The National Institute of Population and Social Security Research (Japan) projects future population both nationally and by region based on the Population Census of the Ministry of Internal Affairs and Communications. According to the latest projections in Regional Population Projections for Japan: 2020-2050, the number of municipalities whose populations will decline in the coming 30 years is 1,651, which is more than 95% of all 1,728 municipalities (excluding 13 municipalities in the Hamadori region of Fukushima Prefecture, which are projected as a single region). The rate of population decline will tend to be higher in municipalities with smaller populations in non-metropolitan areas. The main factor behind this decline is natural decrease caused by low fertility. The pressure of population decline due to natural decrease is growing stronger over time. The Tokyo metropolitan area, which currently has a growing population due to net migration, will also begin to see its population decline.

(Shiro KOIKE)

 Medium- and long-term population trends in Japan's rural regions and factors underlying changes are surveyed. In the 20th century, because Japan was at a stage in its demographic transition with high birth and low death rates, both non-metropolitan and metropolitan areas experienced large population growths. In rural regions, social decreases continued due to migration to metropolitan areas; however, in the 20th century, natural increase compensated for the decline and population increased. The number of births continued to decline because the birth rate remained below the population replacement level, and the number of deaths inevitably increased as a large cohort entered old age. Therefore, populations of rural regions experienced natural decreases, which are expanding. As a result, most population declines in rural regions can be explained by a natural decrease. On the other hand, social decreases also contribute to population declines in rural areas. A comparison of cohort changes reveals that the rate of females remaining at place of origin is declining in some regions. In addition, social increase rates of municipalities are positively correlated with population size. From this, it can be said that the migration of urban residents to rural settlements, as reported in the media, has been limited.

 Spatial population distribution changes within municipalities between 1995 and 2020 in the Tohoku region (Aomori, Iwate, Miyagi, Akita, Yamagata, and Fukushima prefectures) are examined. Using 4th mesh data from the Population Census of Japan to estimate the populations of central and peripheral areas of municipalities, municipalities are categorized into four types: absolute concentration, where both a central area's population and its percentage within a municipality increase; relative concentration, where a central area's population decreases, but its percentage increases; relative dispersal, where a central area's population increases but its percentage decreases; and absolute dispersal, where both a central area's population and its percentage decrease. Geographical patterns of these types are analyzed, as well as the relationships between these types and the municipal typology of Urban Employment Area (Large Urban Core Areas (LUCA), Large Urban Suburb Areas (LUSA), Small Urban Core Areas (SUCA), Small Urban Suburb Areas (SUSA) and Outside of Urban Employment Areas (OUEA). The main results are as follows. First, while the populations of all prefectures and most municipalities have declined since 1995, they tend to be concentrated in Miyagi Prefecture at a prefectural level and in prefectural capitals and their surrounding municipalities, as well as in cities with relatively large populations and their surrounding municipalities at a municipal level. Second, in most municipalities, population tends to be concentrated in the central area of a municipality. Most of these municipalities are of the relative concentration type and are found in OUEAs, SUSAs, and LUSAs. It is highly likely that spongification—sporadic population and household loss from residential areas—occurs throughout these municipalities because it is difficult to assume that the population concentration in a central area through relocation from peripheral areas to the central area has been achieved, given that the social growth rate is negative in both central and peripheral areas. Third, many municipalities of the absolute concentration type are located in LUCAs, LUSAs, SUCAs, and SUSAs, and both social and natural growth rates at the centers of these municipalities are positive. Fourth, most municipalities with absolute dispersion are found in LUCAs, LUSAs, SUCAs, and SUSAs. Among them, there are those with a doughnut-shaped distribution of population concentration areas due to a degradation or hollowing out of the city center, and these municipalities are classified as a type of relative concentration when the coverage of the central area is expanded. These municipalities include those that provided higher-order central location functions, that is, LUCAs and SUCAs. Although it has been argued that population tends to be concentrated near the center at a metropolitan level, it is not necessarily concentrated in the central area of a municipality at a sub-municipal level. Fifth, the natural growth rate of central and peripheral areas has decreased in all types of municipality in recent years, and its contribution to the population growth rate has tended to increase. In contrast, the social growth rate is highly correlated with regional differences in the population growth rate. Based on these results, it can be concluded that spongification occurs in a wide range of areas of the Tohoku region, whereas, simultaneously, population tends to be concentrated in areas that provide relatively higher-order central functions.

