Organizational Cooperation in Crises

Citation preview

Organizational Cooperation in Crises

This page has been left blank intentionally

Organizational Cooperation in Crises

Lina M. Svedin University of Utah, USA

© Lina M. Svedin 2009 All rights reserved. No part of this publication may be reproduced, stored in a retrieval system or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise without the prior permission of the publisher. Lina M. Svedin has asserted her right under the Copyright, Designs and Patents Act, 1988, to be identified as the author of this work. Published by Ashgate Publishing Limited Ashgate Publishing Company Wey Court East Suite 420 Union Road 101 Cherry Street Farnham Burlington Surrey, GU9 7PT VT 05401-4405 England USA www.ashgate.com British Library Cataloguing in Publication Data Svedin, Lina, 1974  Organizational cooperation in crises.   1. Crisis management. 2. Interagency coordination.   3. Intergovernmental cooperation. 4. Cooperativeness.   5. Disaster relief--International cooperation.   6. Organizational behavior.  I. Title   363.3'4525-dc22 Library of Congress Cataloging-in-Publication Data Svedin, Lina M.   Organizational cooperation in crises / by Lina M. Svedin.    p. cm.  Includes bibliographical references and index.   ISBN 978-0-7546-7725-3 -- ISBN 978-0-7546-9441-0 (ebook) 1. Crisis management. 2. International cooperation. I. Title.   JZ5595.S84 2009   658.4'056--dc22 ISBN 978-0-7546-7725-3 (hbk) ISBN

2009003044

Contents List of Figures and Tables   Acknowledgements  

vii ix

1

Introduction   1 Challenges to Crisis Cooperation   4 Theoretical Explanations of Challenges to Cooperation   6 Exploring Crisis Cooperation through Specific Research Questions  11 Methods   12 Data   13 Outline of the Chapters   13

2

Conceptualizing Organizational Crisis Cooperation: The Legacy of Three Traditions   15 Conceptualizing and Examining Organizational Crisis Cooperation  15 The Transboundary Crisis Management Data: Developing a Dataset to Look at Organizational Cooperation in Crises   41 CATPCA: A Way of Empirically Examining Cooperation in Crises  43

3

Crisis Cooperation in Light of the Three Traditions: Case Illustrations   An International Relations Perspective on Cooperation   Case Example 1: NATO and the Kosovo Crisis   An Organizational Perspective on Cooperation   Case Example 2: The Korean Financial Crisis   A Psychological Perspective on Cooperation   Case Example 3: The Red River Flood  

45 45 46 49 49 52 52

4

Organizational Behavior in Decision-Situations   Indicators Included   Extraction Method   Component Loadings after Rotation   Discussion of the Decision-occasion Level Components   Indicators that Cut across More than One Component  

57 57 58 61 62 70

Organizational Cooperation in Crises

vi

5

Cooperative Strategies across Crises   Variables Included   Extraction Method   Component Loadings after Rotation   Discussion of the Strategies across Crises   Variables that Run across More than One Component  

6

Linking Behavior and Strategies   93 The Empirical Relationship between the Cooperation Variables   93 Correlations between Cooperation Behavior and Strategies   106

7

Cooperating in a Crisis Context   Examining the Connection between Cooperative Behavior and Strategies, and the Characteristics of Crises  

115

8

Organizing for Crisis Cooperation: Conclusions and Implications   The Characteristics of Organizational Cooperation in Crises   Implications for Practitioners in the Field  

123 124 131

9

An Agenda for Continued Research   Our Next Research Questions   Testing the Effect of Independent Variables  

139 139 142

Bibliography   Index  

73 73 73 77 77 90

113

145 159

List of Figures and Tables Figures 6.1

Model of correlations between cooperation behavior at the decision-occasion level of analysis and cooperation strategies at the case level of analysis  

8.1 Model of relations between organizational cooperation behavior (decision-occasion level) and strategies (case level) in crises   8.2 Relations between crisis characteristics and organizational cooperation behaviors and strategies  

110

129 131

Tables 2.1 Operationalization of cooperative behavior at the decision-occasion level of analysis ranging from strongest (yield) to weakest form of cooperation (structural violence) 28 2.2 Indicators operationalizing cooperation at the case level of analysis 31 4.1 Descriptive statistics for the decision-occasion level variables 4.2 Eigenvalues and the variance accounted for by each component in the component solution for the decision-occasion level data 4.3 Varimax rotated solution of transposed decision-occasion level variables 4.4 Case example of fight 4.5 Case example of agree 4.6 Case example of talk 4.7 Case example of negotiate 4.8 Case example of manipulate

58

5.1 Descriptive statistics of cooperation at the case level 5.2 Summary of the cooperation components extracted at the case level 5.3 Varimax rotated CATPCA solution of case level variables 5.4 Case example of bureaucratic politics 5.5 Case example of concurrence seeking

74

61 62 64 65 67 68 70

76 76 79 82

viii

5.6 5.7

Organizational Cooperation in Crises

Case example of signaling trustworthiness Case example success-based helping

6.1 Correlations between decision-occasion level indicators and case level indicators 6.2 Descriptive statistics for the correlation analysis of types of cooperation behavior and cooperation strategies 6.3 Correlations between types of cooperation behavior and cooperation strategies 7.1 7.2 7.3 7.4

