Risk Management and Decision Processes Center - The Wharton School of the University of Pennsylvania

Workshop on Interdependent Security

May 31 - June 1, 2006
Room 240, Jon M. Huntsman Hall
3730 Walnut Street
Wharton School, University of Pennsylvania
Philadelphia, PA

On May 31 - June 1, 2006, The Center for Risk and Economic Analysis of Terrorist Events (CREATE) at the University of Southern California, The Center for Human Performance and Risk Analysis (CHPRA) at the University of Wisconsin-Madison, and the Wharton Risk Management and Decision Processes Center sponsored a Workshop on Interdependent Security.

Conference Summary (password required)
Agenda
List of Attendees
Relevant Publications and Links

Much progress has been made over the last 60 years in understanding individual behavior under risk, cooperative group behavior under risk and competitive behavior under risk. This workshop will address an understudied but profoundly important fourth category of behavior under risk, what we call "interdependent risk". When one of us chooses to drive defensively, we lower our own risk, but also the risk of those around us. When one of us installs a (concealed) burglar alarm, we ensure retribution for a burglar of our home, but benefit everyone else by dissuading burglary in the entire neighborhood. The more important applications in today's society have to do with terrorism: if I alert Homeland Security to a suspicion, I help save both myself and my community. There are a number of researchable issues that spring to mind:

1. Do individuals act differently in interdependent risks? Do they go beyond pure self-interest in weighing their actions?
2. What incentives can be provided to citizens to act in society's best interest?
3. How can the private and public sector best persuade individuals and firms to go beyond their own self interest?

This workshop was designed to bring together a diverse set of the best thinkers from many fields to help determine a researchable agenda for this important topic. It addressed a wide variety of collective risk and decision-making scenarios including problems in airline security, corporate governance, natural hazards, computer-network security, crime, supply-chain coordination, and vaccination against diseases. "Interdependent security" (IDS) models involve not only the likelihood of adverse events occurring, but also the ways in which their consequences can propagate through complex systems, or among large numbers of individuals. Results to date generally suggest focusing on: (1) the weakest links in interdependent supply chains or other interconnected systems; and (2) the economic incentives faced by the participants in these systems to invest in mitigation measures. For example, the elements of the system that pose the greatest risk may be neither those with the greatest incentive to invest in security nor the least-cost providers of security, suggesting a need for coordination mechanisms. Models for addressing and managing these kinds of strategic interactions have significant implications for risk management in a wide range of important societal problems.

Highly interdisciplinary in nature, the workshop investigated both theoretical issues (economic modeling, game theory, and learning) and behavioral issues (e.g., behavioral economics, judgment, decision making, and problem solving) using real world examples to highlight interesting features of the problem.

Specific Objectives

The goal of the workshop was to bring theoretical researchers together with practitioners and applied researchers. In particular, the objectives of the workshop were:

1. To enhance understanding and decision making in the areas of the case studies, based on relevant insights from the theoretical and behavioral literatures;
2. To stimulate theoretical and behavioral research on important applied problems involving interdependencies; and
3. To develop strategies for improving our understanding of how to motivate individuals or organizations to invest in cost-effective measures benefiting both those individuals or organizations, and society at large. The emphasis will be on strategies that are designed to foster cooperative rather than competitive behavior among potential defenders.

In the behavioral area, our focus was on Behavioral Models and Experiments. In particular, we explored future experimental and other behavioral work on how humans and organizations actually make decisions in IDS settings, and also descriptive models of choice and learning that are likely to be used in IDS environments.

In the theoretical area, our focus was on Assessing the Strengths, Weaknesses, and Policy Implications of existing theoretical models. Examples of issues that were addressed include models of imperfect information and learning, privacy and secrecy (e.g., the relative merits of secrecy and deception versus disclosure of defensive investments), the relative effectiveness of threatened retaliation and/or positive incentives for modifying the behavior of potential attackers, and the applicability of existing models to systems with complex social networks or hierarchical structures.

