The Dynamics of Risk: Changing Technologies and Collective Action in Seismic Events

By Louise K. Comfort

Earthquakes are a huge global threat. In thirty-six countries, severe seismic risks threaten populations and their increasingly interdependent systems of transportation, communication, energy, and finance. In this important book, Louise Comfort provides an unprecedented examination of how twelve communities in nine countries responded to destructive earthquakes between 1999 and 2015. And many of the book’s lessons can also be applied to other large-scale risks.

The Dynamics of Risk sets the global problem of seismic risk in the framework of complex adaptive systems to explore how the consequences of such events ripple across jurisdictions, communities, and organizations in complex societies, triggering unexpected alliances but also exposing social, economic, and legal gaps. The book assesses how the networks of organizations involved in response and recovery adapted and acted collectively after the twelve earthquakes it examines. It describes how advances in information technology enabled some communities to anticipate seismic risk better and to manage response and recovery operations more effectively, decreasing losses. Finally, the book shows why investing substantively in global information infrastructure would create shared awareness of seismic risk and make postdisaster relief more effective and less expensive.

The result is a landmark study of how to improve the way we prepare for and respond to earthquakes and other disasters in our ever-more-complex world.

• Language: English
• Publisher: Princeton University Press
• Release date: Jun 25, 2019
• ISBN: 9780691186023

Book preview

THE DYNAMICS OF RISK

Princeton Studies in Complexity

Series Editors

Simon A. Levin (Princeton University)

Steven H. Strogatz (Cornell University)

Robert Axelrod, The Complexity of Cooperation: Agent-Based Models of Competition and Collaboration

Lars-Erik Cederman, Emergent Actors in World Politics: How States and Nations Develop and Dissolve

Peter S. Albin, Barriers and Bounds to Rationality: Essays on Economic Complexity and Dynamics in Interactive Systems

Scott Camazine, Jean-Louis Deneubourg, Nigel R. Franks, James Sneyd, Guy Theraulaz, and Eric Bonabeau, Self-Organization in Biological Systems

Sorin Solomon, Microscopic Representation of Complex Systems

Peter Turchin, Historical Dynamics: Why States Rise and Fall

Andreas Wagner, Robustness and Evolvability in Living Systems

J. Stephen Lansing, Perfect Order: Recognizing Complexity in Bali

Mark Newman, Albert-László Barabási, and Duncan J. Watts, The Structure and Dynamics of Networks

Joshua M. Epstein, Generative Social Science: Studies in Agent-Based Computational Modeling

John H. Miller and Scott E. Page, Complex Adaptive Systems: An Introduction to Computational Models of Social Life

Marten Scheffer, Critical Transitions in Nature and Society

Michael Laver and Ernest Sergenti, Party Competition: An Agent-Based Model

Joshua M. Epstein, Agent_Zero: Toward Neurocognitive Foundations for Generative Social Science

Louise K. Comfort, The Dynamics of Risk: Changing Technologies and Collective Action in Seismic Events

The Dynamics

of Risk

Changing

Technologies and

Collective Action

in Seismic Events

Louise K. Comfort

PRINCETON UNIVERSITY PRESS

PRINCETON AND OXFORD

Copyright © 2019 by Princeton University Press

Published by Princeton University Press

41 William Street, Princeton, New Jersey 08540

6 Oxford Street, Woodstock, Oxfordshire OX20 1TR

press.princeton.edu

All Rights Reserved

LCCN 2019934398

ISBN 978-0-691-16536-3

ISBN (pbk.) 978-0-691-16537-0

eISBN 978-0-691-18602-3 (ebook)

Version 1.0

British Library Cataloging-in-Publication Data is available

Editorial: Eric Crahan and Pamela Weidman

Production Editorial: Kathleen Cioffi

Cover Design: Pamela L. Schnitter

Cover Credit: Nigel Spiers / Alamy

Production: Erin Suydam

Publicity: Tayler Lord, Nathalie Levine, and Julia Hall

To the memory of my mother,

Valborg Oline Fjøslien Kloos,

whose love of learning, humane values, and curiosity

about the world profoundly shaped this inquiry

CONTENTS

List of Figures  ix

List of Tables  xi

Preface  xv

1   Redefining Risk on a Global Scale  1

2   Risk in Complex Systems  22

3   Assessing Risk in Complex Systems: Data, Methods, and Measurement  44

4   Risk in Practice  59

5   Toward an Auto-adaptive System: The 2013 Lushan County, China, Earthquake  76

6   Operative Adaptive Systems: 1999 Duzce, Turkey; 2009 Padang, Indonesia; 2011 Tohoku, Japan; and 2015 Nepal Response and Recovery Systems  92

