I INTRODUCTION

In considering the implications of the urban transition, it is necessary to examine not only the absolute size of the urban population, the scale, the proportion of the urban population to the total population, the level, and the growth of the urban proportion, the rate, but also the urban structure of the distribution of the urban population by city size. Countries with similar levels of urbanization can display very different urban structures. In a country where urbanization is characterized by concentration in one large city (the "primate" pattern of urbanization), even moderate overall rates of urban population growth could impose severe pressure on the city that is absorbing most of the increase. Where urbanization has occurred in a more balanced way across a range of large, medium-sized and small urban localities in an articulated spatial hierarchy, the implications may be more conducive for overall development of the country and for reducing regional disparities in social and economic opportunities. Indeed, it has often been argued that the problem in developing countries is not the pace of urbanization but its pattern of concentration in one or a few very large cities.

The focus of this paper is on the implications of the urban structure and the relative rates of growth of large, medium-sized and small cities vis-a-vis the rural areas in Asian countries. The implications of urban structural change are examined within the context of regional spatial transformation in particular countries. But, of course, it is not city size alone that is a matter of concern. From both analytical and policy perspectives, the implications of urban structure and change derive not only from changes in relative size, but also from the source of the urban changes and from factors like the location of cities, the types of linkages with the hinterland and among the urban centres themselves, the regional development context and the stage of economic structural transformation of the country as a whole.

The growth of cities occurs in three ways: through natural increase, migration and reclassification. Each of these components can have different implications for economic and social development. Over time, the growth potential of different cities, which can involve changes in their ranking in the urban hierarchy, reflects their relative dynamism within the urban system. It has also been found that the larger the population of the city, the more self-sustained and stable is its growth, while the population change of smaller cities is directly affected by locational conditions and changes in industrial structure (Yabuta, 1980:701).

The growth of cities is part and parcel of the evolution of a society in its structural transformation from an agrarian to industrial-service society. The importance of the dynamic inter-linkages should therefore be emphasized. "In developing countries, the inter-action between spatial transformation and economic and social development is much more simultaneous rather than the former occurrence being dependent upon the latter" (Richardson, 1981:7). This means that the implications of urbanization and relative city growth must be examined within the overall development process and the historical phase of the economic and social development of particular countries, as well as in terms of the external links of particular countries with the international economic system.

One important perspective is that "urbanization, being at once a vehicle and a consequence of the urbanization process, plays an important role in the development process and this positive contribution should not be ignored when considering its costs" (United Nations 1985:181). "Analyses often incorrectly associate urbanization exclusively with primate city growth and use the adversities of over-expansion in large metropolitan centres as arguments for discouraging urbanization generally. They overlook the crucial role that cities play in economic development" (Rondinelli, 1980:332). In this paper, we consider both the positive and negative effects of urban growth and structure.

In examining the implication of urban structural change, we attempt to distinguish between situations where spatial polarization is reversed "automatically" or where there has been an "urban turn-round" as in the more developed countries, and those where deliberate policy intervention has been used by developing countries to hasten along a "polarization reversal" so as to reverse polarization trends in the national economy towards the primate city and encourage a more rational pattern of urban size distribution.

The policy measures formulated to promote a certain class or group of cities in an urban system are often influenced by conceptual models of the economic, social and political forces affecting urban systems such as the Central Place Theory (Christaller, 1954; Losch, 1954; and more modern statements), the Growth Center Theory (Perroux, 1950; Kuklinski, 1972; Hansen, 1972; and others), the Hierarchical Diffusion Model (Berry, 1973 and other geographers), the Industrial Linkage Model (Pred, 1977), and the Agropolitan Approach (Friedmann and Douglas, 1978; Friedmann, 1981; Lo and Salih,1978, 1981). These models of city systems are closely linked to the development strategy adopted by the country. For instance, the Growth Pole Model, which aims at stimulating regional development through the purposive creation of intermediate cities as regional growth centres, is essentially a top-down approach to development. It is based on the assumption that the urban centre will transmit multiplier effects that will spread or "trickle down" to smaller centres and its hinterland. The Agropolitan Approach, on the other hand, represents a bottom-up approach to development; the agropolitan areas are really rural-urban units or small towns in rural areas intended to promote agricultural development and rural industrialization and service centres or sources of off-farm employment. Complex settlement patterns are assumed to evolve from the bottom up.

II URBAN STRUCTURE AND CHANGE IN ASIA

Bearing in mind the limitations of underlying data (see, for example, Jones, 1982:4; and Debavalya, 1984:153) and the heterogeneity of countries, we review patterns of urban structure and change only in selected Asian countries. Ideally, to examine urban structure, we should look at how cities are dispersed within each country and what the relationship is among these cities. Unfortunately, for cross national comparisons, data are only available for looking at the urban hierarchy in terms of the distribution of the urban population by city size.

Table 1 shows the estimated change in city size classes for various regions of Asia from 1950 to 2000. The most pronounced shift in urban structure concerns the "super cities", those of the size class of 5 million or more inhabitants. As a percentage of the total urban population, these cities accounted for less than 19 per cent in 1950. But by 1980, they made up almost a third of the urban population in East and South Asia and more than a quarter in South-east Asia. By the year 2000, these proportions are expected to rise to around 40 per cent. This is partly due to the graduation of smaller cities into a higher size class by virtue of rapid population growth.

