As part of its long term strategy, the Asian Urban Information Center of Kobe decided in 1990 to undertake a series of comparative studies of port cities. These studies had two major aims. One was the general aim of complementing the Center's biannual Inquiry, the first of which was carried out in 1989-90. The Inquiry was designed to elicit the views of urban administrators, to permit them to speak for themselves, to tell us what they saw as their major problems and what they were doing to address those problems. The Inquiry was conceived as a broad survey, which would sacrifice in depth information to achieve a broad coverage of a large number of cities and their administrators. The original plan was to carry out such an Inquiry every two years. To complement this broad survey, the Center decided to undertake in the off years a set of more detailed studies of a smaller number of cities. The smaller number would permit greater in-depth examination of a series of especially important problems and urban conditions. It was planned that each detailed in-depth study would be guided in part by the results of the Inquiry, and that it would also help to identify the specific questions to be the subject of the next Inquiry.

This chapter provides an interpretive overview of the five country comparisons. lt is organized in nine sections，each of which has a number of subheadings. The first two sections present the introduction and a note on the methods of data conection. Following this, section three provides an overview of urban administration and national　population policy. This provides a general background of each national environment in which the two port cities are located.

Section four begins the analysis of the specific cities included in the study. lt reviews the patterns of population and port growth in each pair of cities. This includes two sections，dealing with population growth and the development of the port. Population growth is further discussed in terms of its three major sources: natural increase，migration and areal expansion. The fifth section examines some of the major causes of growth. These include both location, and site conditions, as well as the political-economic conditions behind the investment strategies that drove the growth process, and themselves derived from location and site conditions. lt also includes the process of economic development in the larger state or national environment. ln section six we examine some of the major conseqences of gowth. These are dealt with under five main headings: urban congestion, urban services，inner city problems,future strategies, and personnel issues. A major source of data in this section is the views of the administrators themselves. Here we attempt to sustain one of the Center's major aims: to give voice to the front line administrators who must daily deal with the problems of urbanization and try to work out practical solutions to those problems. The seventh section deals with solutions that urban administrators have developed to deal with the problems of growth. Again，here we let the urban administrators speak for themselves. The eighth section uses the five countries together to examine the trajectory of population growth. Here we can show how population problems change with the different levels of growth and the different stages of demographic change. We can also use the Japanese experience to suggest how the character of population problems may develop in the future of the other four countries. Finally, section IX provides a conclusion in which we attempt to bring together some of the diverse strands in this analysis.

For the current project two port cities were selected in each of five countries. The countries and their cities are listed in the following table (see also figures 1，2 and 3).

The Center chose to work in five countries primarny on the basis of importance and access. The five countries: Japan, Republic of Korea, China，lndonesia, and lndia, represent the most populous countries of the three subregions of Asia. They are also countries accessible to the Center through past associations. This represents, therefore, a combined importance and convenience sample.　ln each of the five countries, local country directors were asked to select two cities that represented slow and rapid growth.　The country directors agreed in a meeting in Jakarta in December 1990 to use both published objective data and the views of urban administrators to describe patterns, sources, and consequences of rapid growth. Growth was defined primarily in terms of population size, but it will be seen later that a more dominant selection criterion was in fact port growth.

ln all cases, the most important source of data would be the views of the urban administrators. 0bjective and historical data would be used where they were available, but even these data would be evaluated and interpreted by the views of the urban administrators. The following cases studies display a wide range of methods of obtaining and exposing the view of the urban administrators. lndia employed the most deliberately systematic method, asking a predesignated set of urban administrators their views first on general issues of urban problems and then on a detailed set of specific questions, which was taken almost in tact from the 1989/90 Inquiry.

A.Patterns of Urban Administration.  The cities and countries selected for the study show a wide range of formal administrative conditions, which we should keep in mind when we review the patterns, causes and consequences of growth. These conditions are primarily differentiated by the degree of autonomy of the local urban administration. This autonomy has both formal and informal dimensions, and runs in a continuum from high central control to high local autonomy and initiative.　China，Korea and Japan represent something of a central or mixed position，where the larger and more rapidly growing city has a higher degree of legal autonomy and integrity than does the smaller and slower growing city. Kobe, Pusan, and Tianjin all have some form of legal autonomy, and are accorded a special city status that gives them control over their urban and port functions. Their smaller counterparts are all under thecontrol of a higher administrative level, which also controls the port functions. There is an informal, and self reinforcing,dimension to this difference as well. Cities with greater legal autonomy also tend to have greater economic power and therefore greater informal influence in central political circles than do their less autonomous counterparts. Thus for the smaller cities, integrating port and urban planning presents both formal and informal difficulties from which their larger counterparts are substantially freed. As we shall see,  this gives the larger city greater capacities to address their problems themselves. This will raise a series of questions about the utility of local autonomy for dealing with local urban problems.

