Introduction —Howard Kunreuther, Robert J. Meyer, and Erwann O. Michel-Kerjan PART I. BEHAVIORAL FACTORS INFLUENCING DECISION-MAKING UNDER RISK AND UNCERTAINTY Chapter 1. The Arithmetic of Compassion and the Future of Risk Management —Paul Slovic and Daniel Västfjäll Chapter 2. "Risk as Feelings" and "Perception Matters": Psychological Contributions on Risk, Risk-Taking, and Risk Management —Elke U. Weber Chapter 3. Risk-Based Thinking —Baruch Fischhoff Chapter 4. Structured Empirical Analysis of Decisions Under Natural Hazard Risk —Craig E. Landry, Gregory Colson, and Mona Ahmadiani Chapter 5. Mixing Rationality and Irrationality in Insurance Demand and Supply —Mark Pauly Chapter 6. The Disaster Cycle: What We Do Not Learn from Experience —Robert J. Meyer PART II. IMPROVING RISK ASSESSMENT Chapter 7. Using Models to Set a Baseline and Measure Progress in Reducing Disaster Casualties —Robert Muir-Wood Chapter 8. Learning from All Types of Near-Misses —Robin Dillon Chapter 9. Managing Systemic Industry Risk: The Need for Collective Leadership —Paul J. H. Schoemaker Chapter 10. Measuring Economic Resilience: Recent Advances and Future Priorities —Adam Rose PART III. DEVELOPING BETTER RISK COMMUNICATION STRATEGIES Chapter 11. Improving Stakeholder Engagement for Upstream Risks —Robin Gregory and Nate Dieckmann Chapter 12. Improving the Accuracy of Geopolitical Risk Assessments —Barbara A. Mellers, Philip E. Tetlock, Joshua D. Baker, Jeffrey A. Friedman, and Richard Zeckhauser Chapter 13. Efficient Warnings, Not "Wolf or Puppy" Warnings —Lisa A. Robinson, W. Kip Viscusi, and Richard Zeckhauser PART IV. ROLE OF RISK MITIGATION, RISK-SHARING, AND INSURANCE Chapter 14. Threats to Insurability —Carolyn Kousky Chapter 15. The Role of Insurance in Risk Management for Natural Disasters: Back to the Future —Howard Kunreuther Chapter 16. Improving Individual Flood Preparedness Through Insurance Incentives —W. J. Wouter Botzen Chapter 17. Strong and Well-Enforced Building Codes as an Effective Disaster Risk Reduction Tool: An Evaluation —Jeffrey Czajkowski PART V. GOVERNMENT AND RISK MANAGEMENT Chapter 18. Getting the Blend Right: Public-Private Partnerships in Risk Management —Cary Coglianese Chapter 19. The Regulation of Insurance Markets Subject to Catastrophic Risks —Robert W. Klein Chapter 20. Rethinking Government Disaster Relief in the United States: Evidence and a Way Forward —Erwann O. Michel-Kerjan List of Contributors Index

Highlighting past research, recent discoveries, and open questions, The Future of Risk Management provides scholars, businesses, civil servants, and the concerned public tools for making more informed decisions and developing long-term strategies for reducing future losses from potentially catastrophic events.

