Chapter 2:Urbanization: Concepts, Observations and Policies from the Japanes Experience

I. Introduction

The overview paper has pointed to a number of similar conditions in the five pairs of cities studied in this volume. One is the rapid increase of the urban areas. The other is the question of the definition of the city and the relative weight of the three sources of growth: natural increase, net migration and areal expansion or administrative reclassification. Since Japan has gone through the process of urbanization that the other countries of Asia are now experiencing, Japan's experience might provide some useful insights for the future of urbanization in the rest of Asia.

First, regarding urban growth, we found that all of the cities studied here are experiencing rapid growth. Even the smaller of the two cities in each case is growing more rapidly than the total population. That is, each of these countries is becoming more urbanized. Indeed, this is true of most of the countries of the developing world. Will this process end with near complete urbanization, and what is likely to be the composition of larger and smaller cities? Japan's experience on these issues provides at least one model for the future.

Rapid economic growth in post World War II Japan has brought a rapid rise in living standards. At the same time, however, the industrialization, urbanization and heavy migratory movements that accompanied rapid economic development have caused social, economic and environmental problems. In order to overcome these serious problems, diversified public policies have been implemented by the government. Policies to redirect urban development have been important elements of these policies.

Second is the problem of how we define a city. This question emerges when we note that cities grow through three different processes. Natural increase, and net migration are the two dynamic population conditions that lead to urban growth. But there is also an administrative reclassification, which produces a simple areal expansion of the administrative boundaries of the city. This makes it difficult to compare an individual city over time, or to compare different cities at one point in time. As the overview paper has noted, demographers have struggled with this question for some time. Japan developed the concept of the Densely Inhabited District (DID) to deal with this issue. This, too, may provide a useful example for other cities of Asia.

To Top

II. Urbanization in Japan

A. Concepts

The definition of the urban population in Japan is usually based on the shi. This is the population living in the area legally designated as a shi. It usually has a population greater than 30,000, and is generally recognized as a city or urban area. In order to help solve the financial problems of many small towns, laws were passed In 1953 and 1956 to promote their annexation into larger cities. As a consequence of these laws, a major source of urban population increase in the 1950s was simple reclassification or urban boundaries. This often resulted in urban places that contained a substantial rural population. In 1950-55, for example, 85.9 % of urban population increase in Japan was due to the extension of city borders and only 14.1 % was contributed by natural increase and net-in-migration.

To obtain a clearer picture of the urban population, the Statistics Bureau adopted a new concept, Densely Inhabited Area (DID). Since 1960 the Bureau published figures for the DID as well as for the urban population. Table 1 shows the proportion the urban, or ski population, and also that proportion that resided in the DIDs. Between 1960 and 1985 the difference between the proportions living in the Shi and the DIDs constantly shrank. This suggests that the DID has come to represent both a strict and also a comprehensive definition of the urban population. That is, the DID has come to provide a better definition of the urban population, although it is only available from 1960.

Table 1 Urbanization Trend in Japan Since 1960 Showing Ski and DID Population as Percent of the Total Population

Year	Shi Population (%) (A)	DID (%) (B)	(B)/(A)
1960 1965 1970 1975 1980 1985 1990	63.3 67.9 73.1 75.9 76.2 76.7 77.4	43.7 48.1 53.5 57.0 59.7 60.6 63.2	65.0 70.8 74.2 76.1 78.3 79.0 81.7

Source: 1990 National Population Census; Results of Basic Tabulation (National) November 1991

To Top

B. Sources of Urban Population Increase

Urban population increase is basically determined by three factors: reclassification, or the extension of city boundaries and graduation of rural places into urban places; natural increase, or the surplus of births over deaths, and net migration, or the excess of in-migrants over out-migrants. After 1950-55, reclassification declined dramatically from 85.9 percent of urban increase to 36.5 percent in 1960-65, and to 48.5 percent in 1965-70. As a result, both natural increase and net migration became the dominant forces in urban increase.

The reclassification is technical and done for administrative reasons. From the city administrator's point of view, natural increase and migration are the most important sources of growth, because they present distinctive problems for the administrator. For example, growth by natural increase will imply an increased demand for health and educational services. Net in migration, on the other hand, implies an increased demand for employment and housing.