7   Emergent Adaptive Systems: 1999 Marmara, Turkey; 1999 Chi Chi, Taiwan; 2005 Pakistan; and 2008 Wenchuan, China, Earthquake Response Systems  134

8   Nonadaptive Systems: 2001 Bhuj, Gujarat, India, Earthquake; 2004 Sumatra, Indonesia, Earthquake/Tsunami; and 2010 Haïti Earthquake Response Systems  176

9   Evolving Patterns of System Response  208

10 The Logic of Resilience  235

Appendix I: Tables of Transactions by Classes of Adaptation  253

Appendix II: Sources of Electronic Data, 2013 Lushan Earthquake  279

Notes  281

References  287

Index  299

FIGURES

1.1  CASoS engineering design process  10

5.1  Map, Lushan County earthquake, April 20, 2013  79

5.2  Rate of change in 2013 Lushan County, China, response system  85

5.3  Top 20 organizations in 2013 Lushan County, China, response system, with icons sized by betweenness values  86

6.1  Maps, 1999 Duzce, Turkey, earthquake; 2009 Padang, Indonesia, earthquake  93

6.2  Maps, Tohoku, Japan, earthquake, 2011; Nepal earthquake, 2015  94

6.3  Rate of change in 1999 Duzce, Turkey, response system  101

6.4  Top 20 organizations in 1999 Duzce, Turkey, response system, with icons sized by betweenness centrality  102

6.5  Rate of change in the 2009 Padang, Indonesia, response and recovery system  109

6.6  Top 20 organizations engaged in 2009 Padang response system, ranked by betweenness centrality, with jurisdiction and funding sector  110

6.7  Rate of change in 2011 Tohoku, Japan, response and recovery system  118

6.8  Top 20 organizations in 2011 Tohoku, Japan, response and recovery system, ranked by betweenness centrality  119

6.9  Rate of change in 2015 Nepal response system, April 25–May 16, 2015  127

6.10  Network diagram of top 20 organizations, 2015 Nepal response system, ranked by betweenness centrality  128

7.1  Maps, Marmara, Turkey, earthquake, August 17, 1999; Chi Chi, Taiwan, September 21, 1999  136

7.2  Maps, Pakistan earthquake, 2005; Wenchuan, China, earthquake, 2008  137

7.3  Rate of change in 1999 Marmara, Turkey, response system, August 18–September 7, 1999  144

7.4  Top 20 organizations in 1999 Marmara response system, ordered by betweenness centrality  146

7.5  Rate of change in 1999 Chi Chi, Taiwan, response system September 21–October 9, 1999  152

7.6  Top 20 organizations, 1999 Chi Chi, Taiwan, response system, ranked by betweenness centrality, with jurisdiction and funding sector  153

7.7  Rate of change in the 2005 Pakistan response system, October 9–29, 2005  160

7.8  Top 20 organizations in 2005 Pakistan earthquake response system, ranked by betweenness centrality, with jurisdiction and funding sector  161

7.9  Rate of change in 2008 Wenchuan response system, May 12–June 1, 2008  169

7.10  Top 20 organizations participating in 2008 Wenchuan, China, response system, icons sized by betweenness centrality, showing jurisdiction and funding sector  171

8.1  Maps, Bhuj, Gujarat, India, earthquake, 2001; Sumatra, Indonesia, earthquake 2004  180

8.2  Map, Haïti earthquake, 2010  181

8.3  Rate of change in 2001 Gujarat response system, January 26, 2001–February 15, 2001  184

8.4  Network diagram of top 20 organizations, Gujarat response system, ranked by betweenness centrality  186

8.5  Rate of change in 2004 Sumatra, Indonesia, response system, December 26, 2004–January 16, 2005  194

8.6  Network diagram of top 20 organizations in Sumatra, Indonesia, system, ranked by betweenness centrality  195

8.7  Rate of change in 2010 Haïti response system, January 12–February 3, 2010  202

8.8  Network diagram of top 20 organizations in 2010 Haïti response system, ranked by betweenness centrality  203

TABLES

4.1   Assessment Indicators for Earthquake Response Systems  64

4.2   Preliminary Classification of 12 Earthquake Response Systems Based on Technical, Organizational, and Cultural Dimensions, 1999–2015  67