Correlations between cooperation behavior and strategies, and crisis characteristics Significant ANOVA results for regression models of crisis characteristics and their effect on cooperative behaviors and strategies Significant effects of crisis characteristics on specific cooperative behaviors and strategies Correlations between the crisis characteristics

86 89 94 107 108 117 119 119 121

Acknowledgements I would first like to express my gratitude to a number of people that have been, and continue to serve as, my academic mentors. I’m indebted to all of them for their faith in me, their encouragement, and for taking time out of their busy careers to show me the ropes. I would especially like to thank Peg Hermann and Bruce Dayton for the time, the laughs, the tough discussions and the lunches that we shared while developing and coding the TCM dataset. Keith Bybee and Jim Bennett deserve special mention for witty and insightful comments on my work and for being willing and able to think way outside “the box”. I would also like to thank Bengt Sundelius and Peg Hermann for supporting many of my conference participations, for introducing me to some of the greatest minds in our fields, and for breaking ground for me in more ways than I will ever know. A lot could be written about the many ways in which you have both influenced my life, my growth, my career choices, and the way I approach good scholarship. I know that both of you had a hard time seeing how my move to Utah would benefit you, but I am very thankful for all that you have taught me about good decision-making and I am happy to know that you would have me back if I ever needed to “come back home”. Among the many colleagues that I have spent much time harassing with my ideas and coding schemes, many deserve special mention. Most especially Toby van Assche, Katya Kalandadze, and Johan Eliasson. You have provided invaluable support through this writing process, and all the crises that have accompanied it. I owe thank you to Richard Sherman for encouraging me, for supporting me financially for several months while I was focusing on writing this book, and for always being keen on discussing statistics with me no matter when or where. My path has crossed so many times with Asthildur Bernardsdottir, and I am so grateful for the opportunity to work with you and to share life moments with you (good and bad). To my brilliant, fun, and hardworking colleagues at CRISMART. No one could ask for a better group of people to bounce ideas off, work with, and embark on impossible ventures with. Thanks to all of you for the years we have spent growing and creating together. A special thank you goes out to Dan Hansén for his uncanny ability to be there right when I need a hand. I would also like to thank Andy Field. Your humor has saved my life. Although we have never met or spoken, through your writing I feel I have gotten to know you (and a lot about statistics) and that we have gone through some tough times together. While our relationship is purely fictional, by now I consider you an old and dear friend.



Organizational Cooperation in Crises

I am grateful for the Goekjian Summer Research Grant, which lent me financial support to develop the codebook for the dataset and for a forum where I could share early ideas on this study. I am also very grateful for the SLAPP workshop where good people read my far too long chapters, and gave me helpful comments and suggestions. I would like to thank the department of Public Administration at the Maxwell School for giving me the opportunity to teach the Case Writing Workshop for six semesters. A special thank you also to Tammy Salisbury and Debbie O’Toole for being just wonderful people, and for making teaching while writing a book such a smooth process. I would like to thank my students in the MPA Executive Education Program and the Army Comptroller Program at Syracuse University for the energy and tenacity they showed in pursuing their case research. I would also like to thank them for their willingness to believe that we can strengthen administrative crisis management capacity through the process of doing case research. Seldom have I laughed as much, gotten as many good and fundamental questions, and seen so many people grow and flower into something they, themselves, never believed they could, as I have leading the Case Writing Workshop. I would also like to express my fond gratitude to Candy Brooks, who effortlessly manages to find solutions to every scholar’s many practical challenges with a smile on her face. You are living proof that every academic department would come to a grinding halt and fall into highly educated chaos if it were not for the people “disguised” as administrators that really run the show. My last thank you’s I direct to Daniel Patterson, my beloved partner, who has provided brilliant copy editing and much enthusiastic support for this book. My mom, who is probably the proudest of my accomplishments, and to my sister, who has unwaveringly titled me professor since the day I started my undergraduate studies. Their faith in me keeps me going. Lina M. Svedin December 2008

Chapter 1

Introduction There are a number of ways in which cooperation has direct relevance for crises and crisis management. Crises involve threats to individuals and organizations. As some scholars have asserted, crises can be viewed as “a serious threat to the basic structures or the fundamental values and norms of a social system, which – under time pressure and highly uncertain circumstances – necessitates making critical decisions” (Rosenthal and ‘t Hart quoted in Boin 2004, 167). Cooperation is, as argued in this chapter, an effective and efficient way to be more resilient in the face of threat. Furthermore, crises often involve an acute scarcity of resources and, as will be discussed further in this chapter, cooperation is one way of maximizing the access to and use of scarce resources. Furthermore, the management of crises is often organized in an interorganizational way, and effective management in inter-organizational structures requires either a clear command structure (with a possibility of actually allocating the needed resources and if necessary enforce decisions) or cooperation among horizontally organized actors. Inter-organizational responses to crisis have surged post 9/11 as several countries have gone through far-reaching administrative restructuring in order to achieve better coordination among agencies and departments with responsibilities in the area of civil security. In the United States, the creation of the Department of Homeland Security, the contents of the 9/11 Commission report, and the creation of a new national incident management system all emphasize coordination and collaboration across organizational  ������������������������������������������������������������������������� This interest in improving and supporting cooperation within and between organizations in crises is based on the assumption that cooperation in crises, generally speaking, is a good thing. That is, when organizations cooperate, rather than end up acting in conflict, the overall response to the crisis will be better managed in terms of the immediate and long-term costs (economic, political, social costs) for society as a whole. However, there are also instances when we can expect cooperation to have negative effects on the overall response to a crisis and examples of this will be included in the discussion of cooperation as the dependent variable in this study. Determining the effect of cooperation on organizational or management performance is not within the scope of this study, rather the scope is limited to examining organizational cooperation in crises as a phenomenon and what the connections between its varying expressions are across a variation of crisis characteristics.  ������������������������������������������������������������������������������� Another reason why we should study organizational cooperation is that failures to cooperate in crises have important consequences. A failure to respond effectively to a crisis leads to an increased risk of direct harm for those affected and may imply drastically increased societal costs to bring about recovery.