Introduction —Howard Kunreuther, Robert J. Meyer, and Erwann O. Michel-Kerjan PART I. BEHAVIORAL FACTORS INFLUENCING DECISION-MAKING UNDER RISK AND UNCERTAINTY Chapter 1. The Arithmetic of Compassion and the Future of Risk Management —Paul Slovic and Daniel Västfjäll Chapter 2. "Risk as Feelings" and "Perception Matters": Psychological Contributions on Risk, Risk-Taking, and Risk Management —Elke U. Weber Chapter 3. Risk-Based Thinking —Baruch Fischhoff Chapter 4. Structured Empirical Analysis of Decisions Under Natural Hazard Risk —Craig E. Landry, Gregory Colson, and Mona Ahmadiani Chapter 5. Mixing Rationality and Irrationality in Insurance Demand and Supply —Mark Pauly Chapter 6. The Disaster Cycle: What We Do Not Learn from Experience —Robert J. Meyer PART II. IMPROVING RISK ASSESSMENT Chapter 7. Using Models to Set a Baseline and Measure Progress in Reducing Disaster Casualties —Robert Muir-Wood Chapter 8. Learning from All Types of Near-Misses —Robin Dillon Chapter 9. Managing Systemic Industry Risk: The Need for Collective Leadership —Paul J. H. Schoemaker Chapter 10. Measuring Economic Resilience: Recent Advances and Future Priorities —Adam Rose PART III. DEVELOPING BETTER RISK COMMUNICATION STRATEGIES Chapter 11. Improving Stakeholder Engagement for Upstream Risks —Robin Gregory and Nate Dieckmann Chapter 12. Improving the Accuracy of Geopolitical Risk Assessments —Barbara A. Mellers, Philip E. Tetlock, Joshua D. Baker, Jeffrey A. Friedman, and Richard Zeckhauser Chapter 13. Efficient Warnings, Not "Wolf or Puppy" Warnings —Lisa A. Robinson, W. Kip Viscusi, and Richard Zeckhauser PART IV. ROLE OF RISK MITIGATION, RISK-SHARING, AND INSURANCE Chapter 14. Threats to Insurability —Carolyn Kousky Chapter 15. The Role of Insurance in Risk Management for Natural Disasters: Back to the Future —Howard Kunreuther Chapter 16. Improving Individual Flood Preparedness Through Insurance Incentives —W. J. Wouter Botzen Chapter 17. Strong and Well-Enforced Building Codes as an Effective Disaster Risk Reduction Tool: An Evaluation —Jeffrey Czajkowski PART V. GOVERNMENT AND RISK MANAGEMENT Chapter 18. Getting the Blend Right: Public-Private Partnerships in Risk Management —Cary Coglianese Chapter 19. The Regulation of Insurance Markets Subject to Catastrophic Risks —Robert W. Klein Chapter 20. Rethinking Government Disaster Relief in the United States: Evidence and a Way Forward —Erwann O. Michel-Kerjan List of Contributors Index * * * * * Introduction Howard Kunreuther, Robert J. Meyer, and Erwann O. Michel-Kerjan Hurricane Katrina. Superstorm Sandy. Hurricanes Harvey, Irma, and Maria. Contaminated water in Flint, Michigan. Anthrax. 9/11 terrorist attacks. Earthquakes. Oil and chemical spills. Boston Marathon attack. Financial crisis. Refugee crisis. Ebola. Sea level rise. Devastating floods. Cyber-attacks. Geopolitical instability. Whether man-made or naturally occurring, an unprecedented series of large-scale disasters and crises have caused a vast number of fatalities and injuries, destroyed property, devastated communities, savaged the environment, imposed significant financial and emotional burdens on individuals and firms, and challenged political leadership. The Future of Risk Management Thirty years ago, the study of risk management was viewed narrowly. Scientists and engineers provided estimates of the probability of specific types of disasters occurring and their potential consequences. Economists proposed risk management policies based on these experts' estimates with little thought as to how these data would be used by interested parties. Today we live in a new age of catastrophe and increasing uncertainty. Besides the disasters that arise with alarming regularity, we are faced with major challenges in dealing with natural and technological hazards and global issues, such as climate change and terrorism. What makes these risks all the more difficult to address is that the world has become more interdependent and interconnected; what occurs in one nation or geographical region is likely to have ripple effects across the globe. The information age is also more integrated, creating new forms of communication and greater risks that are not easy to evaluate. All of this implies that new approaches to risk management are required. These measures need to combine the latest scientific knowledge on risk assessment with a better appreciation of the importance of improving individual and collective decision-making processes. While significant progress has been made in quantitatively assessing risks, there is also a growing recognition that we need to integrate psychological and behavioral elements into our risk management strategies. By recognizing the systematic biases and simplified heuristics used by decision-makers, we should be better able to develop effective strategies for coping with them. Today, risk assessment and risk management are no longer the sole domain of scientists, engineers, and economists. The disciplines of finance, geography, history, insurance, marketing, political science, psychology, sociology, and the decision sciences now work together, challenge one another, and in the process, build new approaches. In this edited volume we offer a view into the present and future of risk management. We invited leading thinkers in risk management and the behavioral sciences to address the question: What is the future of risk management based on what we have learned from our past research? While this book is not intended to provide a comprehensive analysis of the topic, it allows the reader to benefit from the experience and perspective of a number of respected scholars in the field as they look back at recent discoveries and at some of the important open questions we still face. Over its 35-year history, the Wharton Risk Center has focused its research on how individuals, organizations, and government