Table 2 shows that not only has there been a strong trend towards a high concentration of the urban population in very large cities but also that the absolute size of the Asian large cities' populations is substantial. Urbanization in Asia has been characterized by the explosive growth of big cities, creating an imbalance rather than a gradation between rural and urban areas. In 1950, only two Asian cities, Tokyo and Shanghai, had a population of 5 million or more. By 1980, cities of this class numbered 14.

Table 2 also shows that the populations of the largest cities in the Asia are expected to at least double in size between 1980 and the year 2000. Many Asian "super cities" are expected to be among the thirty largest agglomerations in the world, each with a population of at least 10 million by the end of this century. These cities are expected to reach the size and scale of what have been labelled "megapolises" (Gottman, 1978:53-60) with the concomitant demand for substantial resources to manage them effectively and to deal with the potential social, economic and physical problems of gigantic magnitudes.

Refer to Table 1

Going back to Table 1, we find that the "intermediate" size cities of between 1 to 5 million population increased their population by almost three times between 1950 to 1980. Over this period, these cities accounted for one out of every three urban dwellers. In contrast, the "smaller" cities of 100,000 to 1 million inhabitants have been having a declining share of the total urban population, even though the absolute size has more than doubled. This pattern is different from that generally observed in Africa, Europe and the USSR where the most rapid average growth was recorded for the smallest cities in the range of 100,000 to 250,000 population (United Nations 1980:40-41).

Another way of looking at urban structure is in terms of urban primacy. Urban primacy concerns the dominance of the largest city, usually the national capital, over the other cities in the national urban system. While dominance involves many spheres of activity, the usual surrogate measure of urban primacy is population size. Various formulae may be used for this, but that most commonly used is the ratio between the population size of the largest city and the combined population of the next three largest cities the "four city index". This index is related to the rank-size rule which states that the population of any city tends to be equal to the population of the largest city divided by particular city's rank in the urban hierarchy; hence the second largest city should have one-half the population of the largest, the third largest, one-third, and so on. For the four city index, the assumption is that the exponent of the result is one, hence the population of the largest city is equal to the population contained in the second, third and fourth ranked cities.

Refer to Table 2

Using this index, the problem of primacy is clear for many Asian countries. The trend towards larger and larger cities that are increasing their share of the total population is a matter of concern to national planners. Most Asian countries, other than the city-states of Singapore and Territory of Hong Kong, have a tendency to increase primacy during the early stages of development. In Thailand, for instance, the four city primacy index was 14.59, in the Philippines, 4.6 and Indonesia, 1.34 (Richardson, 1981:9) Of the ASEAN countries, only Malaysia with a primacy index of 0.69 appears to have experienced the emergence of a significant number of secondary industrial centres (a trend observed also in the more developed Asian countries like Japan and Republic of Korea) with no serious primacy problem (Hackenberg, 1980:394).

Components of Urban Structural Change

Urban structural change occurs because urban counters are not static entities. They grow at different rates as a result of population dynamics (natural increase and migration) and administrative redefinitions of urban boundaries. The growth potential of a city is shaped by a multitude of factors, some of which are: size, structure, function and location of the city; productivity and resource endowments of its hinterland; the sectoral composition, level of technology and demographic characteristics of the state in which it is located; its linkages with other urban centres; and government policy. Changes in the ranking of cities across time reflect the dynamism of towns within the urban system (see Wee, 1985:67).

One important reason for interest in the sources of urban growth is that "Asian urbanization on the whole appears to have little to do with population pressure in rural areas and a lack of labour absorptive capacity due to the stagnation of the rural economy. This means that rural-urban migration has not necessarily led to the easing of the problem of poverty in rural areas. On the other hand, those that form part of the migration stream to urban areas have been mainly absorbed into the low-productivity urban service sectors" (Lo and Salih 1985:15).

The relative importance of the components of the growth of urban centres is of significance to policy makers. In preparing policies designed to affect population growth of a particular urban locality, it makes a great deal of difference whether the major source of urban growth is migration or natural increase. Policies directed towards migration require very different instruments than do these directed at natural increase.

Recent studies (United Nations, 1985:201) appear to support the view that at low levels of urbanization, population is characterized by moderately high rates of natural increase in both rural and urban areas, but that net rural-urban migration tends to exceed natural increase as a source of urban population growth. At moderate levels of urbanization, urban population growth is likely to result primarily from natural increase, since natural increase will be generally very high in the country at this stage and the age structure of the urban population will favour a relatively higher natural increase than in rural areas (a situation that appears to characterize many of the currently developing countries). At advanced levels of urbanization, migration again becomes the larger component of urban population growth, as natural increase becomes very low, as demonstrated by experience of the more developed countries. But it should be stressed that although the relative contribution of the components of urban population growth are related to the level of urbanization, they are mostly a function of the rates of natural increase and migration.

Table 3 shows the components of population growth for some large cities in Asia. Apart from Seoul where the contribution of migration and reclassification was more than two-thirds, most Asian cities definitely experienced higher natural increase than in-migration. Of course, Singapore and Territory of Hong Kong, which are almost entirely composed of one large city, could not be expected to have high in-migration.