Organizational Cooperation in Crises

boundaries in crises (9/11 Commission 2004; Department of Homeland Security 2004a; Department of Homeland Security 2004b; Försvarsberedningen 2007). A similar emphasis has been expressed in Great Britain where the establishment of a new inter-agency working procedure in times of crises clearly outlines the need for, and indeed responsibility of, government agencies to work together toward the common goal of preventing and responding to crises (Emergency Preparedness 2004). Inter-organizational work, whether institutionalized or temporarily arranged, requires cooperation among organizations. Effectiveness in an inter-organizational setting is determined by the actors’ ability to coordinate, communicate, and adapt to each other’s goals, preferences and work procedures so that the organizations work as ‘one’ organizational body. Cyert and March’s (1963) outline of the challenges to getting one organization to work as ‘one mind’ seems highly relevant to contemporary inter-organizational cooperation as it is plagued by similar challenges of coordination, motivation and goal orientation. In crises ad hoc decision-making and management groups are common. The actors included in an inter-agency task force, a joint command force, or an interorganizational management team often vary with the type of crisis the society is facing. Furthermore, ad hoc groups in crises seldom have clear command structures, but rely on voluntary coordination and adjustment to achieve effective and efficient responses to crises (e.g., Page 2003). Cooperation, in the loose sense of the word, thereby has the potential to play an important role in shaping the overall response to a crisis. Two additional factors that underline the importance of cooperation in crises are normative, rather than objective, in nature. First, the public often expects organizations, in a time of threat, to rally around the flag (e.g., Baker and Oneal 2001) in order to limit the uncertainty about the threat and to maximize the use of resources to meet a common challenge. In a sense these kinds of expectations go back to the social contract between the public and its governing institutions that serves as the basis of the administrative state. The basic social contract requires the state to take care of its citizens. Political leaders contract to provide protection, order, security, enforcement of property rights and other collective goods in exchange for tax revenue, some of which goes to provide the protection, order, security, enforcement and collective goods, and some which goes to support political leaders (Yarbrough and Yarbrough 2003, 543–44). Failures by public leaders and organizations to cooperate in the face of a threat can be perceived by the public as a lack of protection and care by the state. A failure to cooperate then has the potential to undermine the public’s trust and confidence in its governing institutions. A lack of adequate protection and care by representatives of the state then constitutes a break with the value-based or moral understanding of the social contract between the state and its constituents. In situations where only part of society, close allies or ‘friends’ are threatened the public still expects those not directly threatened (but that are part of a network

Introduction



or cooperative venture) to help those in need. This expectation has a number of learned behavioral and normative foundations. One is the idea of reciprocity that figures in any cooperative venture, i.e., that “I do this for you now so that later, if I need it, you will help me.” Another is the perceived virtue of pro-social behavior, such as helping another. To help someone in need is a ‘nice’ thing to do; a positively sanctioned pro-social behavior. It makes us, as individuals and groups, feel connected and rewards us in exogenous or endogenous ways. The act of helping boosts our perception of ourselves and the perception that others have of us (see for instance Perry 1996, 1997; Perry and Wise 1990). Second, historical examples show that these public expectations are real, and that failure to live up to norms of rallying around the flag and helping ‘friends’ in need may lead to heavy social sanctioning. This may take the form of public inquiries holding government agencies and organizations responsible for failing to cooperate in the aftermath of a perceived failed crisis response. For instance, after the tsunami disaster in Southeast Asia in December 2004, there was a frantic scramble on the ground to rescue people and to get them medical help. The massive impact of the disaster was soon broadcast on international media and intelligence reports of the scope and potential effects of the disaster quickly reached government offices around the world. In Sweden, the response at the national level to information about the disaster was fraught with inaction, a lack of information sharing, and a failure to develop a shared problem frame (Katastrofkommissionen 2005). With a fragmented response and working at cross-purposes, the different departments of the government and their chief executives not only delayed rescue efforts substantially (Katastrofkommissionen 2005), but also inspired a massive critique by the victims of the disaster and the general public that culminated in a long and grueling public inquiry process. During the public hearings, evidence of the lack of coordination and cooperation between departments and ministers of the government shook the public’s confidence in government. Several narrowly averted votes of no-confidence in Parliament and the findings in the commission report have since brought about a massive effort to improve coordination within  ������������������������������������������������������������������������������������ To help is understood here as to provide a person or a group of persons “with means toward what is needed or sought, be of use or service to. Helping may involve costs to the helper, but it may also bring rewards: in most cases, probably, it does both” (Hinde and Groebel 1991, 4).  ����������������������������������������������������������������������������� These expectations may not only become manifest during acute crises, but may also extend to risk elimination and risk management in society. Hopkins (2005) points to the increasing importance of risk in society and suggests that we may today be facing “a new kind of society, ‘risk society’.” With this he means that modern societies are not in fact facing unprecedented levels of risk, rather, “we live in a society in which risks are more controllable than ever; hence, the increasing demand that they be controlled” (Hopkins 2005, xi).  �������������������������������������������������������������������������������� The tsunami disaster was the largest disaster to hit Swedish citizens in modern times. Thailand is a well-known tourist destination for Swedes and the disaster that hit on Christmas day 2004 therefore directly and indirectly affected thousands of Swedes.