4.3   Classes of Sociotechnical Adaptation by Earthquake Response Systems, 1999–2015  68

4.4   Preliminary Characteristics of Response System Tending toward Auto-adaptation in Lushan, China, 2013  69

4.5   Preliminary Characteristics of Operative Adaptive Systems  71

4.6   Preliminary Characteristics of Emergent Adaptive Systems  72

4.7   Preliminary Characteristics of Nonadaptive Systems  73

5.1   Frequency Distribution of Organizations Participating in the 2013 Lushan County, China, Response System, by Jurisdiction and Funding Sector  83

5.2   Top 20 Organizations Participating in the 2013 Lushan County, China, Response System, Ranked by Betweenness Centrality, with Jurisdiction, Funding Source, and Degree Centrality  87

6.1   Frequency Distribution of Organizations Participating in the 1999 Duzce, Turkey, Response System, by Jurisdiction and Funding Sector  100

6.2   Top 20 Organizations Participating in 1999 Duzce, Turkey, Response System, Ranked by Betweenness Centrality, with Jurisdiction, Funding Source, and Degree Centrality  103

6.3   Frequency Distribution of Organizations Engaged in 2009 Padang Earthquake Response System by Jurisdiction, Funding Sector  108

6.4   Top 20 Organizations Participating in 2009 Padang, Indonesia, Response System, Ranked by Betweenness Centrality, with Jurisdiction, Funding Source, and Degree Centrality  111

6.5   Frequency Distribution of Organizations Participating in the 2011 Tohoku Earthquake, Tsunami, and Nuclear Reactor Breach Response System, by Jurisdiction and Funding Sector  117

6.6   Top 20 Organizations in 2011 Tohoku, Japan, Response and Recovery System, Ranked by Betweenness Centrality, with Jurisdiction, Funding Source, and Degree Centrality  120

6.7   Frequency Distribution of Organizational Response System, April–May 2015, Nepal Earthquakes, by Jurisdiction and Funding Sector  126

6.8   Top 20 Organizations in 2015 Nepal Earthquakes Response System, Ranked by Betweenness Centrality, with Jurisdiction, Funding Source, and Degree Centrality  129

7.1   Frequency Distribution of Organizations in the 1999 Marmara, Turkey, Response System, by Jurisdiction and Funding Sector  142

7.2   Top 20 Organizations in 1999 Marmara, Turkey, Response System, Ranked by Betweenness Centrality, with Jurisdiction, Funding Source, and Degree Centrality  147

7.3   Frequency Distribution of Organizations in the 1999 Chi Chi, Taiwan, Response System, by Jurisdiction and Funding Sector  151

7.4   Top 20 Organizations, 1999 Chi Chi, Taiwan, Earthquake Response System, Ranked by Betweenness Centrality, with Jurisdiction, Funding, and Degree Centrality  154

7.5   Frequency Distribution of Organizations in 2005 Pakistan Response System, by Jurisdiction and Funding Sector  159

7.6   Top 20 Organizations Participating in 2005 Pakistan Response System, Ranked by Betweenness Centrality, with Funding Source, Jurisdiction, and Degree Centrality  162

7.7   Frequency Distribution of Organizations Engaged in Response Operations, 2008 Wenchuan, China, Earthquake, by Jurisdiction and Funding Sector  168

7.8   Top 20 Organizations in 2008 Wenchuan Earthquake Response System, Ranked by Betweenness Centrality, with Funding Source, Jurisdiction, and Degree Centrality  172

8.1   Frequency Distribution of Organizations Engaged in Disaster Operations by Jurisdiction and Funding Sector, 2001 Bhuj, Gujarat, Earthquake Response System  183

8.2   Top 20 Organizations in Gujarat Response System, Ranked by Betweenness Centrality and Reporting Funding Sector, Jurisdiction, and Node Degree Centrality  187

8.3   Frequency Distribution of Organizations Engaged in the 2004 Sumatra, Indonesia, Response System, by Jurisdiction and Funding Sector  193

8.4   Top 20 Organizations in 2004 Sumatra, Indonesia, Response System, Ranked by Betweenness Centrality, with Jurisdiction, Funding Sector, and Degree Centrality  196

8.5   Frequency Distribution of Organizations Participating in 2010 Haïti Response Systems, by Jurisdiction and Funding Sector  201

8.6   Top 20 Organizations Participating in 2010 Haïti Response System, Ranked by Betweenness Centrality, with Jurisdiction, Funding Sector, and Degree Centrality  204

