Collapse and Reconstruction: Housing Recovery Policy in Kobe after the Great Hanshin Earthquake


Yosuke HIRAYAMA

Department of Human Environment, Kobe University, Japan
[Paper first received March 1999, in final form September 1999]

Abstract: This paper examines the housing recovery policy carried out in Kobe, a disaster city heavily damaged by the Great Hanshin Earthquake of 17 January 1995. The housing problems in the earthquake-hit city resulted not only from direct damage by the disaster. Urban restructuring, underway beforehand, had been generating socio-economic polarisation and geographical disparity in housing conditions. The earthquake caused especially heavy damage on the inner-city housing of low-income people and the elderly. Housing recovery progress in the post-disaster period has also been unequal. This paper shows the growing socio-economic and spatial polarisation. The framework of Japan's housing policy is a two-tiered system. On the one hand, most people are encouraged to obtain their own houses by their own efforts on the market, whereas on the other, public housing as residual welfare housing is directly provided for those who are marginal to the market. The housing recovery policy followed this framework, and functioned to socio-spatially isolate low-income and/or elderly victims.


Introduction

The Great Hanshin Earthquake occurred in the early morning of 17 January 1995. It was the greatest urban earthquake in Japan since the Second World War. The Hanshin area, including cities located on the southern portion of Hyogo Prefecture, is a major hub of population, industry, transportation and trade. This area suffered over 6000 deaths, 350 000 or more persons injured, approximately 200 000 collapsed and/or burned buildings, and the destruction of infrastructure such as public transportation, water supply, gas and electricity. The damage was extensive and the total cost of the damage was estimated as 10 trillion yen.
This paper examines the actual conditions and features of the housing recovery process for a period of 3 years after the earthquake in Kobe, the largest city in the Hanshin area (see Figure 1).
It is impossible to understand the housing conditions in the earthquake-hit city if they are examined as a completed phenomenon in itself. It is necessary to situate the circumstances around the earthquake in relation to the general nature of Japan's housing problems and policy. The housing damage and the process of restoration cannot be separated from the broader conflict of contemporary housing. Although the public policy for housing restoration included some special considerations such as increased subsidy by the central government, its fundamental framework followed the usual pattern.


The Characteristics of Housing Damage

Reflecting the geographical disparity of housing conditions, which existed before the event, the housing damage occurred unequally. In the post-disaster period, the housing recovery progress has been extremely unequal. The consequence has been an increased polarisation of housing situations in Kobe.
There was an obvious geographical differential in the housing situation in Kobe in the pre-disaster period. There were many new houses in the suburban wards with intensive recent housing investment. The proportion of owner-occupied houses was high in the middle- to high-income areas. This trend was notable in Nishi-ward where new towns had been developed. On the contrary, in the urban wards housing investment was stagnant and dilapidated houses, frequently of wooden structure, were densely crowded. Many of the residents belonged to the low-income strata and the proportion of low-quality private rental housing was high. Hyogo-ward and Nagata-ward in the western area had stronger characteristics of inner-city areas with notable over-aging of houses, residents on low income and an increase in the proportion of elderly population compared with the eastern area including Higashi-Nada-ward and Nada-ward.
Housing policy has magnified these geographical differentials. Urban planning policy by the City of Kobe has continuously implemented large-scale projects represented by developments of new towns in suburbs and artificial islands, and strategically concentrated investment in those areas. The City's housing policy has been related to such urban developments and has channelled housing investment to the suburbs. These policies have been focused on the expanding supply of housing units for middle- and high-income groups and has drawn those people out from the inner-city area to the suburbs.
In recent decades Kobe has been restructured like many other cities in Japan. The stability of urban society had been weakened by economic globalisation,
changes in production patterns and recent stagnation in economic growth. Kobe's economy was made up of typical heavy industries such as shipbuilding and steel industries. The inner-city area had many small-sized factories affiliated with the heavy industries. Local manufacturing industries in the form of chemical-shoe-making had formed the core of the local economy. It was apparent, however, that the heavy industries, the axis of the economy, were declining. Chemical-shoe-making had to compete directly with other Asian countries. The development of new industries related to information, fashion and tourism has not yet developed sufficiently to counteract the decline in manufacturing. Such restructuring functions to deepen the socio-economic polarisation. The inner-city areas have further declined with the outflow of capital, employment and population.
The housing situation in each ward before the event can be seen from the results of the 1993 Housing Survey of Japan (Statistics Bureau, 1995). As shown in Table 1, the proportion of owner-occupied houses was 61 per cent in Kita-ward and 68 per cent in Nishi-ward. However, it was 34 to 50 per cent in the urban wards. The time when houses were built differs greatly according to the wards as indicated in Table 2. The percentage of housing units built before 1960 is less than 5 per cent in Kita-ward and Nishi-ward, but it reaches higher than 30 per cent in Hyogo-ward and Nagata-ward. It shows that the proportion of old housing is growing in the inner-city area with housing disinvestment.
The index related to housing environment indicates that houses adjoining narrow roads and very small building lots are common in the inner-city wards. More than half of the houses are adjacent to roads with a width of less than four meters and approximately as much as 40 per cent of the single-family houses and terraced houses have a site area of less than 50 square meters in Hyogo-ward and Nagata-ward.
There are also marked income differentials between the suburbs and the inner-city area (see Table 3). The proportion of households with less than 3 million yen a year is less than 20 per cent in Kita-ward and Nishi-ward while it is 42 per cent in Hyogo-ward and 39 per cent in Nagata-ward. The ratio of households consisting only of elderly persons is less than 10 per cent in Kita-ward and Nishi-ward, whereas it is as much as 20 per cent in Hyogo-ward and Nagata-ward. Before the earthquake, there was a striking contrast between the suburbs and the urban area.
The housing damage caused by the earthquake reflects this geographically divided situation which had been growing prior to the event. The earthquake itself was a natural phenomenon, but the inequality of housing damage was generated socially.
The City of Kobe has made public the data about the number of housing units lost (see Table 4). The number of units lost in the city was 79 283 (15 per cent). The damage was concentrated in the urban area. The number of units lost in the urban wards was 74 234 and the ratio was as much as 24 per cent. The damage was especially heavy in old houses, wooden multi-family rental housing and wooden terraced housing structures. The ratio of housing units lost in urban wards was 58 per cent in those built before 1945, 45 per cent in wooden multi-family housing, and 58 per cent in wooden terraced housing. The residents affected were mainly of low-income households and the aged. The data by ward show that the proportion of units lost was particularly high in Nagata-ward with 39 per cent, followed by 25 per cent in Higashi-Nada-ward and 23 per cent in Nada-ward.
It should be noted that housing damage caused an enormous number of deaths. More than 95 per cent of deaths were directly caused by the collapse and/or the catching fire of their houses. About half of the deaths were people aged 60 or older.