Organizational Cooperation in Crises

and between departments in the Swedish government. Unfortunately, a list of similar organizational failures to coordinate and cooperate in response to disasters and crises could be made about any country. Challenges to Crisis Cooperation We want and expect actors to come together and cooperate in difficult situations, but we also know from prior experience that organizations often fall short of this expectation, and that cooperation under crisis conditions is particularly difficult to achieve. It is not only the case that organizations face many challenges to cooperation in reality but it is also the case that research with bearing on organizational cooperation in crises has been unable to provide much in the way of suggestions for how to make this type of interaction more successful. That is, how organizations can approach facilitating cooperation and avoiding conflict in high stress and high stakes situations. The international arena unfortunately abounds with examples of states, as members of international organizations or as independent multilateral actors, failing to cooperate in the face of serious challenges to peace and in humanitarian disasters. The EU member states’ failure to organize a coherent and effective response to the breakout of war in the Balkans in the 1990s is one example. More recently, the international community failed to respond to the humanitarian disaster in the Darfur region in Sudan. Members of the United Nations and NATO failed to agree on an intervention, and the situation for the population in Sudan has since proceeded to deteriorate. The Sudanese government’s response to the 2005 rebellion, using the Arab Janjaweed militia, left an estimated 70,000–350,000 dead and unknown numbers of displaced people (Smith 2005). In domestic crises like the Heizel Stadium disaster in Belgium (‘t Hart and Pijnenburg 1989), we have seen organizations with very similar goals fail grossly at coordinating, communicating and cooperating in the face of acute threats to life. The situation, where fans of rival teams clashed violently causing a large section of the stadium to collapse, was aggravated by the failure to cooperate among the two Belgian police forces in charge of security at the soccer game. In a more recent example, stories have surfaced about problems that arose between different levels of government in the aftermath of Hurricane Katrina in the United States  ������������������������������������������������������������������������� Scholars like Williams and Bellamy (2005) have suggested some underlying reasons for the failure to rally international cooperation in this case. Among these are; an increased skepticism about the humanitarian interventionism as a worthy goal (especially in the aftermath of the invasion of Iraq), states’ self-interested protection of strategic interests in Sudan, and link between the current crisis and links to other wars in Sudan. Some of the key problems in this setting seem to be the actors not sharing the same problem frame, acting on self-interest and there being linkages between issues and the stakes involved.

Introduction



in 2005. The city agencies in New Orleans, the Louisiana state organizations, and the US federal government had major problems coordinating not only the logistics of the rescue and recovery phases but also the official messages that were being communicated to the victims of the disaster (e.g., Alkan 2006; Comfort 2006; Kettl 2007). Even locally, where organizations have very strong incentives to work together and familiarity among the actors often is much greater than in national or international responses to crises, there are examples of gross failures to cooperate effectively. During the 1994–1997 outlaw-biker war in Scandinavia (Svedin 1998a, 1998b) many localities experienced waves of violence that threatened the stability of the very organizations set out to protect citizens from harm, such as the police, the judicial system, and local government. Over time, as the response by public officials in one city became more coordinated, the local reaction also became more effective and the outlaw biker gangs moved out of the community. However, as soon as the bikers officially moved to a neighboring city, all coordination activities in the first locale stopped, and the experiences and the response designed were never communicated to the new city of residence. As all these examples suggest, in light of these failings of crisis cooperation, it is important to better understand organizational interactions in crisis situations. Abundant but fragmented research on cooperation provides us with suggestions regarding why organizations face difficulties cooperating in the face of threat, urgency and uncertainty. Some of these factors are structural, such as the fact that increased security or safety for one actor can increase the risks and insecurity of other actors (Jervis 1978, 1985). One example of this would be the storage of nuclear waste, which has created both real and perceptual crises in several countries (D’Agostino 2003; Yoo 2003). Another example is the implementation of flood mitigation measures upstream on a flood prone river (Lamberte 2000; Rosenthal and ‘t Hart 1998; Svedin 2001; Ullberg 2001). A third example can be seen in the problems of organizational structure that often are experienced when civil and military or para-military organizations participate jointly in decisionmaking processes and implementation structures. Large amounts of red tape, different operational logics, or standard operating procedures can cause delays, friction and misunderstandings (Sundelius et al. 1997; Svedin 2001). Other obstacles to cooperation have to do with the processes in place with regard to how organizations interact and the impact of agency manifesting itself in the leadership of the crisis. One example can be seen in the difference between the responses in New York versus the Pentagon in the 9/11 attacks. In New York the response was highly decentralized with a large measure of individual agency leading to coordination and cooperation problems down the line, whereas the response to the attack on the Pentagon was considerably smoother and more coordinated in part due to joint inter-agency training prior to the crisis (Kettl 2007; Langewiesche 2005). Another example of international process failures that seriously hampered cooperation was the international response to the Estonia ferry disaster (Haspers 1998). There the processes set in place were only followed