9.1   External/Internal Index for Earthquake Response System Moving toward Auto-adaptation  215

9.2   Comparison of Observed E/I Index Values with Permuted E/I Index Values for Lushan Earthquake Response System Moving toward Auto-adaptation  216

9.3   External/Internal Indexes for Operative Adaptive Earthquake Response Systems  218

9.4   Comparison of Observed E/I Values with Permuted E/I Values for Operative Adaptive Systems, Significance (P) Values  221

9.5   External/Internal Indexes for Emergent Adaptive Earthquake Response Systems  224

9.6   Comparison of Observed E/I Values with Permuted E/I Values for Emergent Adaptive Systems, Significance (P) Values  227

9.7   External/Internal (E/I) Indexes for Nonadaptive Systems  229

9.8   Comparison of Observed E/I Values with Permuted E/I Values for Nonadaptive Systems, Significance (P) Values  232

10.1  Comparison of Overall E/I Values for Network Moving toward an Auto-adaptive System  238

10.2  Comparison of Overall E/I values for Whole Networks, Operative Adaptive Systems  239

10.3  Comparison of Overall E/I Values for Whole Networks, Emergent Adaptive Systems  239

10.4  Comparison of Overall E/I Values for the Whole Network, Nonadaptive Systems  240

I.5.1 Types of Transactions Reported in 2013 Lushan County, China, Response System, by Jurisdiction and Funding Sector  254

I.6.1  Types of Transactions Reported in 1999 Duzce, Turkey, Response System, by Jurisdiction and Funding Sector  256

I.6.2  Types of Transactions Reported in 2009 Padang, Indonesia, Response System, by Jurisdiction and Funding Sector  258

I.6.3  Types of Transactions Reported in 2011 Tohoku, Japan, Response and Recovery System, by Jurisdiction and Funding Sector  260

I.6.4  Types of Transactions Reported in 2015 Nepal Response and Recovery System, by Jurisdiction and Funding Sector  262

I.7.1  Types of Transactions Reported in August 17, 1999, Marmara, Turkey, Response System, by Jurisdiction and Funding Sector  264

I.7.2  Types of Transactions Reported in 1999 Chi Chi, Taiwan, Response System, by Jurisdiction and Funding Sector  266

I.7.3  Types of Transactions Reported in 2005 Pakistan Earthquake Response System, by Jurisdiction and Funding Sector  268

I.7.4  Types of Transactions Reported in 2008 Wenchuan, China, Response System, by Jurisdiction and Funding Sector  270

I.8.1  Types of Transactions Reported in 2001 Bhuj, Gujarat, India, Response System, by Jurisdiction and Funding Sector  272

I.8.2  Types of Transactions Reported in 2004 Sumatra, Indonesia, Response System, by Jurisdiction and Funding Sector  274

I.8.3  Types of Transactions Reported in 2010 Haïti Earthquake, by Jurisdiction and Funding Sector  276

II.1  Sources of Electronic Data for Network Analysis, Lushan Earthquake, April 20, 2013  279

PREFACE

Like the San Andreas fault, which runs more than 800 miles along the coast of California, seismic risk is hidden from view, and largely from conscious thought of the nearly 40 million residents of the state. The risk is ever present, although the actual ruptures occur every 40 to 60 years for moderate earthquakes, 90 to 150 years for major events. Through these quiet periods, communities develop, grow, thrive, and expand, putting more people, buildings, infrastructure, organizations, and institutions at risk, should a major earthquake occur. The challenge is how to maintain sufficient awareness of risk to inform the design, construction, and maintenance of communities despite the knowledge that a catastrophic event could shatter the interdependent web of social, economic, and engineered systems that support society.

Living with risk requires a framework for managing contingencies and imagining alternative approaches to seemingly permanent structures. In a search to understand how decision makers frame actions to cope with uncertain events, especially large-scale, potentially destructive events that have regional, national, and international consequences, I undertook this study nearly twenty years ago. Examining community response to actual earthquake events as they happened, I sought to understand the characteristics of decision making by people at different levels of responsibility, capacity, and insight into this complex, dynamic problem.