at all levels can deal with extreme events. While many decisions regarding low-probability, high-consequence events are based on intuitive thinking—with emotions, past experience, and systematic biases playing a key role—more deliberative thinking such as the use of decision analysis and cost-benefit analysis can provide important insights to develop effective strategies. Organization of the Book The book is organized into five parts, each of which addresses avenues for practitioners and researchers to consider. We briefly summarize below the chapters in each of these parts. Behavioral Factors Influencing Decision-Making Under Risk and Uncertainty Part I probes the behavioral and psychological factors that influence decision-making in risky situations. Why do some people invest in protective measures to prevent severe losses from the next disaster, while others do not? What catastrophes or personal hardships elicit widespread compassion that other less salient events do not? What factors influence people in their risk perceptions, risk-taking, and risk management? The chapters point to several new directions as this topic is quickly evolving and needs more attention in the future. Paul Slovic and Daniel Västfjäll discuss three nonrational psychological mechanisms that confound the arithmetic of compassion: psychophysical numbing (insensitivity to large numbers of losses of human lives), pseudoinefficacy (being deterred from helping one person because one cannot help others), and the prominence effect (the disconnect between our stated values and the values revealed by our actions). Their chapter contrasts risk as feelings with risk as analysis and suggests that it is time to examine strategies that acknowledge the psychological challenges of intuitive thinking and encourage more systematic and reasoned deliberative thinking. Elke U. Weber confirms that the concept of risk can be interpreted as a statistic via metrics like the variance of outcomes or as feelings driven by emotional responses. She shows how the DOmain-SPEcific Risk-Taking (DOSPERT) scale differentially predicts observed real-world risk taking in different domains by incorporating domain-specific perceptions of risk, and how the Columbia Card Task (CCT) captures risk-taking in dynamic environments. Weber concludes by showing how risk can be better managed by appreciating subjective perceptions of risk, including risk as feelings, so that one can reframe problems in ways that align observed behavior more closely to desirable behavior. Baruch Fischhoff recognizes that assessing the quality of individuals' decision-making is technically demanding. He then raises the question as to how we can communicate the findings from our research on risk and choice to key decision-makers in ways that will make a difference to their behavior. He suggests that future risk research should focus on a longitudinal study of decision-making competence and methods for communicating the quality of risk research. Craig E. Landry, Gregory Colson, and Mona Ahmadiani highlight the limitations of expected utility theory in explaining why individuals do not invest in protective decisions. They then suggest alternative models of choice that reflect the role that subjective probability and preference parameters play in the decision-making process. The authors recommend survey-based studies using panel data over time at the individual and household level for developing models that explain natural hazard risk management decisions. Mark Pauly focuses on what we know about the mix of rationality and irrationality in consumers, focusing on health insurance purchasing in the United States as an illustrative example. He notes that a rational model of choice, such as expected utility maximization, is highly demanding in terms of time and data collection. He also points out that there are individuals who do not make the optimal decision of purchasing highly subsidized health insurance when they should know they are getting a good deal. Pauly concludes that there are fewer deviations from the rational model in health insurance than other insurance markets in part because health-related extreme events are less common and losses are rarely correlated across exposures. Robert J. Meyer concludes Part I with a discussion as to why we often fail to learn as much as we should from disasters. He argues that much of the blame lies in how our brains are innately wired to learn; we repeat actions that give positive rewards and avoid those that produce negative returns. While these instincts work well in most circumstances, when applied to protective decisions for rare events, the reward structure sometimes is reversed. In the immediate wake of a disaster it is easy to see the benefits of protection, but as time passes, the positive returns from these investments become harder to see, eventually leading to their abandonment. As a result, disasters often are destined to repeat themselves. Improving Risk Assessment Part II explores how risk assessment provides the scientific ingredients for developing risk management strategies. What data should individuals, groups, businesses, and government take into account when evaluating risk and determining what strategies to consider and examine carefully? Given the uncertainty of extreme events, what tools can organizations consider when developing long-term strategies for undertaking protective measures and reducing future losses from potential disasters? Robert Muir-Wood begins his chapter by noting that the March 2015 Sendai Declaration, signed by 187 countries, was concerned with achieving substantial reductions in disaster casualties by 2030. He then raises the question as to whether the benchmark years 2005-2015 employed in the Sendai process, for comparison with the future 2020-2030 period, provide a biased sample, capable of indicating spurious