Refer to Table 3

Urban Structural Change and Regional Spatial Transformation

The implications of urban structural change can better be understood if the process is seen in the context of regional spatial transformation in Asian countries. Trends in regional development and spatial transformation are both a product of and an influence on urban structural change.

Concern over the adversities of over-concentration in primate cities and unbalanced regional development has led most Asian planners to devote attention to regional development not merely as independent spatial strategies but as part and parcel of economic structural changes in the country. As a consequence, over the past quarter century, three major trends in regional spatial transformation in Asia can be discerned (see Lo and Salih, 1978,1985).

The first trend is that of "popularization reversal". In line with the growth centre strategy which was in vogue as a tool for regional development in the 1970s, attempts were made to decentralize development to the less developed, peripheral regions. The experience of various countries has shown the growth centre strategy to be successful only in countries where the forces making for an actual urban turn-round were already in place such as in Japan and Republic of Korea (Lo 1978:25-51; Kim, 1978:53-77). As in other developed countries, Japan and Republic of Korea have experienced, and are projected to experience, a relatively slow rate of growth of the largest cities, even stagnation or declining population with urban growth shifting to cities located further down the urban hierarchy. In the less developed countries, attempts to divert industries from the primate areas to lagging regions, as in Malaysia, Thailand, Philippines and Indonesia, do not appear to have been very successful (Salih et. al, 1978:79-119; Malaysia, 1984:181-185; Higgins, 1982; Hackenberg, 1980:404). There has, in fact, been unintentional impact; such as the metropolization of major cities, for instance Metro Manila, the Klang Valley conurbation, Bangkok metro area and Jakarta-Bogor-Tangalang-Bekasi. This pattern of regional development is evident in the high growth rates of the major city, in contrast to the more industrialized East Asian countries where, as mentioned earlier, a number of thriving secondary industrial centres have emerged.

The second major trend is that of purposive regional development with the corresponding urban structural change in terms of growth at the lower order centres (Taylor, 1981). Such a strategy of regional development produces quasi agglomeration economies which in turn promote the growth of towns that perform essentially central place functions. The towns are normally based on resource frontier development (such as oil exploration in Trengganu and North Sumatra, Timber in North Kalimantan) or on green revolution areas (such as the FELDA schemes in Malaysia). They have also been noted to cause structural shifts in Asian economies in their patterns of labour absorption (especially for off-farm employment and into the formal sector) and capital formation. (Lo and Salih, 1985:35).

The third major trend is the growth of lower-order centres that have been stimulated by agropolitan development. This bottom-up strategy of regional development has influenced urban structural change in that local level development has been associated with the rise of small district towns or "cities in the fields" (Friedmann and Douglas, 1978). The penetration of rural areas by urban-like forms of production, infrastructure and administration has been described as a process of ''diffuse urbanization" (Hackenberg, 1980). Experiments with the agropolitan approach have included the Republic of Korea's Saemaeul Undong programme, China's commune system of rural organization, India's community development programme, Bangladesh's territorial self-reliant approach to rural development and many of the Malaysian rural development programmes that involve some degree of community level participation and self reliance.

Implications of Urban Structural Change

To examine the implications of these various forms of urban structural change and associated regional spatial transformation, we combine three related approaches. One is to look at the positive and negative implications by focussing on the roles and functions of cities of varying sizes and the links between these cities and the rural areas, in particular development and historical contexts. The second approach is to examine the role of urban structural change in the development process and the relationships between spatial transformation and economic and social transformation of the country. Thirdly, the related policy implications are obviously important both for influencing patterns of urban structural change and for dealing with the socio-economic impact of such changes.

III THE ROLES AND FUNCTIONS OF DIFFERENT SIZE CITIES

The advantages and disadvantages of size vary from one place to another. What constitutes a large, medium- and small-sized city in a particular country should have to be related to the total population as well as the total urban population, apart from the geographical area involved. Also, the concept of optimal city size has little operational value to planners, and city size has little meaning if divorced from structure, function and position in the urban hierarchy. An efficient size for a city depends upon its role and function. The role and function of a particular size city should be visualized in the wider context of development planning and a more balanced pattern of all human settlements, rather than in the narrow context of the urbanization process per se. The important considerations are the economic function of the city, the location of the city, its linkages with the rural hinterland as well as with the metropolitan areas, the transportation network, the existing infrastructure, energy constraints and environmental factors (see Bose, 1984:186).

In the process of economic development, both economic activities and the population of a country tend to be increasingly clustered in urban centres to the advantage of the wide spectrum of economies of scale made possible by urban agglomerations. These economies are external to the firm, but result from the prior presence of other firms and social infrastructure (United Nations, 1980:38). In brief, agglomeration economies provide scale economies for efficient production, convenient markets for goods and services, economies of proximity that lower distribution costs, and a ready supply of both skilled and unskilled labour.

At the upper end of the urban hierarchy, the growth of large metropolitan centres in Asia "has been an economically favourable development providing economies of scale and proximity that have been conducive to industrialization, allowing the cities to absorb large numbers of people in manufacturing jobs, and stimulating governments to construct modern infrastructure, health, educational, commercial and other facilities that require large population concentrations in order to operate efficiently. The large metropolises in Asia now play crucial roles in regional economic development. They are communications and transport hubs for their nations, providing international ports, harbours and air facilities. Most have also become either national or international financial and banking centres, serving as nodal points in networks of national and international trade" (Ron-dinelli, 1985:73). The metropolises may also serve as the prime importer of modern norms (often western in character) and technology from the world's main centres of innovation. From the metropolitan centre, innovation is diffused throughout the hinterland along the transportation and communication networks.

