Cover Image
close this bookCrucibles of Hazard: Mega-Cities and Disasters in Transition (UNU, 1999, 544 pages)
View the document(introduction...)
View the documentAcknowledgements
View the document1. Introduction - James K. Mitchell
View the document2. Natural disasters in the context of mega-cities - James K. Mitchell
View the document3. Urbanization and disaster mitigation in Tokyo - Yoshio Kumagai and Yoshiteru Nojima
View the document4. Flood hazard in Seoul: A preliminary assessment - Kwi-Gon Kim
View the document5. Environmental hazards in Dhaka - Saleemul Huq
View the document6. Natural and anthropogenic hazards in the Sydney sprawl: Is the city sustainable? - John Handmer
View the document7. Disaster response in London: A case of learning constrained by history and experience - Dennis J. Parker
View the document8. Lima, Peru: Underdevelopment and vulnerability to hazards in the city of the kings - Anthony Oliver-Smith
View the document9. Social vulnerability to disasters in Mexico City: An assessment method - Sergio Puente
View the document10. Natural hazards of the San Francisco Bay mega-city: Trial by earthquake, wind, and fire - Rutherford H. Platt
View the document11. There are worse things than earthquakes: Hazard vulnerability and mitigation capacity in Greater Los Angeles - Ben Wisner
View the document12. Environmental hazards and interest group coalitions: Metropolitan Miami after hurricane Andrew - William D. Solecki
View the document13. Findings and conclusions - James K. Mitchell
View the documentPostscript: The role of hazards in urban policy at the millennium - James K. Mitchell
View the documentAppendices
View the documentContributors
View the documentOther titles of interest

1. Introduction - James K. Mitchell

The Great Hanshin earthquake of 17 January 1995 was a signal event in the history of urban disasters.1 Not only was it Japan's most deadly and destructive natural disaster in over 70 years,2 it also raised disturbing questions about existing hazard-management policies and programmes that had been regarded as among the most effective in the world. Despite decades of attention to the goals of hazard reduction by Japanese governments, industries, and citizens' organizations, over 6,000 residents of the country's second-largest metropolitan area3 were killed, 10 times as many were injured, and large parts of the Kobe - Osaka urban region experienced heavy damage and disruption.4 Fires took hold rapidly and burned out of control, structures and lifelines that had been designed and built to hazard-resistant standards gave way, emergency management operations failed to live up to expectations, and recovery programmes dragged on well beyond their anticipated termination dates.5 Not since the massive Kanto earthquake of 1923 devastated Tokyo and Yokohama has a major Japanese urban area been so grievously stricken by natural disaster. Indeed, this was the first time that Japan's annual disaster death tolls have climbed back above double digits into the thousands since the Ise Bay typhoon of 1959 killed over 5,000 people and triggered a major restructuring of the country's hazard-management systems. Economic losses may have exceeded a staggering US$ 100 billion!6

1 The Great Hanshin earthquake is also known as the Hanshin-Awaji earthquake and the Hyogoken-Nambu earthquake. It occurred at 5.46 a.m. and registered 7.2 Ms (surface wave magnitude) and 6.9 Mw (moment magnitude) on the Richter scale. The quake's epicentre was located in Osaka Bay just south of the port city of Kobe. Major damage occurred in Kobe, with lesser effects in Osaka and adjacent communities. For details, see the special issue of the Journal of Natural Disaster Science 16(3), 1995.

2 It is estimated that 6,300 people were killed by the Hanshin earthquake. This surpasses the 1959 Ise Bay typhoon, which brought about over 5,000 deaths in and around the city of Nagoya, but does not compare with the Kanto earthquake of 1923, whose ground motion and fires destroyed over one-third of Tokyo and most of the adjacent port of Yokohama, killing more than 140,000.

3 The Keihanshin metropolitan region (Osaka - Kobe) contained approximately 14 million people in 1990, second only to the Tokyo region, which had almost 27 million.

4 In this book no attempt has been made to standardize estimates of city populations and city areas or quantitative estimates of disaster impacts. Varying estimates faithfully reflect the diversity of measurement criteria adopted by urban analysts and disaster assessors as well as the provisional nature of information about rapidly changing places and events.