We can see how these two sources of growth changed dramatically for Japan's three largest metropolitan areas, where there has been no reclassification. Until 1965 net in migration provided more than half of the population increase of these great cities. By 1980, migration ceased to be important, and natural increase accounted for all of the urban growth. This is shown in table 2.

To Top

Table 2 Proportional Distribution of Natural Increase and Migration in Population Increase of Three Major Metropolitan Areas*

Year	Natural Increase	Net Migration	Statistical Errors
1955-60 1960-65 1965-70 1970-75 1975-80	43.9 46.7 62.9 79.6 106.4	55.1 53.1 38.1 14.9 -1.5	0.8 0.0 -1.1 5.4 -4.9

* The three metropolitan areas are Tokyo, Hanshin, and Chukyo.
Source: Urbanization and Development in Japan, The Asian Population and Development Association, March, 1986, pp. 81-90.

The role of natural increase and migration in the urban population growth is quite different from country to country and also by the stage of development. In general, natural increase is a much more important component of urban growth in the Less Developed Countries than in the More Developed Countries. This fact has important implications for population policy. The conclusion is obvious. To control urban growth in the Less Developed Countries, it becomes much more important to control natural increase than it is in the More Developed Countries. This suggests that although migration control policy may be important for the LDCs, fertility limitation for both national and urban populations is far more important. It should be also noted, however, that the contribution of the factors to urban population increase tends to be quite different city by city. This is clear in the ten cities included in the studies in the volume.

To Top

C. Structural Change of the Urban Population

The structure of the urban population, or the distribution of cities by size, also provides important insights on the process and character of urbanization. Different rates of growth for the largest cities or for middle-sized cities can show important differences in popular attitudes toward urban living. Changes in size distributions thus also show changes in attitudes toward the city. These often reflect changing economic opportunities, but they also reflect changes in the quality of the environment as well. Finally, these changes can reflect the impact of different urban policies, The recent experience of Japan in this regard may provide a some important insights for other countries less far along in the urbanization process. The basic data for this analysis can be seen in Table 3, which shows the proportion of the urban population living in Large (over 500,000), Medium-sized (100,000-500,000), cities from 1960 through 1990.

Table 3 Distributional Change of Urban Population by Size of City Population

City group by Pop. Size*	1960	1970	1980	1990
Large	31.0%	33.7%	32.5%	33.1%
Medium	32.7	37.7	41.5	42.5
Small	36.3	28.6	25.9	24.4
Total	100.0	100.0	100.0	100.0

* Large= Pop. more than 500,000; Medium= Pop.100,000<n<500,000; Small= Pop. less than 100,000. It should be noted population size categories are arbitrably selected.
Source: Population Census data

To Top

As we can see in Table 3, in 1960 the urban population in Japan was about equally divided among the three size classes. Over the past three decades, the largest cities have amintained a stable one-third of the total urban population. The big change has come in the smaller cities, which declined from 36 to 24 percent ofthe urban population. Their loss was taken up almost exclusively by the medium sized cities, whose proportion grew from 32 to almost 43 percent of the total.

Table 4 shows the same trend using absolute numbers. In the past 30 years, 1960-1990, the population in large cities increased by 13,186,000 or by 71.3 percent. That in small cities increased by 1,639,000, or only 7.6 percent. The medium cities, however, more than doubled, rising by 21,160,000. This remarkable increase in the concentration of population in medium cities indicates stronger orientation of both individuals and private sector businesses to selective location of residence and industries.

Table 4 Population by City-Size Group, 1960-1990 (in thousand)

City Size	1960	1990	Increase (Rate)
Large Medium Small	18,492 19,513 21,653	31,678 40,673 23,292	13,186 (71.3%) 21,160 (108.4%) 1,639(7.6%)

Source: Population Census data

To Top

III. The Changing Pattern of Migration Streams

The typical migration streams in Japan have been the population flows between the three large metropolitan areas -- Tokyo, Hansin and Chukyo -- and the rest of the country. The population of the three large metropolitan areas combined made up almost half (47.8 %) of Japan's total population in 1990. Since these cities together accounted for only 13.2 percent of the total national land area, this reflects a highly unbalanced population distribution. Further, these three large metropolitan areas are traditionally characterized by high concentration of industrial and commercial activities, reflecting a highly unbalanced economic distribution. This imbalance has been a major concern of the government in the post war period.