Indonesia and lndia occupy extremes on this continuum. In Indonesia virtually all city and port functions are controlled by the central government. Local administrators are primarily concerned with implementing the policy decisions made at higher national levels.　Mayors and higher city officials are nominated by the local elected legislative assemblies, but they are appointed by the central government.　This condition of high central control runs parallel with a widely recognized condition of the national political culture. Indonesian administrators tend to follow the source of central power. Their views tend to reflect the views of the center of power,  and they are themselves often reluctant to express views before they know the positions of those in greater power. But there is also a paradox here. Many observers have found that lndonesia shows a high degree of administrative decentralization. That is, the central authority makes the basic policy decisions, but it also permits a high degree of discretion at the local levels in the specific tactics for implementing those decisions. This has helped to promote lndonesian economic development, by permitting knowledgeable local administrators in all areas to adapt national policies to distinctive local conditions.

India is at the opposite pole, where local autonomy, both de jure and de facto, runs high. The lndian constitution leaves substantial legal powers to the states, and to the large cities Though the Center has great financial control and administrative power, it often has more power to obstruct than to initiate local initiative. Local capacity to initiate and sustain developmental activities varies considerably by state or municipal authority. The pontical conditions that determine the extent to which center and local interests are divergent or congruent is also an important fact of life for local administrators. At times Center and Local units work in concert; at other times they are in opposition. When they are in concert,  considerable development can occur. When they are divergent, we often have the conditions of stagnation. We shall see all types of these power relationships working their way out into practice in the cases of Calcutta and Bombay.

Finally,in all countries the personal quality of the local leadership emerges as an important condition. A strong and charismatic mayor can achieve considerable resources and autonomy for his city. A weak and passive mayor implies greater central control.

B.National Population Policies.  The five countries included in this study all have articulated national level population policies that are especiany relevant to the problem we are studying. Policy statements are currently articulated in comparable form in response to United Nations inquiries about policies. Five major population conditions are include in the inquiries. They are: a) population growth;  b) mortality and morbidity; c) fertility, nuptiality and family; d) spatial distribution and urbanization; and e) international migration. The UN Inquiry typically asks if governments consider these conditions of population to be satisfactory or unsatisfactory, and for what reason. For our study the government perceptions and policies are especially relevant for issues of growth, fertility, and spatial distribution or urbanization.　As might be expected,the five countries fall into three major categories, based on the rate of population growth.

China, lndia, and Indonesia all consider rates of population growth and the level of  fertility to be unsatisfactory because they are too high. AIl three have official policies to limit population growth through reducing fertility, and all have energetie national family planning programs to implement these policies. For all three, the population policies are an integral part of the overall strategy for social and economic development. For China there is also concern for the future aging of the population as a consequence of lowered fertility.

The condition of spatial distribution and urbanization of the population is viewed somewhat differently by the three countries. China finds its spatial distribution partially inappropriate, with some regions lagging behind others. Urbanization does not appear to be problematic, however, India finds the condition inappropriate with respect to both regional disparities and rural-urban disparities. Indonesia considers its regional, or inter-island distribution to be inappropriate,but is generany snent on the issue of urbanization.

Korea　has achieved rapid and substantial reductions of fertility, with a total fertility rate now of just 1.7, or slightly below replacement level.　Since this fertility reduction was achieved only recently, however, the population continues to grow at nearly one percent per year, and the government considers this to be too high. The level of fertility, however, is considered satisfactory. The government considers its spatial distribution to be only partially appropriate. lt wishes to reduce both the rate and the level of urbanization.

Japan completed the demographic transition in the 1950s and now shows a slow rate of population growth and low fertility. The government considers both the rate of growth and the level of fertility to be satisfactory. lt is,  however, very much concerned about the aging of the population. The spatial distribution of the population is considered partially appropriate. Government is attempting to slow the rate of urban inmigration and rural outmigration.

A.Population Growth.　One of the striking and unexpected findings of the study is that even through the cities were selected to represent rapid and slow growth, they are not always distinguished by their growth rates. This is in part because growth rates differ over time, and in part because the cities have shown substantial variation in the extent to which their administrative boundaries have been changed over time. Table 1 summarizes the rates of growth overthe past century, roughly dividing the centuly in two equal parts. Figure 1 provides a graphic portrayal of the patterns of urban growth for the pairs of cities.

