one of the world's largest cities has been much increased by the settlement there of immigrants or refugees. These include over 600,000 refugees arriving from India after Partition, large numbers of refugees from Bangladesh during the 1970s and large numbers of refugees from Afghanistan and Iran during the late 1970s and the 1980s. An understanding of Karachi's economy also needs consideration of the role of remittances sent to families based there from Pakistani workers in the Middle East. The very rapid growth of Dhaka in recent decades must also be understood in terms of its much expanded political role, first as part of East Pakistan, then as capital of Bangladesh - and more recently, linked to the rapid growth of garment industries there.

For Asia's largest cities, it is perhaps worth drawing distinctions between those where urban dynamics are strongly linked to the globalization of the world economy and those that are much more linked to political and economic functions of the nation-state. For instance, the size of Delhi and its rapid growth is far more linked to its role as capital of India than to its concentration of enterprises with roles within an increasingly globalized economy. By contrast, urban dynamics in Singapore and Hong Kong are much more their role within the global economic system than as political and administrative centres. Most of the other large cities in Asia come between these two extremes. For instance, Tokyo is the world's largest urban agglomeration because it is both the national capital of the world's second largest economy and one of the three pre-eminent global cities. It is also clear that Tokyo's role within the global economy has helped to counter a tendency towards decentralization out of the central city and during the 1980s, many enterprises opened their new offices in central Tokyo or moved their head offices to Tokyo. In addition, the deregulation and internationalization of Japanese financial market helped to create a concentration of many profitable service activities in Tokyo. For the cities that are strongly linked into the world economy, it is also worth stressing the distinction between those that serve as centres of production for the world market or those that are centres for international tourism - and those that are important "command and control" centres and thus considered "world cities". It is to these latter cities that banks and other financial services, media and the regional headquarters of multinational companies are attracted and they in turn stimulate and support a large range of producer service enterprises while their employees in turn also generate demand for consumer services.

The increasing importance of Asia within the world economy and of its "world cities" was illustrated by a study of the locational behaviour and spatial organization of some major electronics firms between 1975 and 1991. Nine leading Japanese electronics firms expanded production much more rapidly outside Japan than inside, especially in East Asia - but with research and development facilities continuing to concentrate in metropolitan areas such as Tokyo and Osaka.

Urban trends in all Asian nations are also influenced by government actions and structures. Perhaps the most notable change during the 1980s was the relaxing or removal of government controls on urban growth in various countries and the scaling down or removal of special programmes encouraging or directing new investment to peripheral regions. The most dramatic change was in China where there was rapid rural to urban migration during the 1980s and the 1990s. This was made possible by changes in the household registration system that during the 1960s and 1970s had controlled rural to urban migration and by the growing private food and housing markets and employment opportunities that allowed people to find a livelihood and basic necessities outside of the official system; in 1994, the government announced the imminent demise of the household registration system.

The spatial distribution of urban development in many other Asian countries is also likely to change, as their macro-economic policy orientation has changed. For instance, the size and spatial distribution of large cities in India has also been considerably influenced by government priorities to import substitution and heavy industry from the 1950s - and the new priority to export promotion and a lessening of protective barriers will favour different locations to those favoured under the previous policy. In Japan, the slowing in the concentration of population in the three main metropolitan regions during the 1970s was certainly helped by public investment - particularly that in transport facilities that was relatively concentrated in non-metro regions - and public policies that restricted the building of such facilities as large factories in the major metropolitan areas and promoted their moving to the other regions. However, it was also helped by a buoyant economy and by structural changes that encouraged many activities to locate outside the three main metropolitan regions.

Small and intermediate urban centres

A very considerable proportion of Asia's urban population lives outside large cities. For instance, in India, despite the prominence given to its largest cities such as Bombay, Calcutta and Delhi in any discussion of its urban trends - or in general to its 23 "million cities" - in 1991, there were still as many people living in urban centres with less than 100,000 inhabitants as living in million cities - although the proportion living in million cities is growing while that in urban centres with under 100,000 is decreasing.

The economic and urban transformation of many cities in Asia noted above has also been matched by a less visible but perhaps as important a transformation in many smaller urban centres and in particular rural areas. For instance, in China, there has also been a rapid transformation in the rural and "township" economy:.

"The 1980s was a period of unprecedented transfer of China's farm population away from traditional agricultural pursuits. The proliferation of the rural enterprise sector (owned collectively at township and village levels and after 1984 complemented by millions of small privately run operations) has been the cornerstone of an urbanization policy emphasising small and intermediate settlements"

Between 1983 and 1986, the number of rural enterprises increased by a factor of ten; some 90 percent were privately owned. By 1988, the rural enterprise sector employed 95 million people, almost a quarter of the official rural workforce. In many industrial sectors such as building materials and garments, rural industries it now accounts for over half of national output. By 1987, the total value of rural enterprises's output had exceeded that of the agricultural sector and by the end of the 1980s rural enterprises provided up to a quarter of China's export earnings. This growth has helped to absorb the surplus rural labour that lost its means of livelihood with the abolition of collective agriculture. It also had the great advantage of being largely self-financing and of making only modest calls on the higher levels of government for infrastructural investment. The growth of the small-town economies was also a stimulus for agriculture.

But there are also thousands of small and intermediate size urban centres in Asia that do not have dynamic economies. Many became urban centres because they are a minor administrative centre or a market and centre of a local or regional road network.