On Risk and Disaster

Leading experts address questions of public and private roles in assessing, managing, and mitigating major risks to public health and safety in light of the devastation caused by Hurricane Katrina.

On Risk and Disaster
Lessons from Hurricane Katrina

Edited by Ronald J. Daniels, Donald F. Kettl, and Howard Kunreuther. Foreword by Amy Gutmann

2006 | 304 pages | Paper $28.95
Public Policy | Economics | Technology and Engineering
View main book page

Table of Contents

Foreword
—Amy Gutmann

Introduction
—Ronald J. Daniels, Donald F. Kettl, and Howard Kunreuther

PART ONE: THE CHALLENGE OF THE GULF
On Their Own in Battered New Orleans
—Peter G. Gosselin
Using Risk and Decision Analysis to Protect New Orleans against Future Hurricanes
—Detlof von Winterfeldt
Planning for a City on the Brink
—Kenneth R. Foster and Robert Giegengack
JARring Actions that Fuel the Floods
—Carolyn Kousky and Richard Zeckhauser

PART TWO: THINKING ABOUT RISK
Behaviorally Realistic Risk Management
—Baruch Fischhoff
Rationales and Instruments for Government Intervention in Natural Disasters
—Michael J. Trebilcock, and Ronald J. Daniels
Social Inequality, Hazards, and Disasters
—Kathleen Tierney
Equity Analysis and Natural Hazards Policy
—Matthew D. Adler

PART THREE: PRIVATE SECTOR STRATEGIES FOR MANAGING RISK
Why We Under-Prepare for Hazards
—Robert J. Meyer
Has the Time Come for Comprehensive Natural Disaster Insurance?
—Howard Kunreuther
Rethinking Disaster Policy After Hurricane Katrina
—Scott E. Harrington
Providing Economic Incentives to Build Disaster-Resistant Structures
—Harvey Ryland

PART FOUR: THE GOVERNMENT'S ROLE IN DISASTER PREPAREDNESS AND RESPONSE
Role of Public Health and Clinical Medicine in Preparing for Disasters
—Brian Strom
Hurricane Katrina as a Bureaucratic Nightmare
—Vicki Bier
The Katrina Breakdown
—Jonathan Walters and Donald F. Kettl

Acknowledgments
Contributors


Excerpt [uncorrected, not for citation]

Foreword
Amy Gutmann

"The problems that exist in the world today cannot be solved by the level of thinking that created them."—Albert Einstein
More than four months have passed since Hurricane Katrina struck. We now know that our affluent country failed both to take adequate precautions against the hurricane's deadly impact and to respond effectively to its devastation of New Orleans and other Gulf Coast areas. We do not yet know what lessons will be learned—or heeded—from one of the greatest catastrophes our country has ever experienced.

As the suffering grew ever more alarming in the hurricane's immediate aftermath, we at the University of Pennsylvania rapidly mobilized our community to help survivors rebuild their shattered lives. We invited 100 displaced students to take their fall classes at Penn. And as generous donations poured into relief agencies, students and staff volunteers traveled to the Gulf coast to support efforts on the ground.

At the same time, we began to confront the problems that contributed to making the devastation of Katrina so troubling not only in its breadth and depth but also in its unbalanced effects across different segments of the population. Katrina's aftermath raised perhaps the most profound and disturbing moral question that our society has yet fully to confront: how willing is the United States to compensate for the increased risks to life and health associated with poverty, race, growing economic inequality, inadequate emergency preparedness, and antiquated urban infrastructures? This overarching question cannot be answered by one person thinking alone or by a single institution acting alone.

Nor can the underlying social, economic, and environmental problems that have been magnified by Katrina be solved by the level of institutional thought and action that created them. A shocking amount of shoddy thought and action—from the denial of scientifically manifest risks to the disavowal of morally apparent official responsibility—will need to be transcended to resolve the problems that have victimized hundreds of thousands of Americans.

We must find a way to equitably distribute risks in order to push our democracy closer to its promise of liberty and justice, not only for the affluent but for all. Rising to this challenge demands that the public and private sectors collaborate to develop effective prevention strategies and coordinated responses to natural disasters, industrial accidents, terrorist attacks, and pandemics.