evidence of progress. Muir-Wood concludes that goals for casualty risk reduction should be directed at individual countries so they are owned by institutions with the power to alter future outcomes. However, the extreme volatility of mega-catastrophes means that catastrophe models, with their long-term synthetic histories, should be used for setting and measuring national targets for disaster casualty reduction based on "expected" fatalities. Robin Dillon focuses on the importance of learning from near-misses and notes that people's own experience influences how they interpret these events. If they haven't had a disaster they may believe that the near-miss lowers their probability of a future disaster because they previously escaped unharmed. She summarizes recent experiments on false alarms that reveal that individuals exposed to two false alarms are likely not to heed future warnings. Dillon concludes by highlighting the importance of emergency managers understanding people's responses to near-misses to successfully communicate risk of hazardous situations to the public. Paul J. H. Schoemaker contends that business leaders will increasingly have to prevent or mitigate industry-wide risks through better industry-level collaboration. He explains that American credit unions weathered the 2008-2009 financial crisis relatively well because industry leaders recognized the importance of focusing on strategic issues related to regulation, technology, and business models as early as 1997. He also discusses why other sectors in the financial services field did not fare as well. The leadership challenges at an industry level are to deploy strategies and capabilities that are flexible enough to deal with unanticipated black swan events as well as a rising tide of systemic risks. Adam Rose focuses on questions associated with measuring economic resilience by proposing an operational metric and advocating the use of cost effectiveness and cost-benefit analysis to make prudent and careful resource management decisions. He highlights the direct and indirect benefits and co-benefits that emerge from undertaking risk reduction measures, and proposes the use of innovative financing instruments and insurance to incentivize these expenditures now rather than waiting until the next disaster occurs. Developing Better Risk Communication Strategies The chapters in Part III examine challenges in risk communication. How do we issue warnings so that people will pay attention to risks that matter? Why don't people pay attention to the lessons from past disasters when making decisions regarding investing in protective measures? The goal of such strategies is to engage stakeholders more effectively in the risk management process. Robin Gregory and Nate Dieckmann highlight the importance of informed deliberation by stakeholders in evaluating the uncertainty associated with different proposed programs for managing future environmental risks. They propose several methods for eliciting stakeholder opinions to help overcome judgmental biases that often arise when dealing with the unfamiliar and novel choices associated with upstream technologies. Gregory and Dieckmann also note the importance of communicating uncertainty so that people understand why experts' predictions for a specific phenomenon may differ and that individuals presented with identical information may interpret the uncertainty differently as the result of varying beliefs, worldviews, or motivations. Barbara A. Mellers, Philip E. Tetlock, Joshua D. Baker, Jeffrey A. Friedman, and Richard J. Zeckhauser discuss the factors that improved accuracy in a four-year geopolitical forecasting tournament and the characteristics of the best forecasters. Top performers tended to use many distinctions along the probability scale. The authors use the data from the forecasting tournament to investigate the number of categories necessary to maximize forecasting accuracy. The intelligence community in the United States currently recommends a seven-point rating scale for expressing uncertainty. A comparison of forecasters' original accuracy scores with the scores they would have obtained after rounding forecasts to seven bins shows that prediction errors grow if analysts are only allowed to convey seven degrees of doubt. Even worse, the accuracy of the best forecasters suffers the most when probability scales do not have sufficient shades of gray. Lisa A. Robinson, W. Kip Viscusi, and Richard Zeckhauser argue that warnings, appropriately employed, are often superior to command and control regulation in controlling individuals' exposures to risks. Unfortunately, products imposing vastly different dangers are often required to employ the same warnings. California Proposition 65's indiscriminate labeling of 800+ substances as carcinogens or reproductive toxins is an exemplar. Individuals cannot be expected to distinguish among such risks when making consumption decisions. Other warnings, such as about mercury in some seafoods, simply confuse. Consumers may stop eating healthy seafood, failing to identify fishes that are safe so that significant welfare losses can be expected. Warnings should be designed in a manner that leads consumers to, at the least, distinguish between serious and mild risks. Role of Risk Mitigation, Risk-Sharing, and Insurance The chapters in Part IV focus on the need to prepare for the financial impacts of adverse events. How can insurance encourage those at risk to undertake loss reduction measures before the next disaster? How does one deal with issues of affordability? What is the appropriate role of the public sector in dealing with costs of catastrophic risk? Risk-sharing is a key principle in the management of disasters, and discussions about fairness and affordability in designing insurance programs are now getting more attention. Carolyn Kousky discusses many