Thus the population exchange between these Large Metropolitan Areas (LMAs) and the rest of the country is a matter of great importance for our analysis. Table 5 shows historical changes of inmigrants and out-migrants and net migration in a 37 year period, 1954-1990. Four major observations can be made.

First, note that the inflow into the LMAs from the rest of the country exceeded one million per year in the period, 1961-1973. It then declined sharply up to 1980, stabilizing at about 800,000. However, most recently a modest declining trend is seen.

To Top

Second, the out-flow of migrants from the LMAs to the rest also increased, but less rapidly than the in-flow until 1961j62. Thus in this seven or eight year period the net inflow increased steadily, from just under 400,000 to well over 600,000 per year. After 1962 the out-flow grew more rapidly than the inflow, producing a steady decline in net in-migration, which even became negative for four of the five years 1976-81. This reverse of the flow from net in-migration to net out-migration in the LMAs should be extremely noteworthy, because it had never happened before in Japan. At the time it was assumed to be the starting point of redistributive movement of population from excessive concentration in the LMAs to local areas from which the migrants came. This reverse migration was called by the author a "U-turn" movement at that time.

Third, the LMAs have regained their strong pulling power since 1981, reversing the stream once again. The net in-flow grew to about 100,000 in just four years, and may have stabilized at that level, which is notably far below the net in-flow levels of earlier years. It is not yet clear whether the downturn of the net in-flow in 1989 represents another reversal or the beginning of a cyclical movement from small net in-flows to small net out-flows and back again.

Fourth, there are interesting differences among the three LMAs. The Tokyo metropolitan area, the multi-dimensional capital of Japan, has continued to exhibit a strong attractive power, pulling population from allover the nation. The net in-flow grew to over 350,000 per year in 1961-3, then declined steadily to only one-tenth of that in 1981. The migration stream never turned negative, however, and for the past decade appears to be fluctuating cyclically between about 60,000 and 100,000.

Hanshin and Chukyo have seen the same general time path of rising then falling net in-flows, but they two differed from one another quite a bit. Hanshin's migrant stream turned negative in 1973, and since that time it has had a net in-flow for only two of the 18 years. Chukyo's stream turned negative in 1975, but in the ensuing 16 years, it has had a positive migration stream for 11 years. For the past six years the stream has roughly stabilized at just above 10,000 in-migrants per year.

To Top

A declining or stationary trend of population in the capital region and large metropolitan area in the 1960s and 1970s was a common experience in the More Developed Countries. This also implied that in some cases rural populations reversed a long term trend from losing to gaining population. Japan's case is thus more or less similar to the Western trend. However, whether so called counter-urbanization in terms of declining population concentration in the large metropolitan areas and the consequent rural revival is emerging as a major trend is not clear at this moment.

It may well be that the nature and pattern of population distribution is changing. We have seen, for example, that medium sized cities according to our definition have been increasing rapidly and have gained a dominant position in terms of population size since 1970. On the other hand, the population of large sized cities continues their steady increase, but their proportion of the national population has been roughly stationary.

Thus while the pace of urban population growth and concentration is accelerating rapidly across the Third World, it, appears to have come to the end in advanced countries, including Japan. It is not unlikely that we shall see some of the same processes as the Third World becomes more fully urbanized.

In connection with these changing patterns of urbanization and population distribution, a number of questions can be raised about their nature and significance, and about their implications for policy.

To Top

IV. Population Deconcentration Trends and Policies

Even in the pre-war years, there was a great concentration of population in the large cities and metropolitan areas. As early as 1935, more than a third (38 percent) of Japan's population lived in these areas. The concentration of people in the large metropolitan areas has greatly accelerated in the post-war years, however. As we noted above, their share reached almost 50 percent in recent years.

