Migratory flows involving an urban environment include movements into, out of, and within towns and cities (Lewis, 1982, p.118). Urban migration refers to all of such movements, that is, in-migration, out-migration and intra-urban migration. Previous studies have put emphasis on differences between the former two and the last one types of migration. But both yield insights into the urban residential mobility.
The purpose of this study is to elucidate distribution patterns of urban migration in the Kanto district where rapid urbanization has created the most advanced society in Japan. 125 administrative cities in the Kanto district constitute one dimension of the geographic matrix (Figs. 1 and 2). And the other dimension represents attributes of migration, which are divided into four variable groups named as mobility (Table 1), sex ratio (Table 4), age structure (Fig. 5) and migration field (Fig. 8). It sums up to 61 variables and the data are derived from the Population Census of Japan 1970 and 1980. In order to eliminate the garrulous correlation, the principal component analysis is applied to the variable groups individually (Tables 2, 5, 7 and 8). Then, by means of Ward's method of cluster analysis, the 125 cities are classified into three types in the mobility (Table 3 and Fig. 3), four types in the sex ratio (Table 6 and Fig. 4), five types in the age structure (Figs. 6 and 7), and seven types in the migration field (Figs. 9 and 10). The results are summarized as follows (Fig. 11) :
1. The cities with low mobility are distributed in the outer suburban zones of the Tokyo Metropolitan Area. In the inner suburbs, cities show high mobility of the in-and out-migrations. And, in terms of intra-urban migration, high mobility characterizes the cities with large population or high centrality.
2. It can be observed that Tokyo and its inner suburban cities tend to be male-dominant in the in-migration. However, it is not sure whether the distribution has a zonal structure or not, and the sex selectivity is of less significance especially in the intra-urban migration.
3. The youth labor force aged 15-19 is the most important age group in explaining the distribution pattern of age structure. The out-migration of this age group shows a very strong contrast between the high mobility level of cities located in the outer suburban zones of the Tokyo Metropolitan Area on the one hand and the low mobility level of those located in the inner suburbs on the other. Additionally the cities with high centrality are characterized by the high mobility level of in-migrated 15-19 age group.
4. Generally speaking, the number of in-and out-migrants decreases rapidly with the increase of distance to travel, and thus the migration field exhibits a horn-like shape which is common to all of the cities. However, there are differences between those migration fields in terms of the location of another peak of shape, which represents the distance from Tokyo and determines the types of migration field. Therefore it becomes very apparent that the distribution of those types forms a concentric pattern centering around Tokyo.
5. In quantitative comparison between the four distributions of urban migration mentioned above, the intensity of concentric pattern becomes lower in the order of migration field, mobility, age structure and sex ratio (Table 9). Also, in terms of urban population size, it can be confirmed that there is the high level of correspondence between the two distributions, that is, mobility and city size (Table 10).