We at the University of Pennsylvania pride ourselves on being one of the world's major research universities. But we cannot rest on our laurels. Institutional contributions to effective prevention and response are a matter not only of technical expertise but also of moral and social responsibility. Collaborating with others, we have the capacity and expertise to develop a framework within which to address such daunting and urgent challenges. We are committed to embracing our civic responsibilities to help inform public debate and discussion.

For all of these reasons, Penn Provost Ron Daniels took the lead in organizing a symposium on Hurricane Katrina in Washington, D.C. on December 1, 2005. The symposium brought together approximately 250 policymakers, public and private-sector leaders, and scholars from many disciplines to raise the level of thinking on risk management issues and develop more effective strategies that can save lives and speed recovery when disaster strikes.

At the core of these discussions were two recurring questions:


This book represents the first attempts to answer these questions. As this book goes to press, we are already planning a second Katrina conference and town meeting—organized by the Penn Institute on Urban Research—that will build on the analyses and strategies contained in this volume. I am thankful to Provost Daniels and the many eminent scholars and policymakers who are doing their best to bring a higher level of thought and action to bear on the many profound problems of societal response to risk that Hurricane Katrina put into high relief. Our thanks go also to the University of Pennsylvania Press for so quickly bringing the symposium's findings to the attention of a wider public.

* * * * *

Introduction
Ronald J. Daniels, Donald F. Kettl, and Howard Kunreuther

The 9/11 terrorist attacks made it clear to Americans that we are living in a new world of geopolitical risks. Hurricane Katrina has shown the world how vulnerable the United States is to natural disasters. A widely predicted Category 3 hurricane left more than 1,300 people dead and hundreds of thousands displaced in New Orleans and throughout the southern part of four states. It was the most expensive natural disaster in the history of the United States, but its true cost to the region and the nation cannot be measured in dollars alone. It brought a great city to its knees and focused large questions about the nation's ability to prepare for and respond to natural disasters and other large risks.

While much-needed investigations will highlight specific failures at local, state, and national levels, Katrina raises a set of much deeper questions about how we address risk in our society. Could our current policies and infrastructure be as vulnerable in our complex modern world as the levees that were designed to protect New Orleans? What public and private initiatives will be needed to increase security and mitigate losses in an increasingly insecure and inherently risky world? Who—government, market, or the third sector—is and should be responsible for taking these actions? As a catastrophic failure, Katrina presents an opportunity to look at the entire public-private systems for addressing risks, and perhaps build better ones. As Stanford Professor Paul Romer has said, "a crisis is a terrible thing to waste."

Issues of risk and responsibility for natural disasters are of interest not only to researchers and policy makers but also to those affected directly or indirectly by these tragic events. In the aftermath of Hurricane Katrina, disaster preparedness and recovery are now at the forefront of the public policy agenda. On the one hand, the devastation of Katrina raises fundamental questions about the strengths and frailties of the existing institutional, legal, and policy landscape governing the risk management of natural disasters in American society. Does the damage inflicted by Katrina imply that the United States is vulnerable to other types of natural hazards that could impose similar, if not greater, damage on its citizens? If so, what types of reforms are required in order to mitigate the risks faced by individual citizens or communities, particularly where there is evidence that these risks are disproportionately borne by the least advantaged members of society?

On the other hand, the avowed determination of the country's political leadership to rebuild the Gulf region raises vexing questions about how to best use public money. Should governments focus on rebuilding adversely affected communities or on providing targeted assistance to adversely affected individuals? This issue of instrument choice is particularly salient because there is no assurance that individuals who receive targeted subsidies will necessarily decide to remain in those communities devastated by Katrina. Politicians who favor assistance to individuals may face onerous political consequences from local interest groups committed to rebuilding their communities—even at extraordinary public cost.

The commitment to rebuilding communities in the face of devastation is powerful testament to the indomitable American spirit. No one wants to admit defeat in the face of adversity. Most citizens and their elected officials want to take advantage of new opportunities to craft a new sense of community. And national sympathy in the face of large catastrophes often opens up the governmental purse strings wider than is the case in normal politics. No one wants to let a chance like that slip by. This attention is likely to be short-lived unless a set of concrete steps is taken to develop a long-term plan that has a good chance of being implemented.