Only in Korea and lndonesia (until 1971) is there a clear and marked contrast. Pusan has grown far more rapidly than Mokpo in the past three decades. ln lndonesia, Surabaya far more rapidly than Padang until about after 1961, when Padang appears to grow much more rapidly.　This situation illustrates a common problem of unstable administrative boundaries, however, which will be discussed in grater detail in the body of this chapter. Padang's recent
rapid growth comes primarily from a major extension of the city's boundaries after 1971. From 1900 to 1971,  Padang grew at a rate about one percentage point less than Surabaya.

For China we have at this time long term growth figures only for Tianjin. Here we see a slight increase in the rate of growth from 1.3 percent per year in the first half of the century to 1.8 percent in the last half of the century. For the past four decades, however, Lianyungang has grown more rapidly than Tianjin.

Calcutta and Bombay in lndia grew are roughly similar average annual rates from 1901 to 1981, with Bombay only slightly ahead of Calcutta (1.8 versus 1.6 percent average annual growth rates). Their relative positions changed over time, however, with Calcutta growing slightly more rapidly in the first half of the period,  and Bombay growing more rapidly in the second half.

In Japan the overall growth rates of Kobe and Niigata for the past century are almost identical, but they have differed in different periods. Kobe grew slightly more rapidly than Niigata in the first half of this century (4.0 versus 3.3 percent), but Niigata has grown substantially more rapidly since 1940 (2.4 versus 0.7 percent). As in lndonesia, these differences represent different periods of boundary expansion.

Two caveats are in order. The first is that urban growth rates have fluctuated considerably over time, and the relative position of any city depends in part on the period chosen for comparison. These are shown in detan in each of the country papers.　Second, as we shall see shortly,these growth rates are strongly affected by the areal expansion of the city, and thus may not reflect real demographic changes. This uncertainty about growth rates represents a major problem for both urban administrators and scholars, which calls for a discussion later on.

All cities grew by three major movements; natural increase, net in migration and areal expansion. Date are not always available on the relative contribution of each to the overall growth, but some general comments can be made first on the population movements. Areal expansion is usually more fully detailed, thus that discussion will be developed separately below. In all cases the cities have followed the national trend of natural increase. All of these countries have experienced the rapid population growth of the middle stages of the demographic transition. Japan, Korea and China have also experienced the rapid declines in fertility that signal the end of the demographic transition. In China natural increase contributed substantially to the reported growth of both cities until the 1970s, when the birth rate began to fall dramatically. Since population movement was rather strictly controned, migration did not contribute much to reported population growth through the 1970s.　The fall of fertility has been more dramatic in Tianjin than in Lianyungang. For both cities, officially recorded migration has been highly erratic, with net inflows occurring in one five year period and a net outflow in the next. ln no case, however, have these been large migrations. These are, of course, only officially recorded movements. lt is well known that the loosening of controls over the past decade has brought a large but uncounted increase in urban migrants.

In India, detailed migration and natural increase data are available only for Bombay. There net immigration was about four times the level of natural increase through the 1940s and contributed substantially to the city's growth. Since the 1940s, however, migration and natural increase have contributed about equally to the city's growth. Data are not available for Indonesia. Both countries have developed national family planning programs, and in Indonesia and some of the states of India these have been quite effective in reducing fertility. That reduction has reduced the contribution of natural increase to the growth of both cities,  but the actual levels are not known.

Japan and Korea are at an advanced stage of population development, with low fertility and natural increase. In both cases in-migration contributed substantially to city growth in earlier stages, but has fallen off substantially recently.　For Japan the turning point was around 1950, for Korea it is more recent, around 1985, at which time net migration became negative for both cities. For all but Mokpo in Korea, natural increase has been at or below the national average and now contributes very little to city growth.　Mokpo represents an interesting case of higher than national levels of fertility. This has been the case for some time, though the gap has been steadily declining. lt is not known why Mokpo shows this higher than average level of fertility.

All of the cities except Bombay and Calcutta have experienced an expansion in their urban administrative areas. In most cases this has been a gradual accretion as surrounding rural and semi-urban areas have been added to the administrative boundary of the city and have become more urban. ln some cases, however, as in Tianjin, there has been a radical reshaping of boundaries, including first a rapid growth and then a reorganization to reduce the areal size of the city.　Data on changes for Indonesia are not presently available, except it is known that Padang experienced a large areal expansion after 1971. It is generally the case,  however, that in Indonesia cities are underbounded in Java, where Surabaya is located. There the actual urban area extends beyond the administrative boundaries.　In the outer islands, where Padang is situated, cities are typically overbounded, with administrative boundaries including vast tracts of rural land.