5 Contributory factors to the disaster included, among others: a large stock of wooden houses in crowded neighbourhoods that were developed before the Second World War and never subsequently upgraded; infrastructure that was haphazardly installed during the heady post-war years of Japan's economic expansion; and transportation facilities that were seriously underdesigned for the risks that they faced (see Ichikawa, 1995).

6 American usage of the term "billion'' (i.e. 109) is employed throughout this book.

The Hanshin earthquake is just the most recent in a string of natural disasters that have inflicted unprecedented losses on very large cities (table 1.1). Often these have involved earthquakes, but hurricanes and wildfires have also led to heavy losses. Though the upward trend in economic losses is most striking, death tolls have also been substantial. These events have far-reaching implications for much of the world's population. Among others, they suggest that cherished notions about the security of cities in the face of natural extremes are no longer tenable and that disasters in large cities are likely to pose troubling new problems for society. Viewed against the emergence of a predominantly urban world where people increasingly live in giant urban agglomerations (i.e. mega-cities7), recent disasters also underscore the potential for even larger losses in the future.

7 As used here, "mega-city'' means a metropolitan area with a population of at least 1 million people. Other analysts have employed higher population thresholds (e.g. 3 million, 8 million, or 10 million), but the term has no universally accepted definition.

Table 1.1. Recent natural disasters losses in mega-cities

Year

Location

No of deaths

Cost (US$ billion)

1985

Mexico City

5,000-10,000+

9

1989

San Francisco Bay, CA

62

8

1991

Oakland, CA

25

2

1992

Miami, FL

38

30

1994

Northridge, CA

57

20

1995

Kobe, Japan

6,300

100+

Sources Official loss estimates provided by various national and international agencies

Note Figures are best available estimates

This book focuses on natural hazards and disasters in mega-cities partly because of their potential for catastrophe. That does not mean that other kinds of hazards are incapable of producing urban catastrophes. Wars have frequently been associated with large-scale destruction of urban areas, especially in the twentieth century (e.g. Hiroshima, Dresden, Phnom Penh, Kabul). Political terrorism and crime are also potent agents of urban destruction. So too are hazardous industrial technologies (e.g. Bhopal, Texas City). In view of this record, why single natural hazards out for special consideration?

The answer is that natural hazards are joint products of nature and society. Unlike the other threats just mentioned, they are only partly created by humans.8 This gives them a special role in debates about humanity's future because they are not, ipso facto, entirely susceptible to human will. They represent an "other'' that can be modified by humans but is not ultimately reducible to a human construction, in either the material sense or the mental one. In other words, natural hazards invite humans to recognize that our knowledge of the Earth and its peoples is incomplete, uncertain, disjointed, and subject to penalizing contingencies beyond our control. It is likely to remain so in the foreseeable future. We should prepare our institutions and environmental management strategies for the twenty-first century with this firmly in mind, especially in the mega-cities that will likely become the pivots of global society. For mega-cities are, in effect, crucibles where new kinds of hazards are being fashioned and old ones reshaped so that existing ways of dealing with both are thrown into doubt.

8 This is a distinguishing feature of some other contemporary debates, including those about the demise of certain ancient cities and the consequences of anthropogenic atmospheric warming.

Crucibles of Hazard explores the emergence, re-emergence, and transformation of environmental hazards in contemporary mega-cities. Although this process is driven primarily by changes in the location, size, structure, and functions of cities and in the composition of city populations, many other factors - including global economic restructuring and global environmental change - are also involved (International Decade for Natural Disaster Reduction, 1996). As table 1.1 shows, leading indicators of the transformation are already detectable. However, the complete results of this large-scale experiment in Nature - Society relations will not be clear for decades to come.