Although this increasing concentration of population was associated with rapid economic growth and rising living standards, it also generated considerable concerns in government because of some of its negative aspects. The population concentration produced a polarization into two distinct areas: one was heavily over populated while the other was experiencing heavy depopulation. Second, the concentration produced an increasing income gap between the urban areas and the rest of the country. Third urban concentration was having some undesirable consequences for private enterprises as economies of scale were lost to the costs of congestion. Finally, the excessive concentrations began to have a negative impact on the environment and on people's quality of life.

As a result, both public and private sectors and also central and local governments became keenly aware of the serious effects on people's welfare of an extremely imbalanced population distribution.

In 1962, the first National Overall Development Project was adopted. Among other things it aimed to prevent excessive concentration of population and industries in a few great cities and to develop a more balanced regional development. This was followed by the second (1969), third (1977) and fourth (1987) National Overall Development Projects of the central government. In all of these overall plans, high priority was given to deconcentration projects, which aimed to achieve a more balanced distribution of the population.

To Top

Table 5 Migration to and from Three large metropolitan Areas and Net Migrants (Toatal and Each LMA) (in thousand)

Year	In-Mig. to LMAs	Out-mig. (from LMAs)	Net*	Net Migrants
				Tokyo	Hanshin	Chukyo
1954 1955 1956 1957 1958 1959 1960 1961 1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972 1973 1974 1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990	788 718 743 866 815 880 999 1,104 1,184 1,209 1,217 1,186 1,138 1,154 1,202 1,252 1,263 1,214 1,127 1,099 987 912 850 858 837 812 789 799 806 794 782 782 808 803 772 774 763	403 385 342 359 393 389 406 449 536 589 639 705 732 750 784 827 870 926 921 985 949 901 873 867 829 827 795 768 751 711 692 679 658 656 668 682 699	385 353 401 507 421 490 594 655 647 619 578 481 406 404 418 426 393 289 207 114 39 11 -23 -9 7 -15 -6 31 55 83 90 102 150 147 104 92 64	258 235 247 295 273 301 333 359 364 354 327 298 266 255 259 250 248 206 159 97 53 45 26 35 45 30 31 53 65 83 86 95 125 129 98 89 67	103 95 113 168 123 145 189 221 211 185 174 131 103 107 112 121 91 46 24 -5 -21 -30 -41 -45 -38 -41 -36 -24 -14 -5 -2 -4 8 3 -6 -11 -18	24 23 42 44 26 45 72 75 72 81 76 52 37 42 47 55 54 37 24 22 7 -4 -7 -0.3 0.4 -3 -2 2 4 4 6 12 17 15 11 14 14

Source: "Annual Report on Migration Derived from the Basic Resident Registers," Statistics Bureau, Management and Coordination Agency.
Remarks: minus (-) means excess of out-migrants over in-migrants.
* Totals may not agree with the three "net migrants" columns on the right due to rounding.

To Top

In 1962 the central government also enacted a very ambitious law called "New Industrial Cities Promotion Law". The law aimed at establishing several new industrial cities that were expected to be growth poles located between large cities and rural areas. This would provide enough space for factories and housing, which was expected to slow the migration flow into great cities and metropolitan areas.
Prefectural and local governments, with the central government support, invested huge amounts of money to create new spaces available for new industries and people. Tax benefits and other favorable conditions were arranged for the private industries that were expected to come into those new cities.

This new industrial city policy was expected to have a double effect of population distribution. One was to provide economic motivations to attract new industries that otherwise might have selected location in large cites. Another aim was to attract existing industries in large cities, which would wish to move out because of increasing costs of the crowded environment.

In addition, various measures were implemented to stimulate local industrialization and expanded local transportation networks. In this way local employment opportunities were considerably increased.. The policy was based on the recognition that an important factor in population migration is the increase of diversified job opportunities. Jobs in the local areas would keep people from moving to the metropolitan areas in search of work.