But where the primate city has grown faster than management and physical, financial and social resource capacities, there are visible and severe economic and social adversities. "A higher city growth rate means greater pressure on the quality of the local environment, less time for overcoming social, institutional, and political barriers, and the postponement of the resolution of existing problems created by past growth. A slower growth rate buys time to explore alternatives and to catch up with past need" (Rogers and Williamson, 1984:276).

Within the primate cities, some of which have been described as "bursting at the seams" (Debavalya, 1984:154), the economic problems can be traced to the inability of the industrial base to productively absorb the burgeoning labour force, which is then manifested in unemployment, underemployment, increasing pressures in the low-income, low-productivity urban service sector. A significant and problematic trend accompanying the increasing concentration in the urban structure has been the dramatic shift in the incidence of poverty. The World Bank (1980:3) estimated that while in 1975, about 80 per cent of the households living in absolute poverty were found in rural areas, by 1990 nearly half of the absolute poor are likely to be living in urban areas. In Asian cities, the number of households living in destitution is expected to increase from 12 million in 1975 to nearly 38 million by the end of the 1990s.

A major physical problem confronting the fast growing metropolitan areas in the less developed countries is the spread of squatter settlements. The combined effect of the substantial movement of population into the cities, the resultant housing shortage, the spiralling cost of land and the difficulty of obtaining adequate jobs and income is seen in the unplanned mushrooming of shanty towns occupied by squatters in poverty. The over-crowding in the primate cities has produced serious environmental hazards in the form of air and water pollution and waste removal problems and lack of recreation and park areas.

The poverty and jostle for space and jobs in the cities have been linked to various social phenomena - including a high incidence of crime, juvenile delinquency, prostitution especially among new comers to the city, drug addiction, inefficient bureaucracy and corruption. The traditional functions of the family also tend to be weakened -protection functions for instance are taken over by new institutions such as the police, hospitals, social security provisions, etc. and the recreation function is substituted for by commercial institutions such as the cinema, discos, etc. In terms of the political implications, concentration in the primate city may increase the vulnerability of a metro-dominated political system to alienation from the numerically larger rural electorate.

It has often been observed that the socio-economic problems stem not from the urbanization transition per se but from the urban structure in terms of over-concentration in the primate city and relative lack of attention to the growth of medium and small-sized cities. More equitable and widespread economic development requires the organization of national space in such a way as to reinforce the mutually beneficial interactions between cities and the countryside and between agricultural and industrial development. In this context, medium- and small-sized cities are recognized as follows (Rondinelli, 1985:80-81):

(a) they provide convenient and efficient locations for decentralizing public services, thereby offering greater access for both urban and rural residents to public services and facilities. Such intermediate or secondary cities could function as regional administrative and service centres:
(b) they offer economies of scale which allow a reasonable concentration of basic and intermediate-level health, education, social and municipal services, acting as regional or provincial centres for a variety of public facilities;
(c) secondary cities and towns can offer a wider variety of basic household and consumer goods, commercial and personal services and opportunities for off-farm employment in both the formal and informal sectors;
(d) small towns act as regional marketing centres offering a wide variety of distribution, transfer, storage, brokerage, credit and financial services and important outlets for the sale and distribution of agricultural goods grown in surrounding areas;
(e) the population growth and economic diversification of secondary cities and towns creates new demands for cash crops grown in the rural hinterlands;
(f) many small towns provide conditions that are conducive to small and medium scale manufacturing and to artisan and cottage industries which service local markets with low-cost consumer goods. Medium-sized cities support large industries which provide employment for their own residents and for migrants from other towns;
(g) secondary cities act as agro-processing and agricultural supply centres for fertilizers, seeds, cultivating and harvesting implements, irrigation components and pesticides for farmers in their regions;
(h) they provide off-farm employment and supplementary income for people living in nearby rural areas;
(i) they often serve as regional centres of transportation and communications, linking their residents and those of nearby rural areas to large cities and other regions of the country;
(j) many secondary cities function as centres of social transformation, absorbing rural migrants who might otherwise go directly to the largest city or national capital, accommodating and encouraging the integration of diverse social groups that help assimilate rural people into city life; and providing new economic opportunities for people seeking social and economic mobility.

In the more prosperous countries in Asia, such as the Republic of Korea and Japan, the medium-sized cities tend to be emphasized as the vehicle for bringing about a more balanced population distribution and the amelioration of economic and income inequalities within and between subnational regions. The growth of regional metropolitan centres and intermediate size cities accommodate high-threshold activities and larger industries and serve as "counter-magnets" to the primate city for rural migrants. In the less prosperous nations such as Indonesia and the Philippines, medium-sized cities are important as bridgeheads for promoting the growth of rural industries and especially the processing of farm produce for export. They are important for promoting rural transformation, especially where there are serious problems of surplus labour in agriculture. In China and India where the absolute size of the urban populations is comparable to that of highly urbanized and industrialized countries, despite their low level of urbanization and high proportion of the labour force engaged in agricultural activities, the medium-sized cities assume much greater importance than in the developed countries because of the positive role such cities are expected to play in rural and metropolitan development (Bose, 1984:182-183).