Risk Center Home About the Center Directors Research Fellows Visiting Scholars Ackoff Fellows Staff Location / Directions Research Areas of Research Russell Ackoff Award for Doctoral Student Research Research Partnerships Academic Institutions Government Organizations International Organizations Corporate Associates Faculty Faculty in the News Conferences Upcoming Conferences Past Conferences Publications Books Papers Newsletters	Workshop on Interdependent Security May 31 - June 1, 2006 Room 240, Jon M. Huntsman Hall 3730 Walnut Street Wharton School, University of Pennsylvania Philadelphia, PA On May 31 - June 1, 2006, The Center for Risk and Economic Analysis of Terrorist Events (CREATE) at the University of Southern California, The Center for Human Performance and Risk Analysis (CHPRA) at the University of Wisconsin-Madison, and the Wharton Risk Management and Decision Processes Center sponsored a Workshop on Interdependent Security. Conference Summary (password required) Agenda List of Attendees Relevant Publications and Links Much progress has been made over the last 60 years in understanding individual behavior under risk, cooperative group behavior under risk and competitive behavior under risk. This workshop will address an understudied but profoundly important fourth category of behavior under risk, what we call "interdependent risk". When one of us chooses to drive defensively, we lower our own risk, but also the risk of those around us. When one of us installs a (concealed) burglar alarm, we ensure retribution for a burglar of our home, but benefit everyone else by dissuading burglary in the entire neighborhood. The more important applications in today's society have to do with terrorism: if I alert Homeland Security to a suspicion, I help save both myself and my community. There are a number of researchable issues that spring to mind: 1. Do individuals act differently in interdependent risks? Do they go beyond pure self-interest in weighing their actions? 2. What incentives can be provided to citizens to act in society's best interest? 3. How can the private and public sector best persuade individuals and firms to go beyond their own self interest? This workshop was designed to bring together a diverse set of the best thinkers from many fields to help determine a researchable agenda for this important topic. It addressed a wide variety of collective risk and decision-making scenarios including problems in airline security, corporate governance, natural hazards, computer-network security, crime, supply-chain coordination, and vaccination against diseases. "Interdependent security" (IDS) models involve not only the likelihood of adverse events occurring, but also the ways in which their consequences can propagate through complex systems, or among large numbers of individuals. Results to date generally suggest focusing on: (1) the weakest links in interdependent supply chains or other interconnected systems; and (2) the economic incentives faced by the participants in these systems to invest in mitigation measures. For example, the elements of the system that pose the greatest risk may be neither those with the greatest incentive to invest in security nor the least-cost providers of security, suggesting a need for coordination mechanisms. Models for addressing and managing these kinds of strategic interactions have significant implications for risk management in a wide range of important societal problems. Highly interdisciplinary in nature, the workshop investigated both theoretical issues (economic modeling, game theory, and learning) and behavioral issues (e.g., behavioral economics, judgment, decision making, and problem solving) using real world examples to highlight interesting features of the problem. Specific Objectives The goal of the workshop was to bring theoretical researchers together with practitioners and applied researchers. In particular, the objectives of the workshop were: 1. To enhance understanding and decision making in the areas of the case studies, based on relevant insights from the theoretical and behavioral literatures; 2. To stimulate theoretical and behavioral research on important applied problems involving interdependencies; and 3. To develop strategies for improving our understanding of how to motivate individuals or organizations to invest in cost-effective measures benefiting both those individuals or organizations, and society at large. The emphasis will be on strategies that are designed to foster cooperative rather than competitive behavior among potential defenders. In the behavioral area, our focus was on Behavioral Models and Experiments. In particular, we explored future experimental and other behavioral work on how humans and organizations actually make decisions in IDS settings, and also descriptive models of choice and learning that are likely to be used in IDS environments. In the theoretical area, our focus was on Assessing the Strengths, Weaknesses, and Policy Implications of existing theoretical models. Examples of issues that were addressed include models of imperfect information and learning, privacy and secrecy (e.g., the relative merits of secrecy and deception versus disclosure of defensive investments), the relative effectiveness of threatened retaliation and/or positive incentives for modifying the behavior of potential attackers, and the applicability of existing models to systems with complex social networks or hierarchical structures.
© 2006 The Wharton School, University of Pennsylvania