Organizational Cooperation in Crises

by a number of countries engaged in the multinational rescue effort, and many procedures proved too cumbersome to work in a real life situation causing an additional secondary disaster. A clear example of how the animosity among leaders and organizations had the potential to greatly aggravate a crisis situation can be found with the Los Angeles Police Department (LAPD) Chief of Police and the Mayor of Los Angeles during the 1992 Los Angeles riots (Moment of Crisis 1993). Finally, many obstacles seem connected to the psychological reaction that the very nature of crises generate in individuals, such as a narrowing of perception, stress (Hermann 1979), and a decreasing ability to trust and to deal with ambiguity in the face of threat (Vertzberger 1990). The real reason why research has failed to make much headway with regard to organizational cooperation is the result of the theoretical assumptions and the biases in the approaches that this research has tended to rely on. A majority of related international relations’ and crisis management studies are heavily focused on conflict. As a consequence, the theoretical approaches most frequently utilized in research on cooperation, including crises cooperation, have unwittingly stacked the odds against finding solutions and generating practically actionable research (Barnard 1938). This has been a situation where, as the proverb suggests, if the only tool you have is a hammer, then everything you find is going to look like a nail. What we need to do instead is take a broader approach to cooperation and realize that cooperation everywhere is interspersed with and intricately connected to conflict. In fact, we are hard pressed to find one without the other. Cooperative motivations, strategies and behavior are constantly being balanced with, played off and interspersed with competitive and conflictual drivers, strategies and behaviors as organizations interact with one another in situations where the organizations are operating under pressure. If we take an integrative approach to the study of cooperation, viewing cooperation and conflict as integral and often co-existing complimentary parts of organizational behavior and strategies, we will be better able to understand the underlying dynamics of organizational crisis interactions, predict the conditions under which more cooperative or conflictual organizational behavior is more likely, and what measures may help strengthen cooperative sentiments and weaken competitive or conflictual sentiments. Theoretical Explanations of Challenges to Cooperation Common explanatory approaches to cooperation stem from international relations research, public administration as a field of study and social psychology. The first approach focuses heavily on the relationship between the actor and its environment. The second looks more to the structure of and processes involved in the interaction between organizations. The last perspective takes a serious look at how perceptions of the situation at hand and the interacting parties shape organizational responses in interactions.

Introduction



Relations to the Environment – Self-interest and Competition The first theoretical explanation is based on a view of the organization, group, and individual as autonomous, rational, and self-interested. Here ideas of the ‘economic man’ and the self-serving state or organization paint a picture of cooperation as only possible when actors perceive they can sufficiently gain from an agreement. In this view, cooperative outcomes are often thwarted or undermined by parties acting on short-term goals and being ‘envious’ of each other’s relative power and gains. Relations between organizations in this competitive environment are exacerbated when there are no rules and enforcement mechanisms for agreements. Crisis situations, more than any other organizational problem-solving situation, seem to generate ad hoc responses and create temporary groups or task-forces to manage the problem. It is hard to predict exactly which actors will be affected by a crisis and who may need to interact; crises rarely happen exactly where they are expected and seldom between 9am and 5pm. In fact, for many organizations they seem to occur when least convenient. A lack of previously established rules for decision-making and a lack of procedures for holding others to their commitments in ad hoc groups is similar to one-shot interaction games or temporary alliances in an overall competitive environment. Competitive environments have a potential to generate bureau-political behavior among organizations and organizational parts, particularly in the aftermath of a crisis. Bureau-political behavior, whether it occurs domestically or internationally, can be understood, as Halperin pointed out already in 1974, by focusing on the numerous constraints, conflicting interests, and priorities of the individuals that make up the organization or government involved. Arguing that we should not ask why the United States decided to do this or that, Halperin (1974) said that we would get farther if we asked what the various motives, and sources of power of the participants were that led to the decisions made. Decisions, he argued, while often portrayed as if they were made by one person are, in fact, almost inevitably the result of compromises (Halperin 1974). The struggles between individuals and factions of organizations in the aftermath of a crisis are often part of an effort to

 �������������������������������������������������������������������������� Game theoretic notions stemming from economics became influential in both political science and psychology by creating a basic notion of ‘the economic individual’.  ������������������������������������������������������������������������������� Game theory has by far spurred the most research on cooperation and failure to cooperate in strategic interactions, like those that take place in ad hoc inter-organizational groups. While game theoretical conceptualizations of cooperation are rather simplistic this tradition’s strength lies in the independent variables systematically tested through controlled experimental settings. Game theory, originally designed to solve problems in the sphere of economics, highlights self-interest, payoff structures, access to information, and prior behavior in repeated games as key in shaping actors’ strategies of cooperating or defecting.