The thesis that natural hazards are joint products of nature and society is one of two main principles that underlie this volume. The other is that mega-cities are intensely human-constructed environments that - to varying degrees - shield their inhabitants against natural processes. When the shields are inadequate, as is often the case for poor populations in many third world cities, the human engagement with hazard can be immediate, personal, and highly unpleasant (Main and Williams 1994). But, for most people in large cities and for most of the time, extremes of nature are experienced indirectly in the form of signs and symbols that act as cues to thought and behaviour. In other words, mega-city residents are constantly bombarded with information about hazards - which is only occasionally put to the test by actual events. Unlike the simpler hazard-scapes of rural areas and small towns, evidence of mega-city hazard crowds in upon the senses, demanding consideration of many different threats and responses: direct and indirect; past, present, and future; real and imagined; stated and implied - often in startling juxtaposition.

An hour's drive along the Tijuana - San Diego urban corridor provides many examples. Here, in one of North America's fastest-growing mega-cities,9 a plethora of natural and human-created risks is embedded in a jumble of contrasting neighbourhoods, institutions, and peoples from the first and third worlds. Evidence of hazard is plentiful - though often contradictory, ambiguous, and paradoxical. In one locality, affluent homes crowd along an eroding beach in the shadow of a massive bullring. Minutes away, new cars topped with surfboards and golf bags stream through an expressway cutting that is overhung by droop-shouldered wooden shacks clinging to a much-eroded slope. Nearby on an outdoor movie set, a large-scale replica of the ocean liner Titanic is poised in the act of sliding beneath the Pacific, within sight of a tollroad where wrecked vehicles have been placed in conspicuous positions to caution unwary drivers. Concrete flood-control works line the course of the Tijuana River, which usually carries only a trickle of heavily polluted water that is ultimately returned, untreated, to Mexico via a pipeline from the United States. Signs on US Interstate 5 warn motorists to beware of colliding with illegal immigrants, most of whom now prefer to avoid this well-patrolled road and to risk death from exposure in barren mountains that gird the metropolis. Further north, submarines and aircraft carriers ply navigation channels past slumping coastal bluffs and endangered migrating whales. Visitors to the San Diego Historical Museum learn about the central role of droughts and floods in the city's history, while the building - hastily concocted as a temporary exhibit for a 1915 Exposition - vibrates in the wake of large passenger jets threading down through hills and houses to land on a city-centre runway. These examples do not exhaust the range of hazards that occur in just one mega-city, but they are sufficient to illustrate a typical cross-section of threats and to hint at the complexity of contributory factors that must be taken into account by populations at risk and by urban managers. In the contemporary mega-city, knowledge and ignorance of threats help to shape the burden of hazard; so do affluence and poverty; human propensities for taking and avoiding risks; professional training and political adroitness; intended actions and unintended reactions; lessons from past realities and hopes of idealized or fantasized futures, as well as many other factors.

9 The combined populations of both cities are close to 4 million, of whom a slight majority live in the United States.

Global scientific, engineering, and hazard-management organizations have taken note of the increasing disaster potential of mega-cities (Ichikawa, 1995). While curiously silent about natural disasters, the global community of urban leaders, urban managers, and urban researchers has focused attention on other human crises of mega-cities such as rapid population growth, chronic unemployment and underemployment, inadequate housing, poor public services, weak pollution controls, dangerously decaying infrastructures, fractious relations among different ethnic and racial subpopulations, and insufficient citizen access to decision-making systems (see Postscript). To many observers, the crises and disasters of humankind and of Nature are interrelated; a comprehensive approach to all kinds of urban environmental hazards - natural, technological, biological, and social - is warranted. However, efforts to reduce the burgeoning natural hazards of mega-cities have been slow to develop and joint initiatives to address the full range of mega-city hazards have not yet occurred. More progress is urgently needed. This book is a contribution to the task.

Crucibles of Hazard is based on a set of invited papers delivered at the International Conference on Megacities and Disasters (Tokyo, 10 - 11 January 1994) and subsequently updated for publication. The conference was sponsored by the United Nations University, which is also engaged in a broad range of projects on mega-cities.10 Initial contributed papers were later supplemented by others to provide a wider spread of case-studies drawn from different parts of the world. The Tokyo conference also functioned as a springboard for initial studies carried out by the International Geographical Union's Study Group on the Disaster Vulnerability of Megacities (Parker and Mitchell, 1995; Mitchell, 1995; see also other articles in special issues of GeoJournal, November 1995, and Applied Geography, January 1998, Vol. 18, No. 1).