converging trends that might make natural disasters increasingly harder to insure, such as dependent and systemic risks. In the face of such changes, a more thoughtful approach to risk management that stresses targeted mitigation and develops complementary roles for the public and private sectors—in both risk reduction and risk transfer arrangements—can improve insurability. Howard Kunreuther proposes that insurers follow the example of the nineteenth-century factory mutual insurance companies who required those requesting insurance to invest in risk-reducing measures as a condition for coverage. Today, state regulators need to allow insurers to charge risk-based rates so they can incentivize their clients to undertake loss reduction measures. There is also a need to use choice architecture to frame the problem in ways that those at risk will want to purchase insurance and invest in protective measures. The public sector can play an important role by addressing fairness and affordability issues, as well as providing protection against catastrophic losses that are currently uninsurable. W. J. Wouter Botzen examines economists' contention that insurance and risk reduction measures are substitutes so that individuals who purchase insurance are less likely to invest in risk mitigation measures than those who are uninsured. Surveys in Germany and New York City reveal that those who purchase insurance also mitigate, implying that there is not always a moral hazard problem associated with the purchase of insurance as the theory of insurance has traditionally assumed. He suggest additional research on how one can utilize insurance to provide financial incentives to encourage individuals to better prepare for the next disaster before it is too late. Jeffrey Czajkowski provides empirical evidence that windstorm losses due to stronger building codes have been reduced by 40 to 60% and are cost-effective with respect to new construction. At the same time he shows that many local communities do not enforce their building codes in part due to the lack of financial and technical resources. Surveys of Oklahoma homeowners' attitudes toward building codes reveal that most favor a law requiring new homes to be better designed to withstand high winds. Czajkowski concludes that future research should focus on what drives building code enforcement at the local level and the degree of consumer support for improved building codes when the actual costs are explicitly indicated. Government and Risk Management Part V examines important and growing questions about the role of government in dealing with extreme events. What is the right mix for the public and the private sectors to play in providing protection against risk? Should government intervention be more local or more federal? How does such intervention affect personal responsibility and the demand for protection? The degree of public-private partnerships will naturally vary for different risks. Cary Coglianese notes that all risk management requires forming the right kind of partnership between the public and private sectors and illustrates this point by stressing the importance of creating effective relationships between private and public actors, such as when government regulation is coupled with private insurance (and vice versa). He says that, to facilitate a constructive partnership, decision-makers and organizers need to satisfy four core factors to solve a particular problem: interface, incentives, information, institutions—or what he calls the "four i's." Coglianese concludes with a call for more behavioral research on the fit between public- and private-sector risk management interventions, as well as the important role of distributional politics in affecting the success of public-private partnerships. Robert W. Klein contends that insurance regulation can enhance social welfare if it addresses market failures that it can remedy or ameliorate and seeks outcomes consistent with a competitive market. He also argues that regulation that promotes efficient insurance markets contributes to effective risk management. Using these basic principles as guidelines, he examines the performance of current regulatory policies with respect to the following insurance-related issues for hurricanes and other catastrophic risks: pricing and underwriting, adequacy of coverage, the management of residual market mechanisms, and the solvency of insurance companies, as well as the implications of these policies for catastrophe risk mitigation. Erwann O. Michel-Kerjan concludes the book by analyzing how the federal government in the United States has played a more important role in intervening after disasters in recent years than it has done historically. This new reality may have created significant moral hazard: if they expect to be financially supported after a disaster, people, firms, and cities might invest less in pre-disaster preparedness than they would otherwise. He proposes a transparent national accounting system of how much is spent on disaster relief, where the money is coming from and where it is going; in other words, who pays for this spending and who benefits from it. These data should provide inputs to proposing better disaster preparedness policies based on measurable evidence, as well as providing greater incentives for states and other interested parties to invest in catastrophe risk management today. As noted, this volume offers a snapshot of risk management today along with directions for future research. We look forward to deeper financial analysis, thoughtful policy discussions, and conversations among the wide array of stakeholders in the arena of risk management. Our hope is that 35 years from now, scholars, business people, civil servants, and the concerned public will have tools that provide even greater ability to make informed decisions to deal with extreme events.