Organizational Cooperation in Crises

secure the organization or organizational unit’s survival in the face of criticism and subsequent reprioritizations of resources. Institutional Design – Interaction Structures and Training The second type of explanation for the challenge regarding cooperation has to do with the design of the organizations involved and the structuring of their interaction. There are a number of ways that collaboration between organizations can be arranged, all of which have different strengths and weaknesses. While hierarchies are, comparatively speaking, strong in command and control, this type of relation between organizations in crises is uncommon and can still lead to problems of information collection and dissemination essential to organizational coordination. Scholars interested in inter-organizational relations have focused on exchange relationships and networks in order to better understand what shapes organizational cooperation. Organizational networks are particularly common where service provision organizations cater to a common set of stakeholders. For example, networks of NGOs, government agencies, and corporations providing a wide array of services to people who suffer from mental illness. Network arrangements are flexible but have the disadvantage of not having a very clear command or authority structure. The organizations engaged in the network are arranged laterally with equal power and with a great number of connections between each network member making the authoritative structure dispersed and rather flat. Because networks are flexible, the participant organizations’ goals and preferences tend to be interactive and changing, making cooperation fluid. Noted obstacles to cooperation in network structures are different dependencies built into the process of how the organizations participate and make their contributions to the network as well as the maintenance costs associated with these types of interactive networks (Kickert and Koppenjan 1997). Decentralized control structures, similar to that of inter-organizational networks, are especially fit for dynamic decision-making situations such as operational crisis management (Brehmer 2000; Pigeau and McCann 2000). However, these types of arrangements have been known to cause problems in collaboration between different levels of organizations. In these decentralized systems, individual decision makers only see their part of the decision-making process making it difficult to achieve coordination with overarching goals and to maintain the information flow from the bottom up (Hugemark 1991). Furthermore, this type of fast-response structure, where changes in local conditions can more easily be accommodated, tends to be driven by the immediacy of the situation. Prioritizing, focusing on  ��������������������������������������������������������������������������������������� There is a real risk that information in hierarchical systems remains, unprocessed, at the top of the organization due to so-called information overload in the crisis, thus leaving the rest of the organization without adequate information and direction Brehmer (2000) and Hugemark (1991).

Introduction



short-term goals, if the time pressure is great may encapsulate decision-making groups who find themselves overwhelmed on the frontline. An example of this was the Dutch government’s effort to get a grip on how many people had died in the building encapsulated by fire in the Bijlmeer airplane disaster. By publicly promising amnesty to any illegal immigrants living in the building affected by the fire, Dutch government officials attempted to solve the pressing short-term problem of defining the death toll in a disaster where it was very unclear who the victims were. The longer-term problem of being forced to grant a far greater number of amnesties than the building could possibly have had residents, was an unintended consequence of the government becoming wrapped up in the drama of the enfolding of events of the plane crash and the subsequent apartment building fire (Rosenthal et al. 1994). Organizational Psychology – Perceptions and Crisis Situations The third line of explanation for the challenge that cooperation poses focuses on the individuals that make up the cadre of organizations and how these individuals face challenges as they interact with others. The characteristics of crises – presenting threat, urgency, and abundant uncertainties – often trigger psychological reactions in people and organizations that make it harder for those same persons, unacquainted with one another, to trust each other. Stress tends to narrow the scope of attention and the capacity to process complex information (Hermann 1979; Holsti and George 1975). The stress experienced in crises may paralyze individuals or groups or lead to other self-defeating reactions such as denial and hyper vigilance (e.g., Hermann 1990; Miller 2001), making cooperation with others challenging or impossible. Sociological and socio-psychological approaches to cooperation have highlighted a number of factors that influence group cooperation. Some of these include individuals’ identification with what they perceive as their in-group (see Dasgupta 2004; Gaertner and Dovidio 2005; Grzymata-Kaztowska 2005; Yamagishi et al. 2005) and prejudice against groups with whom they do not primarily identify (so-called out-groups) (e.g., Shah et al. 2004; Shinada et al. 2004). These dynamics are heightened by perceived threats to the groups, which in turn promote group cohesion and a circling of the wagons to protect the group. Perception, social psychologists have shown, is key to understanding any type of human behavior and interaction, whether in groups, organizations, or as individual decision makers and leaders of states. In general, it can be said that this line of research shows how social bonds (Huddy 2004; Jost et al. 2004; Sidanius et al. 2004) and individual identity10 (Reicher 2004; Schafer 1999; Tajfel and Turner 1979) are central to social coordination, both in creating social conflict and in getting people to work together. 10 ���������������������������������������������������������������������������������� For a good discussion of competing relevant perspectives on this topic, see Rubin and Hewstone (2004).

10

Organizational Cooperation in Crises

The Crisis Interaction Setting A fourth line of explanation that ties directly into this study focuses on the influence the specific crisis setting has on challenges to cooperation. The main characteristics of crises – the perception of threat to core values, urgency, and uncertainty – are closely associated with a number of the psychological dynamics outlined above. Furthermore, the perceptual definition of crisis is key to understanding organizations’ responses. Crises are only crises when organizations perceive there to be a threat to core values, urgency, and uncertainty. The degree to which these organizations share perceptions held by the public or other organizations involved is going to be key to the management of the situation. Events may well turn into crises for organizations that do not pick up on the potential threat in a situation or the urgency with which it is affecting the organization. Perceptions of threat, of time-pressure, and of uncertainty explain a lot of organizational behavior and can prove fruitful in explaining actions in organizational crisis cooperation. There are, however, a number of other characteristics distinct to crises as interaction settings that pose challenges to organizations’ ability to cooperate. In crises any shortcomings in terms of planning, defining areas of responsibility, and division of authority become acutely obvious. One problem in the preparation phase, that is before a crisis takes place, is that it may be hard to predict which organizations are going to be affected. New constellations of actors adjusted to the nature of the crisis and ad hoc decision-making groups may make it difficult to prepare for cooperation with specific parties. The acute phase of the crisis often leaves organizations little room to renegotiate relations which can have detrimental effects on the outcome of the crisis. A recurring problem of this kind is related to the elimination of red-tape and bureaucratic hurdles surrounding the extension of police powers in crises (often needed in situations of mass-evacuations). In order to forcefully remove people from their homes and to protect their property with force one needs police powers, which in times of crises may be granted supporting organization like the National Guard or the fire department. Other illustrative examples are the legal and organizational problems related to the use of armed forces in the response to civil emergencies (such as floods, e.g., Svedin 2001). There may be good reasons during normal circumstances to have large hurdles in front of the use of military personnel in civilian affairs; however, when there is need to change these principles during crises, it is often quite difficult and timeconsuming. Training and experience may help decision makers and organizations detect crisis situations and to better organize collective efforts to deal with them. “Crisis management logic suggests that planning and preparing for crisis should be a vital part of institutional and policy toolkits” (McConnell and Drennan 2006, 59). Nevertheless, even at a time when crisis management is high on many countries’ political agendas, crisis preparedness, McConnell and Drennan (2006) argue, may be a virtual mission impossible. Crises are known as low probability – high impact events that do not always compete successfully with other everyday demands on