Small market towns and village centres are important in serving as marketing outlets for the agricultural surplus produced in the hinterlands and in providing goods and services to their catchment populations. They can accommodate a wide range of agro-processing, small scale manufacturing and commercial enterprises that could provide off-farm employment for rural workers displaced from farming or for household members not engaged in cultivation. Their significance lies in strengthening the mutually beneficial linkages between urban and rural development in ways that might at the same time promote regional development more effectively.

IV URBAN STRUCTURAL CHANGE AND ECONOMIC AND
SOCIAL DEVELOPMENT

Apart from the problems of absorptive capacity in the primate city, for a country as a whole, the massive concentration of population and economic activities in the primate city generates spatial polarization, economic dualism and dramatic income inequalities that seriously obstruct balanced and widespread development.

In many cases, the concentration in a primate city has been so great as to create marked disparities in levels of wealth and development between the primate city and other regions within the country. The feeling is that a great majority of the people outside these primate cities have been practically left out from the benefits of urbanization and development.

Michael Lipton (1977), the foremost critic of over-urbanization, has argued that despite a rhetoric that emphasizes the role of smaller urban centres and rural development, most national policies in developing countries are biased in favour of the larger urban centres. An urban bias has led to no more than 20 to 30 per cent of a country's capital being devoted to the agricultural sector even though 70 per cent of the population may live in rural areas. According to Lipton, the price of such a misallocation of resources in borne by the whole country in the form of rural stagnation, urban overcrowding and general poverty.

The implications of urban problems should therefore be considered in relation to the larger issues of development, particularly of rural-urban disparities and rural poverty. The situation in most Asian countries is one where the rural populations have become too large to be transferred productively to any conceivable number of present and prospective cities. Without enhanced rural development, rural livelihood is bound to deteriorate. The potential labour force 'explosion' expected by the end of the century in both agricultural and non-agricultural activities demands rural development in order to reduce income disparities between and within rural areas. Without rural development, large sections of the population will simply be dispelled from rural areas and not absorbed into remunerative urban occupations. The result will be a mere transfer of poverty from rural to urban areas (Salih, 1981:11).

In part, the concentration in the primate city was due to deliberate locational decisions and consistent investment decisions to devote national resources for production, physical infrastructure, social services and public facilities to a single city or to a few large metropolitan centres. The concentration of investments in Asian countries was partly the product of a preindustrial and colonial past and partly a consequence of the belief of development planners that primate cities could serve as "beach-heads" of modernization which would act as catalysts for economic growth that would spread outward to revitalize rural peripheries and stimulate the lagging agricultural sector. Development efforts concentrated in the fast growing metropolitan areas were expected to generate "trickle down" effects that would spread to the periphery. This "top-down" approach to development, which was adopted by many developing countries, was based on two basic but wishful assumptions: that industrialization with modern technology can be decentralized to the benefit of rural areas and that national integration through the growth pole strategy can solve the problem of regional under-development (Lo and Salih, 1981:125). But experience has demonstrated that over concentration in one or a few growth centres may not trickle down, producing instead "backwash effects" that tend to drain capital, labour and raw materials from the rural hinterlands to support the further development of the primate city.

These implications are important because in the typical national space economy of a developing country, there are both spread and backwash effects around the primate city. But the problem is that spread decays much faster than backwash, especially at relatively low levels of national development. While the favourable impacts of the growth of the primate city tend to be associated only with the surrounding hinterland usually in the form of decentralization of economic activity, the backwash or polarization effects are typically felt throughout the country, even in the isolated periphery. The failure of development to spread has been largely attributed to the lack of an adequate spatial system in terms of a hierarchical network of urban centres of different sizes performing different functions, widely dispersed but linked to each other and to their rural hinterlands.

It has only been in the countries which were already undergoing accelerated industrialization, such as Japan, or where diseconomies in manufacturing industries had set in, as in Republic of Korea, that the growth centre strategy linked to the top-down approach to development has been successful (Lo and Salih, 1985:34). But the growth pole pattern has not succeeded in promoting regional economic and spatial transformation where the purposive creation of intermediate cities has not been able to generate and transmit economic activities which in turn must be transformed into new economic activities. Hackenberg (1980:404) argued that urban structures and systems of communities do not develop from the top down, they evolve from the bottom up. But for this evolution to take place the prerequisites include substantial investment in infrastructure (highways, irrigation, electrification) by the government; and acceptance of Green Revolution technology by both large and small scale farm operators.

In spite of the limited success that many Asian countries have had to date with the promotion of a more articulated hierarchy of different sized cities, it has generally been agreed upon that appropriate policies for medium and small-sized cities' development could offer feasible, long-range benefits. From a development perspective, the implications are especially significant for balanced agricultural and industrial/ rural and urban development. Often, policies in developing countries have been based on a perception of rural-agricultural development and urban-industrial development as conflicting sources of economic and social change. Pointing out the fallacy and lack of success with approaches that either advocate concentrating on industrialization in large cities as the "engine of growth" or emphasize the importance of rural agricultural development, Rondinelli (1986:240) suggested that developing countries should instead recognize that "rather than urbanization being detrimental to rural development the growth of urban centres could provide economies of scale that increase the efficiency of agricultural support services, essential commercial and financial services and physical infrastructure ... could also accommodate a wide range of agro-processing, small scale manufacturing and commercial enterprises that provide off-farm employment". Increasingly, there is agreement that such urban structural change would be beneficial to both agricultural and industrial development efforts.