Over the years, many people have assisted me in this journey: practicing emergency managers, scholars, students, and community residents in twelve different countries in local, national, and international organizations. Their names are too numerous to mention, but in each country, several persons were extraordinarily thoughtful and supportive of this effort. In Turkey, Rușen Keleș, Polat Gulkan, Husein Guler, and Yeșim Sungu guided me through the Turkish laws, policies, and context of seismic risk in the country. Suleyman Celik and Sitki Corbacioglu were thoughtful interpreters of the changing response to disaster as earthquakes continued to threaten Turkey in repeated seismic events of 1999 and later. In Taiwan, Jay Shih, Chung Yuang Jan, Kai Hong Fang, and Wen Jiun Wang provided thoughtful guidance, insight, and importantly, translation in interviewing local personnel and analyzing local news reports following the 1999 Chi Chi earthquake. In Gujarat, India, I am grateful to Haresh Shah for organizing an unusual reconnaissance study that included senior and junior researchers in the social sciences, engineering, and geography. I am especially indebted to V. Thiruppugazh, deputy director of the Gujarat State Disaster Management Authority, Gandhinagar, in 2001, for his thoughtful insight into the uses of information technology that India was beginning to adopt for disaster management. In Indonesia, I am indebted to my longtime colleagues and friends, Harkunti Rahayu, Bandung Institute of Technology, and Febrin Ismail, Andalas University, Padang, for continuing collaboration since the 2004 Sumatran earthquake and tsunami, through the 2009 Padang earthquake and continuing efforts to find credible methods of informing residents of Indonesia’s coastal cities regarding methods of self-organization and resilience to confront seismic risk.

In Pakistan, several practicing managers made significant contributions to this study, among them Abrar Ismael, Zafar Shah, and Dr. Mohammad Daud Khan, CHEF International, as well as Nauman Afridi, who assisted me with coding news reports from the Pakistani newspaper Dawn, Islamabad, regarding the Pakistan earthquake in 2005. I thank Brent Woodworth for his field observations in both Gujarat and Pakistan, which enriched my understanding of the context of disaster operations in both countries. In China, Haibo Zhang, Nanjing University, facilitated my first visit to the Wenchuan region in 2008. Since that time we have worked together as colleagues and friends, seeking to understand the changing disaster management system as successive earthquakes disrupted communities in China. Xing Tong, distinguished professor at Nanjing University, facilitated the scholarly exchange in the study of disaster management between Nanjing University and the University of Pittsburgh, and Hongyun Zhou, Zhongnan University of Economics and Law, Wuhan, graciously provided updates from local newspapers on the Lushan earthquake, 2013.

In Haïti, my thanks go to Jacques Gabriel, then minister of public works in Haïti as well as dean, School of Engineering, L’Université d’Etat d’Haïti; Even-son Calixte, dean of science, engineering, and architecture, Quisqueya University; and Dr. Yolène V. Surena, faculty of medicine, UEH. Importantly, I thank my former students Leonard Huggins, Ted Serrant, Michael Siciliano, Steve Scheinert, Rebecca Jeudin, Maria Escorcia, Sebastian Gasquet, and Edgar Largaespada, all of whom traveled with me to Haïti on research trips and engaged with Haïtian students at UEH and Quisqueya Universities. In Japan, my warmest thanks and appreciation go to Professor Yasuo Tanaka, Kobe University, for his assistance and support over the years, and especially in facilitating arrangements for a field study following the 2011 Tohoku, Japan, triple disasters. My deep appreciation goes to Aya Okada, former doctoral student, research assistant, and now professor at Kanazawa University, Japan. Dr. Okada served as translator, analyst, and essential guide to Japan following the Tohoku disasters and also participated in the research and analysis for the 2010 Haïti earthquake. In Nepal, warm thanks and sincere appreciation go to James Joshi, my colleague and coinvestigator on the Nepal reconnaissance trip, and especially to Prabin Joshi and the Joshi family for facilitating our study in Nepal and their gracious hospitality on our visits.

Collecting the data for this book was one major task; still another was coding and analyzing the data collected over a period of 16 years, all of which needed to be checked, recoded, and validated, using advanced methods of network analysis in R. For this major task, undertaken largely during the spring and summer of 2017, I was fortunate to have a very talented set of advanced doctoral students at the Center for Disaster Management who assisted me with this painstaking, careful work. My warmest thanks and appreciation go to Jee Eun Song, Seunghyun Lee, Lucy Gillespie, Jay Rickabaugh, Nauman Afridi, and Farhod Yuldashev. Fuli Ai graciously produced the maps for the 12 earthquakes in GIS format, and Mark Dunn assisted with database management in organizing and managing the data sets, greatly facilitating the research process. Karen Cuenco provided a careful expert review of the network analysis findings to validate the results.