10 See, for example, Buendia (1990), Fuchs et al. (1994), Gilbert (1996), Lo and Yeung (1996), and Rakodi (1996).

The core of the book consists of 10 chapters that highlight environmental hazards in specific mega-cities on five continents. The case-studies are: Tokyo, Seoul, Dhaka, Sydney, London, Lima, Mexico City, San Francisco, Los Angeles, and Miami. The emphasis is on geological, meteorological, and hydrological hazards, but biological, technological, and social hazards are also addressed. A wide spectrum of places and communities is included (table 1.2). There are older slow-growing cities in more developed countries (e.g. London) and newer fast-expanding ones in less developed countries (e.g. Dhaka). Half of the cases are drawn from English-speaking states that share many other cultural traits (the United States, the United Kingdom, and Australia) and half from culturally varied Asian and Latin American countries. Two of the world's most populous urban centres are included (Tokyo and Mexico City) together with a variety of smaller cities (e.g. Miami and Sydney). Some of the case-study cities are fulcrums of the international economic system (e.g. the "global cities'' of Tokyo and London) whereas others are less strategically placed (e.g. Lima), but most can be counted among the so-called "world cities.'' Apart from the three US cases and Sydney, the cities are national capitals and the biggest urban agglomeration in their respective countries.

Table 1.2. Profiles of case-study cities

Citya

Population (million)

Annual population growth (%)

Political function

Economic functionb

Tokyo

268

140

National capital

Global city

Seoul

116

198

National capital

World city

Dhaka

78

574

National capital

-

London

73

00.0c

National capital

Global city

Sydney

36

047

-

World city

Lima

75

330

National capital

-

Mexico City

156

078

National capital

World city

San Francisco

39

124


World city

Los Angeles

124

172

-

World city

Miami

21

167

-

World city

Sources United Nations Commission on Human Settlements (1996), pp 21-22, 451-456, World Resources Institute (1996), pp 10-11, Brunn and Williams (1993). p 19.

a. Mega-cities in less developed countries are underlined.

b. Global cities are commanding nodes of the global economic system, world cities articulate large national economies into the global system.

c. Not known, but assumed to be static.

All of the case-study cities are acknowledged to be "large'' or "very large,'' with populations ranging in size from 2 million to nearly 27 million. Precise rankings by size are unavailable and may be misleading because territorial definitions of cities are not consistent. The case-studies employ the definitions and size criteria that are used in the different mega-cities. The results may or may not be the same as those found in standardized city rankings published by the United Nations or other sources (see appendix 1). Thus, here Tokyo (i.e. the Tokyo Metropolitan Region) contains around 27 million people and London (i.e. Greater London and the Outer Metropolitan Area) houses approximately 13 million. Likewise, the urbanized area of Dhaka is 155 sq. miles whereas that of Los Angeles is 34,000 sq. miles. To some extent, gross disparities such as these are a function of choices made by the experts who establish urban classifications, but they also reflect real contrasts between sprawling automobile-based cities in more developed countries and compact animal-powered ones in some less developed countries.

The case-study authors are drawn from various countries and professions. Half (six) are geographers trained and working in the United States or the United Kingdom. The authors of the Asian and Latin American case-studies come from other fields: urban studies and urban planning (Kumagai, Nojima, Puente); anthropology (Oliver-Smith); landscape architecture (Kim); and development planning (Huq).

In chapter 2, I contend that mega-city hazards are difficult to address not just because urban natural processes are complex or because there are many competing urban issues, but also because urban areas are changing rapidly. I believe that the members of hazard interest groups and urban-management interest groups are typically mobilized by different contingencies and separated by wide gaps of experience, training, problem conceptualization, and professional outlook. Before hazard-sensitive urban sustainability can become a reality, the potential for mobilizing action in response to varied contingencies must be known and the gaps that separate the interest groups must be bridged. I liken the task to one of managing contacts between mutually unfamiliar cultures and I recommend that the goal of reducing mega-city hazards be pursued in concert with the resolution of other problems that face these places.