Katrina has proven to be much more than the kind of catastrophe that has afflicted parts of the nation in past years. It hit with a fierce power, in the area long identified as most vulnerable to hurricanes. It raised tough and fundamental questions for which officials in both the public and private sectors had few good answers. Some citizens had insurance against such risks, but many did not. Much of the region's infrastructure—from telephone and electric lines to gas and water systems—was still not functioning months after the event. Some residents who wanted to rebuild returned to discover that their homes had been leveled, or that they simply could not find them at all.

As the first private insurance dollars and public relief checks started to flow, everyone discovered a new problem. Especially in New Orleans, it was impossible to begin recovering from the last catastrophe without preparing for the next. The levees have remained vulnerable to even modest storms, and everyone wants to know: Will new levees be built? What kind of storm could they withstand? Might we pour tens of billions of dollars into the city only to risk having it wiped out again? Housing is the biggest need, but no one in Mississippi or Louisiana wants to rebuild without thinking hard about how best to protect their investment.

Katrina is thus proving to be a story both about recovering from and managing risk. Private insurers are contemplating how best to structure insurance for natural disasters and whether to provide more comprehensive levels of coverage in one omnibus policy. Insurers also recognize that they have an opportunity to encourage families and businesses to protect themselves by making their house more disaster resistant, but that they cannot do this alone. Other stakeholders from the private sector need to play a role here in complementing government risk mitigation and risk insurance objectives by, for example, enforcing building codes.

With the federal government's commitment to pay large amounts of disaster relief, a puzzle naturally emerges. Is the message from Katrina one that there is little economic incentive in reducing risk and limiting losses in advance of a disaster—what experts in the field call mitigation—because the government will act as the insurer of last resort? If that is the case, there is the possibility that enough people will respond in ways that guarantee that the next event will be even more catastrophic—that victims will not have insured themselves voluntarily against risk, that they will not have done what they could to reduce their exposure and limit their losses, and that taxpayers will incur an even bigger price tag than after Katrina?

This dilemma is nothing new. Students of risk have increasingly come to understand a fundamental problem in our policy toward managing disasters. All such catastrophes are, inevitably local. When money pours in, the benefits—from new homes to new businesses and roads and a refreshed civic life—are local as well. But given the propensity of the federal government to provide ex post assistance to affected citizens, the costs of disaster become nationalized—borne by all taxpayers. Economists have long argued that sensible planning begins by linking the expected benefits of taking action with the costs, recognizing that what actions one takes prior to a disaster should be coupled with the response following the event. It is important to link so call ex ante planning with ex post recovery. However, our policies have increasingly disconnected the expected benefits from the costs of disasters and the advance planning to reduce risks from the financial consequences of the disaster.

The dilemma would be serious enough if it applied just to Katrina, or just to Katrina-like disasters. However, a complex set of forces is conspiring to translate the same problem into many other realms. As our population grows, more of the nation's citizens live in areas prone to natural disasters, from floods and tornadoes to earthquakes and hurricanes and some states have taken action to encourage this development for immediate economic gain (in the form of a larger tax base and other benefits). That means the consequences of such events are exacerbated. We also face significant perils from other exogenous risks, like public health threats, such as avian flu. And we now must confront the threat of terrorism, on many different scales and from many different fronts, which raises additional issues about our national security.

Many of these risks are highly interdependent. Weak links in any supply chain for food and rescue equipment can prevent key materials and personnel from playing a useful role. Katrina raises both short- and long-term issues, both political and normative, of evacuating victims to other parts of the country and the impact that this will have on rebuilding the Gulf coast. The design of a poor levee system can have devastating consequences to the residents in the areas that were viewed as being protected against flooding.

Hurricane Katrina is thus not only important in its own right. It also raises a new class of problems that demands rigorous analysis, prudent planning, and courageous political leadership. Further compounding these challenges is the fact that neither the public nor the private sectors can, on its own, appropriately protect the country's citizens from a range of devastating risks. Robust policy demands well-designed private-public partnerships to address these issues.