It has also been pointed out, however, that for diffuse urbanization and the agropolitan development approach to succeed, it is necessary to have some degree of regional closure such that the intermediate and small towns are able to play their role in reducing leakage from agrarian regions. Such closure will help reduce the overwhelming influence of national and international forces on the regional economies outside the core, which drain their surplus without in return generating any observable top-down spread effects (Friedmann and Douglass, 1978; Lo and Salih, 1981). The benefits of regional closure are twofold. First, it serves to internalize multipliers and external effects through the emphasis on local linkages and complementarities between agriculture and industry; and second, aided by appropriate policies, it helps in the equalization of the ownership of productive assets and the redistribution of income.

A final point that must be discussed is the implication of urban structural change for new patterns of spatial and social mobility. "Development activities which generate off-farm employment usually stimulate migration to places experiencing employment growth. Since most 'rural' enterprise development occurs in small urban centres, induced migration tends to be directed towards small towns in, the short run. However, as workers gain non-farm occupational skills they may move on to metropolitan areas" (Rhoda, 1979:2-3). Jones (1982:25) also pointed out that "developments in transport mean that many people even in poor countries can commute up to 50 miles to work - an option not available in industrializing nineteenth century Europe. Besides, enabling the bedrooms of the urban workers to be located farther from the heart of the city, transport developments facilitate patterns of circular mobility which do not require continuous residence in the city (Hugo, 1981; Streeton, 1981). They also facilitate the locating of factories and other non-agricultural activities in the rural areas, where they can take advantage of the large and increasingly educated workforce. Thus the rural-urban dichotomy, so clear-cut in the medieval fortress towns of Europe, begins to lose its meaning".

Patterns of social mobility are also vastly enriched by changes towards diffuse urbanization. Hackenberg (1980:407) cited the case of Mindanao in the Philippines where, with the development of a system of bustling service centres in rural areas, the options and opportunities available to farmers have been widened beyond just the purchase of land and they can now invest in farm machines, commercial and transportation enterprises and speciality crops like flowers and fruits that can be moved to urban markets with improved transportation. Household members from less affluent farm households can now seek wage employment in industrial agriculture instead of being confined to tenancy or unpaid farm work. Other mobility channels include the government bureaucracy which has expanded into these secondary towns offering employment opportunities, and branches of urban universities which offer college education to rural girls.

V SOME POLICY IMPLICATIONS

The multiplicity of socio-economic implications of urban structure and change discussed about obviously carry a range of policy implications. We cannot offer policy prescriptions of general applicability to all countries, but we can highlight some policy considerations.

Urban structure and change are important for the socio-economic development of a country because they are related to the availability and exploitation of natural resource endowments, the production and distribution of goods and services, a government's capacity to provide public facilities and amenities, the people's participation in national development programmes, and the degree of crowding and stress in the environment under which people live. In this light, policies to affect or respond to urban structural change are also important.

Broadly, policies to influence urban structural change aim to achieve a more balanced pattern of urbanization with a more clearly articulated hierarchy of different size cities and towns, integrated spatial development and accelerated economic growth with equity. "Experience with three decades of development in Asia suggests that a broad spectrum of human settlements - rural villages, market towns, small cities, intermediate regional centres, and large metropolitan areas - is needed to build strong internal economies. Moreover, urban and rural development are not incompatible, and indeed, must be pursued together. Cities of various sizes must be integrated with rural settlements through physical, social, economic and political linkages that forge them into a mutually sustaining network of production, exchange and consumption centres" (Rondinelli 1980:337). Jones (1984:185) further added that "the objectives should be not so much to slow urbanization as to develop more harmonious rural-urban linkage at the regional level, with the aim of an integrated economy where income and employment growth in rural areas and neighbouring towns are mutually supportive and the benefits are not 'creamed' off by a few metropolitan areas".

A burning question has been whether governments should, in fact, intervene to affect urban structural change. The question arises because some theorists (see in particular Mera 1975; 1981) argue that as economic growth accelerates in developing nations, equity problems are ameliorated and spatial polarization is reversed automatically without deliberate policy intervention and that direct intervention tends not to be effective in achieving the objectives to which they are directed; it often proves to be costly in relation to what can be achieved or the results are inferior to alternative measures that are non-spatial. But others such as Richardson (1977:21) point out that even if primacy reversal "is bound to happen eventually, it may not happen for a very long time and the continued polarization in the meantime may conflict with national policy objectives" and that "developing country concern with inter-regional equity, national spatial integration and other spatial objectives will have a strong incentive to 'nudge' primacy reversal along with policy measures. The problems are when to intervene and how to intervene".