 In Japan, while depopulation, aging, and urbanization have caused inter-prefectural migration to decline, net migration to the Tokyo metropolitan area (Tokyo), the Osaka metropolitan area (Osaka), the Nagoya metropolitan area (Nagoya), and the non-metropolitan area (Non-MA) has generally been maintained. The mechanism of this phenomenon is clarified. Using the number of migrants in the Annual Report on Internal Migration in Japan derived from the Basic Resident Registers and the number of people from the Population Census of Japan, out-migration, crude out-migration rate, and standardized out-migration ratio by destination are analyzed, and changes in out-migration are decomposed into demographic and mobility factors. The results show that propensities of inter-metropolitan migration to and from the Tokyo have been increasing, especially from Osaka and Nagoya to Tokyo, and are more pronounced among women than among men. However, migration propensities between Osaka and Nagoya weakened slightly. On the other hand, migration propensities from metropolitan areas to Non-MA declined substantially, whereas those from Non-MA to metropolitan areas were moderate, and those to Tokyo and Nagoya increased or were maintained. In the discussion, first, the mechanism behind the persistence of net migration in the four regions is explored from the viewpoints of migration propensity and the hypothesis of mobility transition. It is argued that while the attractiveness of Tokyo to people from Osaka and Nagoya increased, the attractiveness of Non-MA to people from the metropolitan areas decreased significantly, and that the attractiveness of Osaka relative to Nagoya weakened. In addition, the results shown for Japan are generally consistent with those presented in the hypothesis of mobility transition proposed by Zelinsky. Their association with the socioeconomic context behind the persistence of net migration is discussed. Because Japan's economy has stagnated and local economies of non-MA have been shrinking due to population decline and aging, people are more likely to think that better opportunities exist in Tokyo or Osaka or Nagoya. This could have increased the attractiveness of Tokyo among people in Osaka and Nagoya, and decreased the attractiveness of Non-MA among people in metropolitan areas.

 The hypothesis is tested that municipal mergers promote migration from peripheral areas, where a municipal office has been lost through a merger, to central areas. Multiple regression models are developed to predict migration during the period 2005-2010 between pre-merger municipal areas using year 2000 census data, which is just before the Heisei Municipal Mergers. In these models, migration is classified into four categories: internal, from periphery to center, from center to periphery, and between peripheral areas. If the hypothesis is true, migration from periphery to center should increase compared to the other categories. In fact, a regression analysis reveals that migration from periphery to center did not increase, but the opposite form of migration, from center to periphery, increased remarkably. Therefore, the results indicate that municipal mergers do not promote the centralization of population but may promote suburbanization.

 Changes in population migration are an important research theme in relation to the demographic dynamics of regional areas. Actual trends and changes in career paths and destinations of high school graduates (university, junior college, professional training college, and employment) since 1990 in Nagano Prefecture are analyzed, and the relationships between these paths and demographic dynamics are examined. The results show: declining birth rate and changes in career paths of high school graduates relate to a decrease in the number of high school graduates who remain in the prefecture after graduating from high school; characteristics and changes in destinations of high school graduates who go on to professional training colleges and universities and junior colleges differ; and, careers and further education courses of high school graduates differ between cities and localities where high schools are situated. This study has social and academic value in that it succeeds in identifying trends and characteristics of migration to professional training colleges and in understanding the migration of high school graduates on a micro-scale rather than on a prefectural basis, which had been an issue with previous studies. It is an important research task to clarify actual trends in the region using not only statistics that have often been used in the past but also those surveyed by local governments.