Introduction

11

resources. Further complicating the prospects for preparedness and training is the fact that “contingency planning requires ordering and coherence of possible threats” (McConnell and Drennan 2006, 59); crises do not readily lend themselves to this type of clear foresight. “[P]lanning for crisis requires integration and synergy across institutional networks, yet the modern world is characterised by fragmentation across public, private and voluntary sectors … robust planning requires active preparation through training and exercises, but such costly activities often produce a level of symbolic readiness which does not reflect operational realities” (McConnell and Drennan 2006, 59). Exploring Crisis Cooperation through Specific Research Questions The exploration of organizational cooperation in this study is guided by a set of research questions. The first question this study aims to answer is: How do organizations interact in crises in terms of cooperative and competitive behaviors and strategies?

There has been very little systematic study of cooperation in crises across a large number of cases. By engaging in such a study this research makes a quantitative contribution to our cumulative knowledge of how organizations interact under pressure. This first question is part of an effort to construct a conceptual framework11 with regard to organizational cooperation in crises. Consequently, in addition to specifying and providing an analysis of generic patterns of cooperation in crises, the results of this study also serve as a point of a departure for formulating hypotheses about what influences organizational interactions in crises. The outlines of this theory building effort are presented in Chapters 7 and 9. The second question the study examines is: How are cooperative and conflictual behaviors in individual decision-situations related to more overarching cooperation and competition strategies that stretch over whole crisis cases?

This question seeks to establish what the relationship is between choices and displayed behavior in individual decision-making situations, and the overall strategy that an organization pursues over the course of repeated interactions with other organizations. A large number of empirical studies of cooperation have been presented in game-theoretic terms. Game theory emphasizes strategic interaction in decision-situations. In order to maximize gain, an actor in a game-theoretic 11 ���������������������������������������������������������������������������� The distinction is made here between a conceptual framework and a theory or theory building is based on Sabatier’s (2007) discussion of frameworks, theories and models.

Organizational Cooperation in Crises

12

model needs a strategy for determining what to do (cooperate or defect) in each turn. These strategies may look different and steer the actions and behaviors displayed in individual decision-situations (turns in a repeated interaction game) as well as affect the overall payoff in a longer n-turn interaction. However, even if the strategy is just to do whatever it is the other actor is doing, the parties engaging each other are assumed to have a strategy. The second research question of this study examines this particular type of interaction to determine the relationship between observable behavior in individual decision-situations (which can be likened to interaction turns) and the overall patterns of cooperation strategy that can be identified at the case level of analysis. Finally, the study examines the question: How are the identified behaviors and strategies linked to the characteristics of crises, such as threat, urgency, and uncertainty?

The third and final research question explores the relationship between types of cooperative behavior and cooperative strategies and the specific characteristics of crises. The purpose is to see if types of crises (with variation on the crisis characteristics) are connected to particular types of behavior or strategies, i.e., are situations with perceived high threat but medium urgency more likely to see a certain type of behavior by organizations or does that type of crisis situation make it more likely that an organization will pursue cooperation strategy x? We can also think about this relationship in reverse terms, i.e., if organizations display x cooperative behavior or z cooperation strategy, they are likely to view the crisis situation as high threat but with little urgency. Does empirical variation in the crisis setting correspond to variation in behaviors and strategies on behalf of the organizations involved? Methods The first part of the present study applies a type of factor analysis called Categorical Principal Component Analysis (CATPCA). This factor analysis provides an empirical mapping of organizational cooperation across two levels of analysis; decision-making situations within a crisis and crises as wholes (for more details see Chapter 2). The results point to common clusters of variables that represent types of cooperation in crises and pin-point the contribution of individual variables to those clusters (see Chapters 4 and 5).12 The second part of the study examines the relationship between cooperation at the above noted two level of analysis using correlational analysis. The correlational 12 ������������������������������������������������������������������������������� Each empirical observation is given a numerical score on the cooperation types identified that are then used (as new continuous variables) in quantitative tests of the different types of cooperation.