In both Japan and Korea, the city areas grew substantially over time. Pusan more than doubled between 1960 and 1989 (219 to 526 square knometers) and the smaller Mokpo grew by about four times (10 to 46 square kilometers). Kobe in Japan grew by a factor of 25, from 21 to 545 square kilometers and Niigata grew by a factor of 17, from 12 to 209 square kilometers. One of the problems Niigata faces, as we shall see later is the political obstacle to increasing its area.

This represents an important source of uncertainty in assessing urban growth. In effect, we do not really know how much or how rapidly the cities have grown, because the boundaries of the city change so that we are not comparing the same city at different points in time. This identifies a major need for both more data and different data conection processes in urban studies. To be comparable, we should have the population numbers of a stable area of the central city. Two useful approaches to this problem are developed in urban demography. 0ne is to use the Standardized Metropolitan Statistical Area (SMSA), which permits us to compare different cities at any point in time and hold roughly constant their areal boundaries. The second concept is developed in Japan, called the Densely lnhabited Districts (DID). In the next chapter, Professor Toshio Kuroda reviews Japan's development of the concept of the densely inhabited districts. He shows that the DID is coming to represent a more exact, comparable, and stable definition of the city．

B. Port Growth.  If population growth rates did not always differ greatly, and sometimes not in the expected direction, port growth rates did indeed differ as predicted. The ports of the large cities grew more rapidly than those of the smaller cities. The one exception to this generalization is Lianyungang, whose port activities took a sudden recent upturn, growing more rapidly than the port activities of Tianjin in the past few decades. The differences in growth rates and levels of port activity are graphically represented in Figure 2.

A. Location. One of the main causes of the different rates of port city growth is the location of the city on world trade routes. This is especially clear in the case of Japan, Korea and Indonesia. Before the opening of Japan in the 19th century, Kobe did not exist, and Niigata was a port of some consequence.　When Japan was opened,  the trade routes shifted to accommodate world trade, eclipsing Niigata on the Japan Sea, and turning Kobe into Japan's leading port.

Similarly, in the 19th and early 20th centuries Padang in Indonesia was a major port of call for steamships entering Southeast Asia. There was a technological force operating here as well, since Padang was located near West Sumatra coal deposits.　The shift from coal to oil changed world shipping lanes from the lndian Ocean to the Straits of Malacca and Padang declined as a port. ln Korea, Mokpo had for centuries been a major port for trade with China. The Japanese colonial government further developed the port for the export of Korean rice to Japan. With the liberation of Korea in 1945, rice exports declined and the port of Mokpo declined a pace. Pusan then became a major port for Korea's new involvement in international trade.

The impact of changing trade routes on Chinese and Indian port development is more local than world wide.　Calcutta was a major port when Britain controlled India, including East Bengal (now Bangladesh) and Burma. The demise of British control after World War II removed some of the local stimulus to Calcutta's growth, and Bombay became the country's major port. The two Chinese ports were more affected by national rather than world wide conditions. Tianjin has been the port of the capital at Beijing for years, and Lianyungang has been eclipsed by major ports just to its north and south. In the past decade, however, Lianyungang has been designated the terminus of the rail "landbridge" from Europe to China's Pacific coast, producing a sudden upsurge in port activities.

B. The National Environment of Economic Development.  Port city growth in these five pairs is in part caused by the rate of economic development in the immediate environnlent of the port. This was suggested first by the case of India. West Bengal, the state in which Calcutta is located is one of the country's most depressed states. On the other hand, Maharashtra, Bombay's state, is one of the fastest growing in economic terms. Once the Indian case explicitly states this proposition, it is evident that simnar forces work in other countries as well. East Java has been growing economicany more rapidly than the more isolated West Sumatra. The southeast of Korea, with Pusan, has been growing more rapidly than the southwest, with Mokpo. Japan's Pacific coast, with Kobe, has been growing more rapidly than the Japan sea coast on which Niigata is located. The Beijing area, with its rich and densely populated hinterland and its port at Tianjin, has been growing more rapidly that the coast around Lianyungang.

This observation raises another question, however, which is made explicitly in the Korean study. Who makes the investment or development decisions that lead to the difference in growth rates of different regions of a country? In all cases the central government has played a major role through its economic development planning and its decisions on the allocation of public investment.　In some cases these investment decisions have been driven primarily by economic considerations. The Korean central government could choose to develop Pusan for its rapid national industrialization program because Pusan already had a substantial infrastructure development. This did not, of course, prevent the people of Mokpo from claiming that their stagnation was a result more of political than economic considerations of the central government.

APPENDIX I
Major Development Projects