This study aims to clarify U-turn migration, i.e., the return migration of people who once left peripheral regions for metropolitan areas in Japan. U-turn migration has been widely discussed, but it has not been satisfactorily substantiated due to a lack of adequate data. We conducted a questionnaire survey of male graduates from 12 highschools in Nagano Prefecture (1956-1958, 1966-1968, and 197.6-1978 graduates). We investigated their migration histories and, in those cases in which the relocation was a U-turn, their reasons for migrating. Twelve thousand questionnaire sheets were distributed, and 3, 825 were returned.
The analysis of the survey results revealed, first, that the rate of U-turn migration from the three largest metropolitan areas to Nagano was progressively higherr for the younger generation. Among the younger generation, the return migration at graduation from university accounted for the largest proportion of all the U-turn migration, reflecting a rise in educational background.
Second, the majority of the U-turn migrants returned to the municipalities where they had grown up, and the tendency gradually became stronger. Although U-turn migrants contributed to the population growth of the leading cities in the prefecture, attention must be paid to the role of the migrants who came directly from other municipalities in the prefecture to the leading cities.
Third, our analysis confirmed that a man's decision to U-turn was affected by his academic background and by whether he was the eldest son, as suggested by previous research. This study, however, also demonstrated that the wife's birthplace was a much stronger factor in U-turn decisions. In other words, graduates who married women born and raised in Nagano Prefecture were more likely to return than those who married women from other prefectures.
Fourth, among the graduates whose U-turn migration was accompanied by a job change, the obstacles to relocation were perceived to be the shortage of desirable jobs in Nagano Prefecture and the accompanying decrease in income. These obstacles were verified by analyzing the respondents' objective conditions, such as the actual change in their income level.
Finally, in terms of the timing of a U-turn migration with job change, it was revealed that the majority of the migrants studied had returned to Nagano within five to eight years after first getting a job. This indicates that U-turn migration mostly took place in a relatively early phase of their lives in metropolitan areas. Therefore it can be concluded that U-turn migration is an option that the young migrant population who once migrated from rural regions may choose at the first turning-point in their careers, with relatively little restrictions due to occupational and residential choice.

View full abstract

Assuming that building damage caused by the Hanshin-Awaji Earthquake was due not only to the strength of the quake but also to the social and physical environment, the author investigated the varying scale of building damage related to tenure type of the land and houses. The study also aimed to examine why housing damage was concentrated in the inner-city area. This study is based on the author's field work of damage inspection and on the analysis of both official statistics and registry records of land and houses. The results are summarized as follows:
1) A clear difference building damage was found based on the tenure of land and houses: noncollapsed or slightly damaged houses totaled 29.1% of “owner-occupied freehold” houses, but only 2.2% of “rented” houses.
2) The older the buildings were, the higher the rate of damage: 81.1% of houses built before 1959 were completely collapsed. On the other hand, only 13.0% of houses built after 1980 collapsed.
3) Buildings with smaller ground floor space suffered more serious damage: 79.2% of buildings of less than 30 square meters were completely collapsed.
4) Building structure was also related to the damage : more than 90% of dwellings in the study area consisted of wooden houses, of which 70% were completely collapsed. Of wooden houses, however, only 20% of buildings constructed after 1980 were completely collapsed. Because most steel frame and reinforced concrete structures were built in the past two decades, they did not collapse or were only slightly collapsed.
5) The damage ratio of old rented houses was higher than that of owner-occupied freehold houses.
6) Regardless of ground floor space, the ratio of completely collapsed houses was higher among rented than among owner-occupied freehold houses. This tendency was more noticeable among houses with large ground floor space.
7) The proportion of small buildings was higher in the category of rented houses than in the owner-occupied freehold category.
8) The larger the ground floor spase, the lower the ratio of old buildings: most buildings with less than 30 square meters of ground floor space were built before 1959. Less than 50% of the buildings with 40-50 square meters were built before 1959, although the ratio of old buildings was higher in the category of larger houses (50 square meters or more). New buildings and rebuildings were mainly found in the 40-50 square meter category. This suggests that (a) the earthquake caused moredamage in the category of small, old houses, most of which were rented houses, and (b) medium-sized houses were newly built or rebuilt in the past 20 years, which mainly comprise the owner-occupied freehold category.
9) Individual rebuilding of small, old houses is difficult, because most are located in the physical environment of small sites along narrow streets less than 4 meters wide. The most typical small, old buildings are tenement houses, which comprised 80.1% of all houses before 1959 and 86.6% of houses with a floor space of less than 30 meters.
10) Until the end of the Taisho era, the study area had been developing into residential quarters for workers, which mostly consisted of small rented houses. Most of these rented houses continued to exist in densely built-up environment of the inner area of Kobe until the earthquake, sinse many parts of this area were not affected by the air raids of World War II.
11) After the war, house and land ownership became progressively more fragmented because of the constant increases in taxes. This fragmentation has mainly led to a change from rented houses, to owner-occupied freehold houses, leading to a very complex ownership pattern. This is one of the main factors that put a brake on the renewal process.
12) One of the reasons for the delay in the renewal process is that most of the small, old houses were occupied by long-term elderly residents.

View full abstract

Some peculiar fissures cutting across mountain ridges appeared after the 1995 Hyogoken-Nanbu Earthquake in the Rokko Mountains. This paper describes the characteristics of the fissures and examines the process by which they were formed, with special reference to the response of the mountain mass to the force generated by the earthquake.
Detailed mapping was carried out at the sites where the fissures are located. The magnitude and direction of horizontal acceleration on the ground surface due to the earthquake were also estimated, in addition to determination of the dislocation of artificial constructs such as asphalt pavement and concrete side-drainage ditches. The horizontal acceleration was determined through the dynamics of mass using the measured friction coefficient and the inclination of the contact surface between moved objects, such as large boulders or gravestones, and the underlying bases.
Cross-ridge fissures were found in three locations in the Rokko Mountains: the first is on the Gakko-rindo ridge situated in the western part of the Rokko Mountains; the second is around the Mt. Yubunedani-yama ridge, Arima, in the northern part ; and the third on the Kanten-sando ridge to the south. The fissure systems at the study sites generally consist of a series of fissures which cross obliquely, cutting through the mountain ridge and the shoulder of the landslide escarpment. They are characterized by an echelon-shaped or parallel arrangement in plane view, and reach a maximum of 50 m in length as a whole. Moreover, the fissures are associated with an open crack of several to 20cm wide, a small-scale cliff with a maximum height of 50 cm, or both.
Based upon the estimated values of horizontal acceleration, all the research sites with fissure systems received force of 1g (g=gravitational acceleration) or more. These values are significantly higher than expected. Therefore a concentration of strong force at the given sites resulted in fissure formation.
Based on both the direction and magnitude of the inferred horizontal acceleration on the ground surface, two types of pattern generated the fissure systems. One is the Gakko-rindo type, in which the directions of acceleration are same over the fissures, but with sufficient difference in magnitude to create a fissure across the ridge. The other is the Arima Yubunedani-yama type, in which the magnitude of acceleration is nearly same, but the direction is entirely different or almost opposite over the fissures. The fissure system of Kanten-sando may be transitional from the Gakko-rindo type to the Arima Yubunedani-yama type. Thus it can be concluded that the fissure systems discussed here are tectonic ones triggered by the earthquake.

View full abstract