Importantly, I owe a major debt of gratitude to Todd R. LaPorte, University of California, Berkeley, and Friedemann Wenzel, University of Karlsruhe, Germany, for reading the entire manuscript, chapter by chapter, and providing thoughtful and welcome comments that helped to sharpen the argument and clarify points that I had overlooked. Warm thanks to my longtime friend and colleague Sidney Verba, who offered encouragement, thoughtful advice, and wise insights into the comparative contexts of decision making in the disaster-stricken communities. My deep appreciation goes to the anonymous reviewers of the manuscript who offered substantive comments that greatly improved the final manuscript. Eric Crahan, as editor at Princeton University Press, provided consistent, firm, but gracious guidance to bring the manuscript successfully to publication. Throughout this process, I thank John T. S. Keeler, dean, Graduate School of Public and International Affairs, University of Pittsburgh, for providing research support through the Center for Disaster Management and a collegial working environment that enabled me to complete this years-long project.

Throughout this period of nearly twenty years, my family has provided unwavering support, perceptive questions, patience, and much needed laughter to guide me through times that were heartbreaking, challenging, sad, but ultimately hopeful in dealing with wrenchingly painful accounts of human suffering, courage, and determination. To my son, Nathaniel, my daughter, Honore, and their families, I give my deepest love and appreciation. This book would not have been possible without their steadfast encouragement, lively questions, occasional skepticism, and, most of all, willingness to listen.

Louise Comfort

Oakland, California

October 15, 2018

THE DYNAMICS OF RISK

1

Redefining Risk on a Global Scale

Seismic Risk as a Global Policy Problem

Nepal will rise again, T-shirts worn by young Nepalis proudly proclaimed, after the Dharahara Tower in the center of Kathmandu, Nepal, collapsed in the April 25, 2015, earthquake. The city reeled from the severe shock late Saturday morning. Search and rescue teams in hard hats and field gear scoured buildings, checking for persons trapped inside. Families huddled around open cooking fires, fearful of returning to houses still trembling from aftershocks. Groups of young volunteers opened their laptops to create lists of neighbors who were safe and those who were missing. Neighbors looked after neighbors, sharing what food they had and comforting one another in collective grief. As the community began to count its losses and build a web of support for one another, questions regarding human capacity to assess and reduce risk reverberated around the world.

Earthquakes are a known risk in Nepal, but dust from the shattered Dharahara Tower belied the ability of the city’s decision makers to curb the sudden, destructive force of the earthquake and its massive disruption of the region’s daily operations. The historic tower, built in 1832 and designated as a UNESCO World Heritage site, was the symbol of the city, connecting its history over two centuries to a modernizing Nepal that engaged in significant programs of disaster risk assessment and reduction, supported by international agencies and humanitarian assistance (NSET 2015). At risk was not only the structure of the tower, but the design of policies and programs to maintain a complex net of public safety actions stretching across local, district, national, and international agencies against movements of the earth beyond human control.

Seismic risk presents an even greater challenge in developed countries, not only for the nations in which they occur, but also for disruption of the interconnected global trade, transportation, communications, and financial networks that sustain the world’s growing populations. Twelve of the 15 most costly natural disasters in the years 1985–2017 have been due to seismic events.¹ Collectively, these events have resulted in estimated losses of 732,102 lives, dead or missing, and over $1 trillion in direct economic costs, with almost certainly larger indirect costs due to relocation, anguish, and pain. While two of the largest seismic events over the last 30 years have occurred in Japan, a major earthquake in San Francisco or Los Angeles would generate catastrophic social and economic costs, not just for the United States, but for the global society (USGS 2016a).²

Although this book focuses on seismic events, the findings apply broadly to other natural hazards—floods, hurricanes, wildfires, and rising sea levels—as well as human-made disasters—nuclear meltdowns, large-scale terrorist attacks, hazardous materials releases. The technical systems that support the functions of an expanding global society are dependent for design, construction, operations, and maintenance on governance systems that are straining under demands from increasing populations, changing climate systems, and vulnerable social groups.