The case-studies begin with Tokyo, one of the premier "global'' mega-cities, one of the most hazard prone, and one that possesses a highly sophisticated system of hazard management. Chapter 3 reviews the far-reaching influence of management policies for earthquakes, fires, and related hazards on the development of Tokyo during the past century. Yoshio Kumagai and Yoshiteru Nojima point out that the city is now undergoing its third major rebuilding since the Great Kanto earthquake of 1923. Whereas the once frenetic pace of national legislation on hazard management seems to have slowed, the government of Tokyo continues to plan for and implement many metropolitan hazard-reduction policies. These undergird a three-tiered strategy that combines: (1) extensive training of citizens to assume immediate responsibility for local hazard-fighting in the event of a sudden disaster; (2) land-use and building controls that permit evacuation and sheltering of exposed populations while retarding the spread of acute hazards; and (3) advanced prediction and warning systems that are intended to be linked with burgeoning technologies for the prevention of disasters. This chapter reflects the unique blend of inputs from science, technology, government, management, and citizen action that has been the hallmark of Japanese approaches to urban hazard reduction for many decades and that has been held up as a successful example to others. It was prepared before the Hanshin earthquake levelled much of Kobe and called into question widely held assumptions about the appropriateness of existing Japanese policies. Whether this system will continue unchanged in the light of Kobe's unwelcome experience remains to be seen.

Though floods, storms, and landslides are among the natural hazards of Seoul, they are neither as widely recognized nor as well documented as other environmental problems such as air pollution, traffic congestion, and shortages of affordable housing. But floods have inflicted substantial damage since the Second World War and - without proper attention - are likely to move into the front rank of Seoul's public concerns early in the twenty-first century. In chapter 4, Kwi-Gon Kim reaches this conclusion by way of a quantitative flood hazard assessment based on Geographic Information Systems that is believed to be the first of its kind employed in this city. A number of factors suggest that existing measures for reducing floods are already severely constrained. Flash-flooding challenges current prediction systems, engineering works are subject to failure, and there are no flood-control reservoirs in the surrounding hills. In addition, adequate flood insurance is unavailable in Korea, natural hazards are not taken account of in Seoul's master development plan or in local action plans, and flood-risk zoning is unheard of by those who manage municipal lands. Without a more determined effort to incorporate flood hazards into future development strategies, the prospects are for fast-rising losses.

From the viewpoint of effective public programmes for hazard reduction, the situation in Dhaka (chap. 5) is considerably worse than in Seoul. Not only is the city surrounded by areas that are at grave risk from severe storms and vast floods; it also faces multiple burdens of very rapid population influx from rural communities, grinding poverty, under-financed and inefficient urban services, and a disadvantageous location on the periphery of the global market-place. Saleemul Huq offers an overview of these problems that underscores the links between rural hazards and urban ones, and between local hazards and the global economy. He also illustrates the importance of dykes and embankments not just as flood-protection works but as centrepieces of planning strategies for land reclamation in support of urban expansion and as barriers to the drainage of polluted water that becomes trapped behind them. Increased industrialization is beginning to displace inner-city residences and to fill the last remaining buildable spaces within existing municipal boundaries. The city now seems poised on the edge of a major expansion phase that will push new developments into locations that were once subject to frequent floods and now are likely to be affected by less frequent but larger events.

John Handmer explores the growing vulnerability of Sydney to an expanding list of environmental hazards and addresses the implications of these for urban sustainability (chap. 6). He organizes the analysis around several contradictions. One is the national myth of Australia as a land of severe natural hazard, which does not match the experience of a city where such problems have not historically been very serious. Second is a carefully nurtured image of Sydney as a vibrant centre of hedonistic recreation, high culture, and international investment, which sits uncomfortably with the daily realities of life in sprawling undistinguished neighbourhoods that are chronically underserviced, subject to growing social polarization, increasingly polluted, and at risk from environmental extremes. Third is the ambiguity that surrounds the concept of planning in Australia; planning is often paid a kind of lip-service that undermines its effectiveness, so that formally "planned'' communities are frequently less well matched with their environments than those that grew fortuituously. These and other contrasts throw up challenging questions about Sydney's prospects for a sustainable future. How, for example, can organizations whose purpose is to protect lives and property against immediate environmental hazards contribute to the achievement of urban sustainability when their actions usually have the effect of buttressing people and encouraging behaviours that are undermining longer-term environmental stability? Handmer's assessment is that without institutional and attitudinal reforms Sydney faces an uncertain future. Today, different types of acute hazard are generally well managed by narrowly specialized agencies. However, neither these bodies nor any foreseeable alternatives seem capable of dealing with the growing number of chronic slow-developing hazards that are emerging to threaten this mega-city.