A Framework for Learning?

This book builds on a framework that recognizes the importance of risk assessment and risk perception as a basis for developing meaningful hazard management strategies. To reduce the potential losses from future disasters and aid the recovery efforts after a catastrophe occurs requires both private and public sector involvement.

The conceptual framework for investigating issues of risk and responsibility requires inputs from many different disciplines. Engineering, medicine, and the natural sciences provide data on the nature of the risks associated with disasters of different magnitudes and the uncertainties surrounding them (Risk Assessment). Geography, organizational theory, psychology, sociology and other social sciences provide insights on how individuals, groups and organizations perceive the risk and make decisions (Risk Perception and Choice). Economics, insurance, health care, public policy and other disciplines form the basis for alternative disaster management strategies. (Risk and Crisis Management). Political science and law underpin important issues related to public and private responsibilities about how to develop a program that allocates resources before and after a disaster efficiently, while taking into account distributional issues and incentives for appropriate preparedness (Implementing Public-Private Partnerships).

Experts in risk assessment have long argued that policy makers ought to base their decisions on assessments of risk—the chances that specific events might occur and the consequences that would are likely to flow from them. Just how probable is another hurricane of Katrina's force? Or of a nuclear power plant accident, a smallpox attack, a major terrorist explosion, or the spread of a new flu pandemic?

Producing good estimates of the direct impacts of a disaster—such as physical damage, injuries, and loss of lives—is difficult enough. The field of risk assessment, however, has expanded to include a careful look at indirect losses (such as business interruption, political disarray, and personal stress). Moreover, some activities impact on other action, so analysts increasingly need to consider feedback effects when they propose a particular program. For example, new development projects can increase water runoff and, hence, hurricane flood damage. On the other hand, better mapping of areas especially susceptible to hurricane damage can alter development plans.

Better risk assessments, if heeded by public and private decision-makers, can reduce the losses from future catastrophic disasters. For example, improved weather forecasts may lower the number of deaths or injuries by providing early warning, which can help residents evacuate threatened areas before a storm arrives. Better forecasts can also eliminate unnecessary evacuations.

When undertaking risk assessments for a specific disaster, experts are likely to seek more precise information to help define the event. Consider the chance of another hurricane like Katrina occurring in the United States in the next 10 years and its potential consequences. Ideally, experts will want to know the geographic area that would be affected by the hurricane, what individuals (and communities) are at risk and how easily they will be able to evacuate the area if there are warnings far enough in advance. They will also want to know the design of the structures in the hazard-prone area and the resulting direct damage and indirect impacts, such as evacuation and relocation of victims from their homes and business interruption losses.

In the case of Katrina's devastating strike on the Gulf coast, experts had forecast much of what actually occurred. The Department of Homeland Security, for example, had identified such a storm, along with a terrorist attack on New York City and an earthquake in California, as the three biggest threats the nation faced. National Geographic, almost a year before the storm struck, published an eerily accurate story on the damage a big storm could inflict. The Federal Emergency Management Agency's "Hurricane Pam" exercise in 2004 also gave clear warnings about what could happen. Yet despite the advance plans and warnings, government officials were woefully under-prepared.

Risk assessment requires that one not only specify the likelihood and potential consequences of disasters of different magnitudes and intensities but that one undertakes systematic benefit-cost analysis to prescribe a set of actions that should be taken in response to those risks. Ideally, such an analysis would reveal the expected costs and benefits of different public policy strategies. For instance, should government invest in large, public infrastructure projects such as dams and levees, and, if so, how elaborate should those projects be? Alternatively, should the risks of disaster be addressed by private investors in the form of safer building design and construction (perhaps motivated by more stringent building codes and municipal approvals)? Or, when measured against the up-front costs of mitigation, are the expected risks of certain hazards economically unavoidable, and, therefore, best addressed by insurance?