Managing risk in changing environments is a classic challenge for decision makers in public, private, and nonprofit organizations. Twenty years ago, in Shared Risk: Complex Systems in Seismic Response (1999), I examined disaster response systems following 11 earthquakes in nine different countries from 1985 to 1995. In that study, I proposed an interacting set of indicators for assessing response operations across organizations, sectors, and countries. The analysis demonstrated how insights from complex adaptive systems could improve response operations to seismic events. In the intervening two decades, mounting evidence has reinforced the core concept that catastrophic risks represent the intersection of increasingly interconnected systems and require a complexity framework for analysis and understanding. The rapid change in wireless and electronic technologies has transformed digital operating systems in travel, finance, communications, education, commerce, and in electrical power, water, and gas distribution systems, to mention only obvious areas of societal performance. When a major disruptive event such as an earthquake occurs, the consequences cascade through interdependent systems, escalating the impact across organizations, sectors, and even national boundaries with unimagined costs. For example, the Mw = 9.0 earthquake on March 11, 2011, in Tohoku, Japan, triggered a massive tsunami that flooded the generators of the Fukushima Daiichi nuclear reactor, causing radioactive contamination, electrical power shortages, and economic disruption that affected all 43 prefectures in Japan, with an estimated cost of 200% GDP for Japan, the third-largest economy in the world.

In this study, I build on the initial analysis presented in Shared Risk but offer a deeper, enriched refinement of the earlier concepts of complex adaptive systems, examining an additional 12 response systems following earthquakes from 1999 to 2015. Further, I use methods of network analysis and process tracing to assess the interactions among actors, documenting the degrees of interconnectedness and adaptation observed in the response systems. This set of field studies extends the initial set of 11 earthquakes in 9 countries to a total of 23 earthquake response systems in 14 countries over a 30-year period as the basis for informed judgment on treating seismic response operations as complex adaptive systems.

The challenge of mitigating risk deepens as decision makers discover the interconnectedness of the systems they manage and the unexpected consequences of decisions made to reduce risk in one set of circumstances that actually increase risk for different groups of population under different conditions. In specific contexts, risk entails a set of dynamic, interacting conditions involving multiple actors, changing technologies, conflicting assumptions, different scales of operation, and the need for timely, informed, collective action to mobilize a coherent response. Risk is the probability that harm will occur (Beck 1992), and consequently, decisions made to reduce risk include judgments of uncertainty and trade-offs of present versus future costs. These estimates are made daily but cumulatively create a culture of risk that calibrates acceptance against avoidance. Decisions regarding risk are especially challenging when risk is shared, that is, when adverse conditions affect all members of a community indiscriminately (Comfort 1999a). Learning to recognize, assess, and manage risk on a global scale is a fundamental challenge of the 21st century.

Seismic risk exemplifies this global policy problem in unique ways. Seismic risk creates the potential for catastrophic disruption and loss to the world’s population, as 36 nations of the world are exposed to significant risk from earthquakes, landslides, and related hazards of tsunamis and soil instability. Further, risk escalates as communities exposed to seismic hazards increase in size and scale. For example, the megacities of Tokyo, Istanbul, Jakarta, Mumbai, Mexico City, and Los Angeles are each exposed to significant seismic risk. Their respective metropolitan regions now encompass tens of millions of residents and serve as centers of government, commerce, communications, finance, and transportation in their respective nations. The interdependence of these operational networks in a globally connected world means that if any one of these megacities experiences a major earthquake, the losses would not only affect the immediate population, but also seriously disrupt vital transactions on a worldwide scale.

While interdependence among global infrastructure and organizational systems creates cumulative risk for the global metasystem, this same interdependence creates opportunities for the redesign of interconnected operations to reduce expanding risk. Decisions made in one set of operations create the basis for adaptation in other stages and other locations, enhancing innovative capacity and reducing the overall risk to the global system. Capturing the potential to transform risk into opportunity for redesign of functional systems and infrastructure into more resilient forms represents a positive goal of the risk dynamic. For example, recognizing the potential of landslides generated by earthquakes to disrupt the roads and bridges of rural Nepal creates opportunity to redesign and rebuild these vital roads in ways that strengthen the transportation network nationally. The transformation of risk to constructive redesign of operational systems, however, is not trivial.

Prior efforts to analyze and model disaster risk have not resolved the inherent issues of complexity, uncertainty, size, and scale (Wildavsky 1988; Beck 1992; Bruneau et al. 2003; Blaikie et al. 2004; Cutter et al. 2008). Ulrich Beck raised the fundamental issue of the impact of technology on society, and the consequent risks that technology engendered in transforming an industrialized society into a modern one. Left unanswered in his profound analysis is how institutions and organizations could adapt to newly generated risks in constructive ways. Aaron Wildavsky framed the issues of cost versus control in changing environments, acknowledging that the search for safety was essentially a process of selecting risks that one could recognize and had the capacity to address. Other risks were either ignored or delayed, given scarcity of resources and public attention. One relied on resilience, or anticipation of potential threats, often viewed in individual terms, to adapt available resources to meet unexpected threats.