It has been suggested that the prospects for decentralization strategies are increased if policies are implemented close to the time when polarization reversal begins rather than when polarization forces are still strong. The efficiency cost of premature intervention may be very high. Primacy reversal has been considered as having begun when the "backwash" effects of resource movements (including migration) into the core region begin to be outweighed by increasing spatial diffusion of technical knowledge, by a rising demand for complementary goods produced in lagging regions; and by the setting up of branch plants made viable by the expanding size of dispersed markets, lower input costs (especially of labour), interregional transportation improvements and mobile external economies (see Richardson, 1981:14).

Experience has demonstrated that attempts to control primate city growth through direct regulation alone, tend not only to be costly but also are not highly successful. Combining regulation with strategic location of public investments, locational incentives to private investors, land use controls, tax penalties and other elements of a decentralized urbanization policy offers greater potential. If controls on the size of the primate city are to be effective, the growth and diversification of intermediate cities must be stimulated to provide viable alternative locations at which high-threshold economic activities and larger industries can operate efficiently and profitably and to act as alternative destinations for rural migrants. But it must be kept in mind that the creation of a few major industrial growth poles in remote rural regions or satellite cities around existing metropolises generally has little effect on the growth of primate cities or on the development of a more balanced system of cities. Furthermore, cities that are selected for development should have strong linkages to their rural hinterlands to ensure that they do not themselves become enclaves that drain the resources of their surrounding rural areas. On the other hand, attempts to slow rural out-migration by dispersing investments widely among rural villages tend to scatter resources so widely as to have little impact on rural and agricultural development. Rather, existing and incipient market towns and small cities should be strengthened as rural service centres. Plans to strengthen these should fruitfully look to self-help, community-based efforts rather than to top-down development efforts.

It is not enough just to have a hierarchy of different sized cities and towns. Different components of the spatial system playing different and crucial economic and social functions in the development process must be linked to each other through a network of physical (road and other transportation and communication networks), economic (production linkages, market interaction patterns, capital and commodity flows, service delivery), technological, social and administrative interactions. Such linkages are essential for generating and spreading economic growth, for helping to integrate regional spatial systems into a strong national space economy, for creating multiplier effects of further growth and change, and for building up the potential for mutually beneficial economic interaction. For example, improved transportation between villages and towns could help to reorganize and expand periodic and regular markets which, in turn, could change the flow of economic and social interactions and the movement of people and goods. Closer linkages among different sized cities make it less expensive and more convenient to integrate technology and to distribute services more widely.
The need for integrating urban structural change and other spatial policies with social-economic development policies has been emphasized throughout the paper. In the words of Renaud (1979:113), "national economic planners must be made more aware that most of their decisions are not spatially neutral, and physical planners must acknowledge the limits placed on their plans by the state of the national economy, if national spatial policies are to improve the national environment".

References

Berry, B.J.L., 1973. Growth Centers in American Urban System Vol.1 Cambridge, Mass.: Ballinger.

Bose, A., 1984. "The Role of Medium Size Cities in the Urbanization Process" in ESCAP, Third Asian And Pacific Population Conference Selected Papers, Bangkok:ESCAP.

Christaller, W. 1954. Central Places in Southern Germany, Englewood Cliffs.

Debavalya, N., 1984. "Working Paper" in ESCAP, Third Asian and Pacific Population Conference Selected Papers , Bangkok: ESCAP-.

ESCAP Population Division, 1984. "Urbanization in the ESCAP Region" in ESCAP, Third Asian and Pacific Population Conference Selected Papers, Bangkok: ESCAP.

Friedmann,J., 1981. "The Active Community: Towards a Political Territorial Framework for Rural Development in Asia", Economic Development and Cultural Change, Vol.29 No.2.

Friedmann,J. and M.Douglas, 1978. "Agropolitan Development: To wards a New Strategy for Regional Planning" in Lo, F.C. and K. Salih (eds.). Growth Pole Strategy and Regional Development Policy : Asian Experience and Alternative Strategies, Oxford: Pergamon Press.

Gottman, J., 1978. "Megapolitan Systems around the World" in Bourne, L.S. and J.W. Simmons (eds.), Systems of Cities New York: Oxford University Press.

Hackenberg, R.A., 1980. "New Patterns of Urbanization in Southeast Asia: An Assessment", Population and Development Review Vol.6 No.3.

Hansen, N., 1972. Growth Centers in Regional Economic Development, New York: The Free Press.

Higgins, B., 1982. "Development Planning" in Fisk.E.K. and Osman-Rani, H. (eds.), The Political Economy of Malaysia, Kuala Lumpur : Oxford University Press.

Hugo,G., 1981. "Road Transport, Population Mobility and Development in Indonesia" in Jones, G.W. and H.V. Richter (eds.), Population Mobility and Development: Southeast Asia and the Pacific, Monograph No.27, Development Studies Centre, Canberra: Australian National University.

Jones, G.W., 1982. "Structural Change and Prospects for Urbanization in Asian Countries", Paper prepared for the Conference on Urbanization and National Development, East-West Population Institute, Honolulu, Hawaii.

Jones,G., 1984. "Structural Change and Prospects for Urbanization South-east and East Asia with Special Reference to Indonesia", Paper prepared for the Conference on Urban Growth and Economic Development in the Pacific Region, The Institute of Economics, Academia Sinica, Taipei.