London (chap. 7) has been a major city for almost 1,000 years and a mega-city for almost 200. During its history different types of disaster have severely damaged the city including, among others, disease epidemics, fires, floods, windstorms, and aerial bombardment. But, as revealed by Dennis Parker, the lessons of these experiences have frequently not been incorporated into improved hazard-management policies and programmes. Hazards are typically - and erroneously - viewed as problems that are separable from their social, economic, and political contexts. Instead of comprehensive responses that target the linkages among different contributory factors, the dominant approaches are usually incomplete, piecemeal, and poorly coordinated. In the coming decades, increased exposure to risk and differential vulnerability among the city's multiple interest groups are likely to be the driving forces in the hazards adjustment equation.

Chapter 8 switches the focus to Latin American mega-cities, as represented by the capital of Peru. Lima's history and hazard profile are laid out by Anthony Oliver-Smith. For almost five centuries, the city's capacity to rebound from repeated earthquakes, economic shocks, and other hazards has been seriously hampered by Peru's lack of control over the international market and political forces that permitted foreign investors to exploit its resources. During this period Lima presided over incoming flows of investment capital and metropolitan cultural influences and outgoing flows of minerals and other exports. Though repeatedly damaged by earthquakes and other extreme phenomena, the city was buffered against the worst hazards because it remained small and because it housed the most powerful Peruvian institutions and leaders who could command available resources in time of need. Unfortunately, Lima's capacity to provide for the safety of its residents has been swamped in recent decades by an unprecedented wave of poor peasants fleeing landlessness, a precarious rural economy, and the civil and military violence of rural areas in pursuit of uncertain but seductive futures in the burgeoning metropolis. What was once a medium-sized city that kept its élites well protected against most hazards, and also provided the subdominant classes with a modicum of security, has become a sprawling metropolis that increasingly offers a hazardous future to all its residents.

The vast metropolis of Mexico City is the focus of chapter 9. Sergio Puente describes the experience of living there as "chaos that is permanently on the verge of catastrophe'' and points to continuing problems of seismic instability, flooding, fire, poor air quality, industrial explosions, and a variety of other hazards. Mexico City is viewed as the product of interactions among three groups of stakeholder: real estate developers, the state, and poverty-stricken citizens. In recent years the weakening of the state has contributed greatly to changing patterns of urban vulnerability, as evidenced by the emergence of citizens' action organizations that largely took over the task of meeting public demands for services in the wake of a catastrophic 1985 earthquake. Yet the state is one of the few third world institutions that is capable of acting against market-driven forces that have often played a large role in creating conditions of vulnerability in mega-cities. Puente indicts rapid population growth and deep social polarization as the two overriding factors that hamper state-backed attempts to address urban environmental hazards in third world cities, but he argues that the vulnerability of such places also varies widely in proportion to the weakness of their infrastructures, their degree of dependence on rural hinterlands, their positions in national urban hierarchies, and their peripherality in the global economic system. He then proposes and demonstrates a methodology for analysing vulnerability to environmental hazards in Mexico City. This employs a matrix that permits different factors of vulnerability in different urban neighbourhoods to be scored. The use of vulnerability matrices for analysing urban hazards is widely accepted among professionals, but Puente's version is more sensitive to the social dimensions of Mexico City's vulnerability than others that have preceded it. Maps of composite vulnerability based on matrix scores show that only about one-third of the mega-city is actually vulnerable to hazards, compared with over one-half as suggested by conventional analyses that rely solely on natural and physical factors.