Bruneau and his colleagues at the Multidisciplinary Center for Earthquake Engineering Research (MCEER) at the State University of New York, Buffalo, developed a promising model of resilience that sought to counter risk through the interaction of four basic conditions in any area exposed to risk: robustness, redundancy, resourcefulness, and rapidity (Bruneau et al. 2003). Widely accepted in engineering and considered in social science disciplines, this model has been largely applied in engineering contexts and the design of individual buildings, rather than to the wider context of societal risk. Blaikie and colleagues (2004) document the vulnerability of populations exposed to natural hazards and the extent to which this risk, if not acknowledged, may unwittingly be increased by policies of responsible institutions. Susan Cutter and her colleagues (2008) added the central consideration of geographic location to the recognition and assessment of risk, acknowledging the interaction between the physical characteristics of a specific community and the types of risks to which its residents were exposed. While each of these analytical approaches to characterize and analyze risk contributes a useful perspective to understanding the fundamental problem of managing potential threats to communities, none captures the potential for inquiry to change the means of coping with risk in practice. The five perspectives cited above also share a broad focus on technology as an agent that generates risk in an increasingly complex world but differs in application to individual, organizational, and institutional capacities to reduce risk in changing environments.

An initial characterization of organized social complexity framed by Todd LaPorte (1975) and his colleagues identified the problem of sociotechnical systems and the critical role of information in managing the interaction between technical systems and the organizations that design, implement, and operate them. Importantly, major shifts in information and communication technologies (ICT) create the possibility to redefine decision processes for large-scale, sociotechnical, complex systems of systems. While advances in ICT had been building for several decades, they reached a critical point of mass distribution, access, and use in the mid-1990s. The advent of desktop computers and widespread adoption of cell phone networks fundamentally altered the way in which information was collected, organized, stored, and distributed within and among organizations and institutions. Gone were the manila file folders and green filing cabinets, with weeks of effort required to catalogue records of past performance. Gone, as well, was the ability to hide negative information in obscure reports that were buried in an avalanche of paper. Instead, personal computers made access to information readily available, and communication via internet made sharing information over wide distances and time zones easily possible. These technical changes in managing information and communication within large-scale, sociotechnical systems enabled a different form of near-real-time communication and exchange of information within and between organizations and institutions. Although the technology is neutral, the organizational processes through which information and communication flow may either enhance or inhibit the capacity for information exchange within and among systems of systems.

Earlier analyses of risk that focused on the technical vulnerabilities of systems missed the inseparable connection to vulnerabilities in the organizational systems that operate the technologies. Consequently, large-scale, sociotechnical systems that manage the critical infrastructure and communications exchange on which the global society operates constitute fundamentally interdependent systems of systems, prone to both error and innovation. Such systems operate on multiple scales with networks that span continents and require collaboration across institutional and organizational boundaries that respect different laws, policies, and cultural norms. For example, airline networks that routinely link cities across the globe illustrate both extraordinary capacity in the ability of this interconnected transportation system to enable people to move easily over long distances, but also the fragility of the whole system, if one hub breaks down, delaying or canceling flights in hundreds of other routes simultaneously. Each boundary that is crossed offers an opportunity for further expansion of performance or vulnerability that limits functionality of the system of systems.

This study focuses on the extent to which ICT is used effectively to redesign organizations and decision strategies to manage seismic and other continuing risks in more productive, innovative ways, transforming potential risk into resilience in practice. Seismic risk represents the classic dilemma of low probability/high consequence events that has stymied risk analysts for decades (Petak and Atkisson 1982; Kartez 1984; Kunreuther 2010). Given its global reach and the long-term consequences of managing a potentially severe but intermittent hazard, seismic risk verifies uncertainty as the core issue driving complex adaptive systems.

Shared Risk as Public Risk

Shared risk represents a distinct challenge, as it threatens potential harm to all persons exposed to hazards that affect the whole community, whether they have contributed to the threat or not (Comfort 1999a, 3). For example, individuals exposed to risk from earthquakes, hurricanes, tornadoes, drought, or tsunamis may have separate insurance plans for protection of their lives and property but still incur major loss if the bridges are out, electrical power is cut off, ATMs, gasoline pumps, and cash registers are not