The precise form of insurance also raises complex design issues. Is there a need for publicly provided (or, at least, subsidized) insurance? If so, how can this insurance be supplied in a way that does not subvert economically desirable investments in risk mitigation? Further compounding the challenge is identifying the citizen groups likely to be affected by disaster. For instance, special treatment may have to be given to low income residents who cannot afford to purchase insurance or to undertake protective measures that will reduce losses from future disasters.

It is one thing to assess the risk. It is quite another to estimate how people—whether individual families, private sector executives, or elected officials—perceive them. While traditional risk assessment focuses on losses most frequently measured in money, risk perception incorporates the psychological and emotional factors that have an enormous impact on behavior.

In path-breaking studies begun in the 1970s, psychologists such as Paul Slovic and Baruch Fischhoff began measuring laypersons' concerns about different types of risks. These studies showed that for some technologies (like nuclear power) and activities (like storing radioactive waste), there was a wide disparity between the views of ordinary citizens and experts. This important finding raises important questions about policies for dealing with risks. Elected officials often face conflicting imperatives: citizens demand attention to some risks, owing to fear, dread, or catastrophic potential, while experts feel the risk is not worth considering because of the small chance of its occurrence. On the other hand, experts may focus on other risks for which there is little public concern. And, of course, there is the challenge of how elected officials will mediate between these conflicting imperatives in the context of political cycles.

Compounding the problem is the difficulty individuals have in considering the probabilities when making their decisions. It is not easy to assess low-probability events, and people often disregard them. There is ample evidence that people often do not want to consider data on the likelihood that an event might occur, even when it is available to them. Only after a disaster do most people pay attention to it, and then they overestimate its likelihood. For example, when a single, rare disaster attracts a great deal of media attention, individuals focus on the consequences of the event and behave as if it will happen again in the future. Following the terrorist attacks of September 11, many people refused to fly. They believed the chances of being on hijacked plane were extraordinarily high, even though it could be argued that the probability of such events occurring in the future were extremely low given increased vigilance and added protection by the federal government.

This pattern spills over into natural disasters and has raised especially difficult questions in the aftermath of Katrina. Even people living below sea level in New Orleans did not consider their vulnerability to a major storm and failed to purchase flood insurance or invest in loss reduction measures with respect to their property. When the disaster struck, they quickly demanded federal help.

More generally, demands for protection and insurance often increase only after disasters hit—and long after it is too late. In California, insurers note surges in demand for earthquake insurance after a quake occurs, when the risks became more salient in everyone's minds. When asked whether the probability of a future quake is more likely, the same, or less likely than before the disaster, most people correctly respond by saying "less likely."

The issue of risk perception thus compounds the puzzle of risk assessment. It shapes both the help that individuals expect after a major disaster—and how they shape the way that individuals prepare for the risks that lie ahead. It also raises challenges with respect to how one communicates risk so as to encourage individuals to take steps to reduce the losses from future disasters and to act appropriately after a catastrophe occurs.

The way that individuals perceive risks—and behave in dealing with them—often creates enormous disaster management problems. When crises do occur, and when individuals are not ready to meet them, they naturally demand that public and private institutions rise to the challenge. And that, in turn, poses big issues for these organizations. In developing risk and crisis management strategies that have a chance of being implemented, there is a need to incorporate the data from risk assessment and the factors that have been shown to influence risk perception and choice. A risk management strategy should reflect the likelihood and tangible impacts consequences that are likely to occur following a disaster but also taking into account other dimensions that concern the public such as fear and dread and the types of heuristics that are utilized for making choices.

Consider a person living in a flood prone area who believes that the likelihood of a future disaster is so small that it is not worth worrying about. Then a strategy of providing highly subsidized flood insurance premiums will not induce these people to purchase policies. One may have to require coverage as a condition for a mortgage. The federal government resorted to such a strategy on federally insured mortgages after observing the low sales of highly subsidized flood insurance. However, this policy has not been well enforced, as evidenced by the large number of homeowners in flood-prone areas of New Orleans who did not have insurance against rising water.