Three chapters on North American mega-cities (San Francisco, Los Angeles, and Miami) round off the case-studies. All three cities have recently experienced major events that inflicted record losses - but none was the much-feared "Big One'' of the respective communities. In other words, these are places where worse is not only possible; it is likely! Conversely, all three cities are also associated with (carefully cultivated) images of earthly paradise; they hold out the prospect that ambitious dreams of material success can be realized in an exotic location by the bold, the resourceful, or the fortunate. The tensions between these contradictions are expressed differently in each of the cities and all face daunting challenges that invite inspired responses.

In chapter 10, Rutherford Platt focuses on earthquake and wildfire problems of the San Francisco Bay Area. High levels of disaster planning and management expertise already exist among many Bay Area institutions, although coordination among different levels of government, different jurisdictions, and different service areas is often very difficult. Mutual aid agreements and other arrangements have improved the area's capacity to cope with disaster by means of greater system interconnectivity, but these adjustments are being hard pressed by changes in the broad context of hazard: increasing risks of wildfires on the urban periphery; a disturbingly long interval since the last great earthquake; the probability of renewed seismic activity along the heavily populated and poorly protected Hayward Fault; and the emergence of San Francisco as one pivot of a disaster-sensitive global economic system. Reduction of infrastructural vulnerability is perhaps the most pressing problem of the Bay Area, especially the ease with which vital lifelines can be disrupted for long periods. This is a problem that affects many other mega-cities, particularly those built near lakes, major rivers, or other physical impediments to the movement of people and resources. In light of the social and physical changes that are occurring in greater San Francisco, there are no guarantees of permanent success for the varied approaches to hazard management that have been tried there. None the less, the Bay Area's record of constructive engagement with environmental hazards has made this mega-city an important teaching laboratory for the international community of hazards scholars and managers.

Platt's guarded assessment of Bay Area hazards is followed by a vivid case-study of another metropolis that accentuates the deterioration of prospects for successful hazard management (chap. 11). This focuses on the "improbable city'' of Greater Los Angeles. While acknowledging the skills of researchers, managers, and other professionals who have diligently pursued "top - down'' adjustments to natural and technological hazards, Ben Wisner offers the opinion that the best efforts of such groups have been steadily undercut by a combination of post - Cold War economic shocks and widening ethnic and class divisions between rich and poor Angelenos. These have the effect of sorting and segmenting populations at risk, so that some hazards are characteristically associated with certain groups and the aggregate burden of hazard falls disproportionately on the urban poor. Alternative grass-roots (bottom - up) initiatives are recommended and examples are provided of inner-city community action groups that have incorporated hazard reduction within their agendas. From Wisner's perspective, a collaborative strategy that unites top - down and bottom - up components is called for. Though the twin forces of global economic restructuring and rampant individualism tend to work against such broad-based metropolitan policy-making, Los Angeles may - as so often in the past - once again be an exception to conventional explanations of urban development.

The final case-study examines a North American mega-city that is still recovering from unprecedented losses inflicted by a catastrophic storm (chap. 12). Greater Miami experienced relatively few deaths as a result of hurricane Andrew (1992), but property damage dwarfed the record of previous hurricanes anywhere in the world. William Solecki analyses the role of natural hazards in the development of Miami and assesses Andrew's impact on regional social ecology, land-use patterns, and demographic trends. He points out that the stereotypical image of Greater Miami as a series of more or less tranquil beach resorts is at variance with the place's dynamic and hazardous setting. Ambitious drainage and water-management projects, which expanded the originally small area available for building, have their roots in efforts to avoid a repetition of damage inflicted by a series of hurricanes that swept through the area in the 1940s. Solecki notes that this mega-city was ripe for disaster when Andrew struck, both because its vastly increased population lacked recent experience of major storms, and because the community had suffered serious economic reverses combined with the fragmentation of its political structure into fractious groups that lacked a perception of common interests. In agreement with impact studies of similar events elsewhere, Solecki reports that the disaster accelerated existing development trends but did not initiate any fundamentally new ones. However, the vulnerability gap between the poorest residents and others widened in the wake of the storm. As this study's analysis of post-disaster interest group coalitions makes clear, because of their volatility, political arrangements in large cities of the developed world pose major difficulties for the construction of long-term disaster recovery and mitigation programmes, but they also open up opportunities for creative approaches to these subjects. The challenge for scholars of hazard will be to find ways of exploiting these openings.