Kim,A-J., 1978. "Industrialization and Growth-Pole Development in Korea: A Case Study of Ulsan Industrial Complex" in Lo, F.C. and K. Salih (eds.), Growth Pole Strategy and Regional Development Policy: Asian Experience and Alternative Approaches, Oxford: Pergamon Press.

Kuklinski, A.R. (ed.) 1972. Growth Poles and Growth Centers in Regional Planning, Paris and The Hague:Mouton.

Lipton, M. Why Poor People Stay Poor: A Study of Urban Bias in World Development, Cambridge, Mass.: Harvard University Press.

Lo, F.C., 1978. "The Growth Pole Approach to Regional Development: A Case Study of the Mizushima Industrial Complex, Japan", Lo, F.C. and K. Salih (eds.) Growth Pole Strategy and Regional Development Policy : Asian Experience and Alternative Approaches, Oxford: Pergamon Press.

Lo, F.C. and K. Salih, 1978. Growth Pole Strategy and Regional Development Policy : Asian Experiences and Alternative Approaches. Oxford:Pergamon Press.

Lo, F.C. and K. Salih, 1981. "Growth Poles Agropolitan Development and Polarization Reversal : The Debate and Search for Alternatives" in Stohr, W.B. and D.R.F. Taylor (eds.), Development from Above or Below? The Dialectics of Regional Planning in Developing Countries, Chichester: John Wiley and Sons.

Lo, F.C. and K. Salih, 1985. "Structural Change and Spatial Transformation: Review of Urbanization in Asia, 1960-80", paper prepared for the Conference on Population Growth, Urbanization and Urban Policies in the Asia-Pacific Region, East-West Centre, Honolulu, Hawaii. '

Losch, A. 1940. The Economics of Location, New Haven: Yale University Press.

Malaysia, 1984. Mid-Term Review of the Fourth Malaysia Plan 1981-1985, Kuala Lum-pur: Government Press.

Mera, K., 1975. Income Distribution and Regional Development, Tokyo: University of Tokyo Press.

Mera, K., 1981. "Population Distribution Policies: The need for Caution" in United Nations, Population Distribution Policies in Development Planning, New York: United Nations.

Perroux, F. 1950. "Economic Space: Theory and Application", Quarterly Journal of Economics.

Pred, A.R. 1977. City Systems in Advanced Economies, New York: John Wiley and Sons.

Renaud, B., 1979. National Urbanization Policies in Developing Countries, World Bank Staff Working Paper No. 347, Washington D.C. : World Bank.

Richardson, H.W., 1981. "Defining Urban Population Distribution Goals in Development Planning" in United Nations, Population Distribution Policies in Development Planning, New York : United Nations.

Rhoda, R.E., 1979. Development Activities and Rural-Urban migration : Is it Possible to keep them down on the Farm?, Washington D.C.: Agency for International Development. Also in International Migration Review Vol.17 No.1, 1983.

Rogers, A. and J.G. Williamson, 1984. "Migration, Urbanization and Third World Development: An Overview" in Rogers, A. et. al., Migration. Urbanization and Spatial Population Dynamics, Boulder and London : Westview Press.

Rondinelli, D.A., 1980. "Balanced Urbanization, Regional Integration and Development Planning In Asia", Ekistics Vol.47 No.284.

Rondinelli, D.A., 1985. "The Crisis of Urbanization in Asia : Finding Alternative to Megapolitan Growth", The Journal of Social, Political and Economic Studies, Vol. 10 No.1.

Rondinelli, D.A., 1986. "The Urban Transition and Agricultural Development: Implications for International Assistance Policy", Development and Change, Vol. 17.

Salih, K., 1981. Urbanization : Its Challenges to Development Planning, Occasional Paper No.5, Kuala LumpunLembaga Perancang Keluarga Negara.

Salih, K. et. al., 1978. "Decentralization Policy, Growth Pole Approach and Resource Frontier Development : A Synthesis of the Response in Four South-east Asian Countries in Lo, F.C. and K. Salih (eds.), Growth Pole Strategy and Regional Development Policy : Asian Experiences and Alternative Approaches, Oxford:Pergamon Press.

Stohr, W.B., 1981. "Evaluation of some Arguments Against Government Intervention to Influence Territorial Population Distribution" in United Nations, Population Distribution Policies in Development Planning, New York: United Nations.

Streeton, A.W., 1981. "The Building Industry and Urbanization in Third World Countries: A Philippine Case Study", Economic Development and Cultural Change, Vol. 29, No.2.

Taylor, D.R.F., 1981. "Role and Functions of Lower Order Centres in Rural Development" in Lo, F.C. (ed.), Rural-Urban Relations and Regional Development, Singapore: Maruzen Asia.

United Nations, 1980. Patterns of Urban and Rural Population Growth, New York: United Nations.

United Nations, 1985. World Population Trends, Population and Development Interrelations and Population Policies 1983 Monitoring Report Vol. 1, New York:United Nations.

Wee, V.E.L., 1985. "Urbanization in the Context of Socioeconomic Development in Peninsular Malaysia", M.A. thesis submitted to Brown University.

World Bank, 1980. "Poverty", Shelter, Poverty, and Basic Needs Series, Washington D.C. : World Bank.

Yabuta, J., 1980. "Service Employment and Population Growth in Japan", Habitat International, Vol.5 No.5/6.