Complicating the government's response is the fact, which Katrina made all too clear, that disasters do not respect the boundaries defining governmental or organizational missions and responsibilities. Almost by definition, they fall outside normal routines, and they typically require extraordinarily rapid response to reduce the loss of life and to minimize property damage. Government thus not only needs to shape different policies to deal with the mismatch of risks and perceptions. It needs to devise different structures and processes to improve the coordination of public services.

Moreover, just as individuals tend to ignore disasters until the costs are all too clear—and then to overact, there are powerful incentives for government to postpone action until crises hit and then to respond with strong, but often not more than symbolic, action. These crises demand resilient organizations that are able to cope with surprise such as an unexpectedly large scale disaster.

Devising more effective organizations is one challenge, but the vast scope of Katrina-like natural disasters and September 11-like terrorist attacks make it clear that no one organization, no one level of government or, indeed, government on its own can possibly cope with the impact. An effective strategy demands effective public-private partnerships, coupled with better information, balanced economic incentives, reformed disaster insurance, and better regulations and standards. Yet the task of harnessing the strengths of different private and public actors and devising a coordinated risk control and insurance strategy is a daunting exercise.

Because people have difficulty in processing data about low-probability events, the nation needs better information provision. Most people feel small numbers can be easily dismissed, while large numbers get their attention. By communicating information on the risk in different ways individuals may behave differently with respect to the actions they take before and/or after a disaster.

Positive and negative economic incentives can encourage individuals to take cost-effective protective measures prior to a disaster based on risk assessments. How people process information on the costs and benefits of reducing the risk can play an important role in whether they decide to make decisions to reduce those risks. One needs to understand what role the status quo (e.g., inertia), budget constraints (e.g., the behavior of people who live from paycheck to paycheck), and short time-horizons (e.g., individuals who seek a return on their investments) play in the process of designing sound policies.

Suppose people and businesses think only about the potential benefits of these protective measures in reducing risk over the next year or two. Then they are unlikely to view these mitigation measures as financially attractive should there be a large upfront cost associated with the investment. Had they considered a longer time period in their evaluation, the proposed measure may well have been viewed as worthwhile. Economic incentives, such as long-term loans, may be helpful in overcoming this resistance to investing in mitigation measures.

Insurance helps individuals who, before a disaster, reduce their risks and invest in policies to help them recover from any losses they suffer. For insurance to be effective in both these roles, those who are at risk must bear a substantial portion of the costs of residing in hazard-prone areas. Otherwise they will have limited economic incentive to take protective actions and will rely on others to bail them out after the next disaster. Insurance can play a key role in determining who should bear the costs of making hazard-prone communities safer and who should pay for the losses caused by disasters. If private insurance is to play a central role in a hazard management program, then those in hazard-prone areas need to bear a substantial cost of making their communities safer and should be responsible for most of the losses after a disaster occurs. The larger the subsidy provided by the general taxpayer, the less important the role that private insurance will play in signaling through premiums to how hazardous a particular region is and inducing individuals residing there to adopt risk reducing measures in advance of a disaster.

If the government anticipates it will have to step in to pay a large share of the costs, it might increase its regulatory role. Hence, well enforced standards and regulations are an important part of a hazard management strategy. Building codes may be desirable when property owners would otherwise not adopt cost-effective mitigation measures, because they either misperceive the benefits from them or underestimate the probability of a disaster occurring. When a building collapses it may create negative externalities in the form of economic dislocations and other social costs that are beyond the financial loss suffered by the owners. Losses from these and other externalities are normally not covered by an insurance policy. A well-enforced building code would help reduce these interdependent risks and obviate the need for financial assistance to those who would otherwise suffer these uninsured losses.

Principles for Shaping Hazard Management

These issues shape the way that policy makers, in both government and the private sector, must think about the issues raised by Katrina. But behind these issues lie basic principles about fairness and efficiency that must provide the foundation for policy makers' decisions. The searing images of New Orleans residents, marooned at the Superdome without food or water, provide an inescapable lesson about the importance of setting broad principles to shape these policies.

Equal Treatment. In developing a strategy for minimizing risk and responding to disasters, all Americans should be treated equally regardless of sex, race, ethnicity, and class.