In chapter 13, I draw together findings and examine a series of new departures that they stimulate for urban hazards researchers and policy makers. I find that there is a strong case for ranking natural hazards and disasters among the most seriously underestimated problems of urban management and suggest that, without changes in the outlook of urban managers, they will only become worse. In recent decades, natural hazards and social hazards have been more troubling in large cities than have technological hazards and biological ones. Certain kinds of potentially valuable adjustment to hazard have been neglected in mega-cities. These include: non-expert systems, informal procedures, non-structural technologies, and a wide range of private sector initiatives suitable for families, neighbourhood groups, and small firms. Shifts in patterns of exposure and vulnerability are among the most striking changes in the parameters of hazard during recent years. A wide range of spatial indicators reflects both the dynamics of hazard within cities and the operation of causal linkages between hazard at the mega-city level and sociopolitical and economic processes at other scales of analysis. This chapter concludes by identifying new and potentially fruitful research areas that are indicated by the case-studies. Among these are the roles of hazards as agents of urban diversification, as catalysts of contingency in decision-making, and as sources of metaphor for urban policy-making. High priority should be accorded to the mapping of urban hazard ecologies, to exploring the influence of indirect signals of hazard in the urban environment, and to charting the role of hazards in conflicts among the major interest groups that shape contemporary urban policy.

The book ends with an extended postscript prompted by recent proceedings of the Second United Nations Conference on Human Settlements (Habitat II). Despite preparatory meetings that identified environmental hazards as an important urban problem, this bell-wether conference paid scant attention to them. Failure to recognize natural hazards as a worsening urban problem suggests a peculiarly myopic view of urban management and signals flaws in the conceptualization of sustainable development as a principle of urban management. It is to be hoped that renewed efforts will be made to correct these deficiencies in the near future.

REFERENCES

Brunn, Stanley D. and Jack F. Williams. 1993. Cities of the world: World regional development, 2nd edn. New York: HarperCollins.

Buendia, Hernando Gomez, ed. 1990. Urban crime: Global trends and policies. Tokyo: United Nations University Press.

Fuchs, Roland J., Ellen Brennan, Joseph Chamie, Fu-chen Lo, and Juha I. Uitto, eds. 1994. Mega-city growth and the future. Tokyo: United Nations University Press.

Gilbert, Alan, ed. 1996. The mega-city in Latin America. Tokyo: United Nations University Press.

Ichikawa, Atsushi. 1995. "Coping with urban disasters.'' OECD Observer 197 (December 1995 - January 1996): 15 - 16.

International Decade for Natural Disaster Reduction. 1996. Cities at risk: Making cities safer - before disaster strikes. Supplement to No. 28, Stop Disasters, Geneva: IDNDR.

Lo, Fu-chen and Yue-man Yeung, eds. 1996. Emerging world cities in Pacific Asia. Tokyo: United Nations University Press.

Main, Hamish and Stephen Wyn Williams, eds. 1994. Environment and housing in third world cities. Chichester and New York: John Wiley.

Mitchell, James K. 1995. "Coping with natural hazards and disasters in mega-cities: Perspectives on the twenty-first century.'' GeoJournal 37(3): 303 - 312.

Parker, Dennis and James K. Mitchell. 1995. "Disaster vulnerability of mega-cities: An expanding problem that requires rethinking and innovative responses.'' GeoJournal 37(3): 295 - 302.

Rakodi, Carole, ed. 1996. The urban challenge in Africa: Growth and management of its large cities. Tokyo: United Nations University Press.

United Nations Commission on Human Settlements. 1996. An urbanizing world: Global report on human settlements 1996. New York: Oxford University Press.

World Resources Institute. 1996. World Resources 1996 - 97. Special issue on "The Urban Environment''. Washington, D.C.: World Resources Institute, United Nations Environment Programme, United Nations Development Programme, World Bank.