To Top

The people's increasing awareness of environment disruption also motivated people to move to the local areas. Air and water pollution, traffic congestion, inadequate housing, noise disturbance, worsening social environment and all the costs of high urban concentration increased people's willingness to move to smaller cities, as long as jobs were available. In addition, young people's changing attitudes toward job selection also appeared to change migration streams away from the larger cities.

These policy and legal changes may have had an impact on the migration streams. We have already seen that in-migrants into the three large metropolitan areas (Tokyo, Hanshin and Chukyo) increased rapidly in the second half of the 1950s and they exceeded one million per year in the first half of the 1960s. Net in-migration in the these metropolitan areas attained highest number of about 650,000 in 1961 and 1962. After 1962, when the new policies were enacted, net-in-migrants began to decline and seemed to enter a new phase of migration streams.

We also noted above that the population of the medium-sized cities, 100,000 to 500,000, has shown the highest rate of increase and these cities have gained a dominant share of the country's urban population. At the same time, it was not the largest metropolitan areas, but the smaller cities that lost population to the medium sized cities. Thus the policies appears to have been only partly successful in changing the structure of urban population distribution. They halted, but did not dramatically reverse the attraction of the major metropolitan areas.

To Top

Net-in-migration to the three largest metropolitan areas began to decline in 1962, and finally reversed to net-out-migration. What are the factors responsible for such a drastic change of migration streams? We do not yet have general theory to explain migration mechanisms. I have discussed government policies and popular attitudes. There is, however, another set of strong forces in business fluctuations. The oil shocks, which occurred in 1973, caused an economic recession. Economic stagnation is supposed to playa dominant role in reducing the in-flow of population into the large metropolitan areas and at the same time in pushing the "U-turn" migration. It should not be overlooked, however, that "V-turn" people were fortunate to find job opportunities in local areas which had been developed much earlier by local development project. Thus policies can help to amplify the deconcentration migration streams that come with economic down turns.

In the 1980s another new trend is emerging reversing the previous "u-turn". Net inmigration into the large metropolitan areas is replacing the net-out-migration of the second half of the 1970s. We do not know the background factors causing this new concentration of people, especially in the largest, the Tokyo metropolitan area. Evidence to demonstrate any clear direction of migration flows is not yet available. It may be hypothesized, however, that concentration and deconcentration will be cyclical phenomena in a highly urbanized society, and will bring about changing patterns in the distribution of urban population in terms of the hierarchical structure of cities by size of population.

To Top

In any event, the unbalanced distribution of population over the national landscape is now very common in both developed and developing countries. It causes serious government concern even though the nature and extent of the social and economic implications of the imbalance are different in the two regions. The effective implementation of redistribution policy is a formidable task for any government. The policy requirements touch on various field of social, economic, and fiscal arrangements and of urban, rural and environmental problems.

The problem is rendered even more complex because the overall national situation is not necessarily applicable to individual cities. Some urban administrators are seriously concerned about the heavy migration from rural areas and the high fertility of migrants. Other administrators are concerned about the loss of population and the consequent social and economic stagnation that this implies. Eventually, however, the objective of administrators everywhere is to provide appropriate housing and employment opportunities to people and to provide the infrastructure and services needed to improve the people's welfare. In order to implement these basic policies, necessary data on population should be fully prepared.

Our major interest in this study is to make a comparative analysis of two contrasting cities in each of five countries. From this we hoped to learn about the perception and policy arrangements of responsible city administrators who are trying to cope with the urban problems facing them. Their social economic and demographic conditions may be quite different. It should be noted, however, that any effective policies would not be possible without taking into account of national background situation. Of course, policies and programs adopted by individual cities should be specific and appropriate for their own conditions.

In drawing attention to Japan's experience, we hope to show one possible future that may be facing the other less developed and less urbanized countries of Asia. This is most likely to include increasing urbanization, until the society is predominantly an urban society. Although this urbanization is expected to be an integral part of modern social and economic development, and to be therefore associated with rising standards of living, the resulting urban concentration can also cause serious problems. In some cases it can lead to a reduction of welfare. Japan's experience shows, however, that urban policies can help to reduce the great imbalance of high urban concentration. Perhaps closer attention to this experience can be of benefit to the other countries of Asia.

To Top

Back