Risk Analysis. Public and private organizations should consider the likelihood of disasters—including alternative strategies to minimize risks and reduce costs—in developing emergency preparedness and recovery plans, as well as plans for economic growth and development.

Cost-Benefit Analysis. Public and private organizations should assess the relevant costs and associated benefits of alternative policies. The review of benefits should include social and psychological effects as well as the direct and indirect economic impacts.

Samaritan's Dilemma. Federal disaster assistance creates a Samaritan's dilemma. If individuals and organizations assume that the federal government will provide significant assistance after hardship—that the government will be insurer of last resort—there will be fewer economic incentives for those in hazard-prone areas to reduce their risks before the next disaster and to purchase adequate insurance coverage.

Uninsured Losses. A policy by both government and the private sector that implicitly anticipates high rates of uninsured losses should not be tolerated unless one explicitly acknowledges that the public sector will respond with financial assistance after the disaster.

Assisting Low Income Residents. Subsidies should be provided to low income residents in hazard-prone areas so they can afford to protect their property and to purchase insurance in advance of a disaster. Such a program would reduce the need for disaster assistance following the next catastrophic event.

Mitigation Measures. A disaster management program should encourage those at risk to adopt mitigation measures. However, disasters inevitably involve complex and interdependent risks, and policy makers need to carefully examine public programs to avoid unintended consequences.

Loss Distribution. The disaster management strategy must consider who is most likely to suffer losses and how the costs are distributed: victims (residents and private organizations), businesses in the private sector that cover some risk (including financial institutions, insurers, and reinsurers), all levels of government (local, state and federal), and the impact on taxpayers.

Relocation of Residents. In determining strategies for relocating residents to other areas following a disaster, policy makers need to consider the economic, psychological, and social effects on victims. The same careful consideration should drive strategies for rebuilding the region (especially the Gulf Coast and New Orleans in the aftermath of Hurricane Katrina).

Governance. Policy makers should devise effective public policies that avoid creating large bureaucracies, establishing complex programs, or incurring excessive costs. The governance system should be transparent, effective, and responsive.

Appropriate Roles for the Public and Private Sectors

Even after struggling with the basic issues and the broader principles noted above, a fundamental question remains: which sectors of society ought to tackle which of these problems. On one hand, all solutions must be closely coordinated through new partnerships. On the other, even effective partnerships require some sorting out of responsibilities. Everyone cannot be responsible for everything, and, as the government's stumbling response to Katrina showed, someone has to be responsible for each important step. We often try to sort out the roles of the public and private sectors by creating bright-line tests—to set a tripwire that would limit government's role. These issues are so complex and important, however, that government inevitably plays a role in setting the rules of the game: how to set the right incentives and sort out the roles of private players.

Perhaps nothing illustrates this point better than the tale of one relief worker in Mississippi. A wheelchair-bound Mississippi senior citizen had lost her home in the hurricane. After weeks of misery and waiting, a trailer from FEMA finally arrived. But she took one look at it and recognized her dilemma. She was confined to her wheelchair and the trailer's door was a foot or more off the ground. She simply could not get into the door of her new home. Separate institutions, working separately had produced an answer that paradoxically failed to get her any closer to a solution. Fortunately for this senior citizen, a coalition of faith-based organizations was working the area. They located her, listened to her problem, and got a wheelchair ramp built. It took a complex coalition of different organizations, public, private, and nonprofit to solve her problem by building bridges across existing institutions. Without this intervention, the relief worker would have been stuck, tantalizingly close to an answer but no better off than she was before.

We cannot, in this volume, sketch out all of the solutions to the issues we face and strategies in mitigating losses from future disasters and facilitating the recovery process following a mega-catastrophe like Hurricane Katrina. However, the outline of a strategy emerges clearly from many of the chapters. Far more important is understanding the problem. Such storms—as well as other disasters like it and other catastrophes we dare not imagine—are inevitable. Katrina revealed a large gap between the capacity of our policies and institutions and our needs, as individuals and as a society. We need a fresh understanding of the problems, and new and creative solutions to tackle them. That is the most important lesson of Katrina and if, we fail to learn it, Katrina's legacy will not be "bigger and better." It will be "bigger and worse."