Introduction

13

analysis provides a basis for the development of a number of hypotheses about the relationships between cooperation as behavior in decision-situations and cooperation as an overall crisis strategy (see Chapter 6). The third part of the study looks at links between the identified cooperation behaviors and strategies and the specific characteristics that make up the crisis setting. These relationships are investigated using correlational analysis and multiple linear regressions (see Chapter 7). Data The data are from the Transboundary Crisis Management (TCM) Project and were developed by the author together with Margaret Hermann and Bruce Dayton. The dataset contains information about crises and various aspects of the management of these crises that has been drawn out of in-depth qualitative case studies. The structure of the data and the dataset will be discussed in greater detail in Chapter 2. The vast majority of crisis research has been qualitative, typically in the form of detailed narrative accounts of crises and the management of these by a number of persons or organizations. Only a small number of researchers have been involved in examining crises and crisis management with any quantitative ambition.13 The dataset, TCM, to be described later allows us to systematically map and study crises from across the world, across time, and at different levels of analysis. The aim of developing this dataset was to make the variables and variable measures as fine grained as the empirical material in the case studies would allow. The dependent variable – cooperation – is measured at two levels, the occasion for decision level and the case level. Organizational cooperation in each crisis consequently has a number of indicators and is assessed, through the occasions for decision, at several points in time across each crisis. Outline of the Chapters The study is presented in the next eight chapters in the following manner. Chapter 2 discusses organizational cooperation in crises and proposes ways to operationalize this concept. This chapter also explains how the TCM dataset was constructed and how we can use it to look at organizational cooperation in crises. Chapter 2 ends with an introduction of CATPCA, the quantitative technique, used in Chapters 4 and 5. Chapter 3 introduces case examples of cooperation in crises as seen through 13 �������������������������������������������������������������������������������� The best known work of a quantitative type is the International Crisis Behavior Project set up by Brecher and Wilkenfeld (e.g., Brecher and Wilkenfeld 2007). Their event dataset on foreign policy crises takes a rational or objective standpoint that is different from the perception-based research usually done. Details of this research project and the dataset are available on , accessed 30 June 2006.

14

Organizational Cooperation in Crises

the lens of each of the perspectives informing the conceptualization of crisis cooperation in this study: international relations, public administration, and social psychology. Chapter 4 reports the first set of findings from the principal component analysis. This part of the empirical study looks at organizational behavior in decision-situations. It identifies and typologizes cooperative and conflictual behavior as different organizations interact at key points in the management of a crisis. Chapter 5 conveys the second set of findings of the principal component analysis. This chapter examines and analyzes a set of cooperation strategies that organizations pursue across whole crisis cases. Chapter 6 links organizational cooperation behavior and cooperation strategies in crises. The chapter particularly looks at what significant relationships there are between the empirical behavior and strategies. This following chapter, Chapter 7, examines the connection between the different organizational cooperation behavior and identified strategies and characteristics of crises such as threat, urgency, and uncertainty. Chapter 8 summarizes the major findings of the empirical parts of the study and highlights a number of conclusions. This chapter also addresses how practitioners may use the findings in this study to better meet the challenges of cooperation in crises. The final chapter, Chapter 9, introduces a number of ways in which the results of this study can and should be used to expand research on organizational cooperation in the future.

Chapter 2

Conceptualizing Organizational Crisis Cooperation: The Legacy of Three Traditions This book empirically examines organizational interactions in times of crises, taking as a point of departure three distinct traditions of cooperation research. The focus of this study is organizational cooperation strategies and behavior in a setting where the actors’ perceive core values to be threatened and there is a sense of urgency and uncertainty about the situation. Scholars of international politics and economics, administration and bureaucracy, and of social psychology have devoted considerable effort to figuring out when and how cooperation can be achieved. This interest in cooperation, however, has yet to enter into the study of crisis in any systematic way. The lack of cooperation research in the area of crisis research is related to a number of unarticulated biases and some methodological challenges researchers interested in crises have faced. This study bridges the gap between qualitative case research and event data by systematically looking at new quantitative data on crises and organization behavior. Conceptualizing and Examining Organizational Crisis Cooperation It makes better theoretical sense to stick to simple concepts and models that clearly describe the mechanisms behind any classified phenomenon. Bringing in factors and mechanisms from other fields of research tends to make the picture much more complex and the terminology in much greater need of definition.  ����������������������������������������������������������������������������������������� International relations scholars interested in inter-state relations and security in the international system, intergovernmental institutions, ‘common-goods problems’, gametheory and those scholars focusing on international political economy, particularly trade and, have a lot to contribute. In public administrations scholars looking at inter-organizational relations, or organizational networks, particularly with regard to service provision, put a strong focus on cooperation. Finally, psychological research that focuses on perception and stress in group decision-making, intra-group dynamics, like social cohesion, and individual attitudes and behavior, such as altruism and risk aversion, have laid some important ground stones for understanding cooperative group behavior. What unites these three research traditions is an interest in the interaction among, or exchange between, different actors/units even though their conception of the unit of analysis differs.

16

Organizational Cooperation in Crises

Interdisciplinary research is taxing on the individual researcher and requires a high tolerance of conceptual messiness (translating these separate understandings and making them fit together is neither easy nor glamorous). I argue, however, that an interdisciplinary approach toward any one research topic is the best way to capitalize on research investments already made, and it is the most effective way of accumulating knowledge of practical relevance. Apart from the problems associated with translating theoretical concepts and ideas, there have been empirical hurdles to comparing and merging theories on cooperation across academic fields. One problem is at the level of analysis problem. International relations scholars tend to focus on states’ interaction at the international level, public administration scholars examine organizational interactions within states, and social psychologists look at interactions between groups or individuals’ embedded in a larger context (be it states, civic society or organizations). The empirical data gathering of these scholars has naturally followed their line of interest and the results have not necessarily been comparable. Nor has there been much desire to test theories of one kind on data centered on another type of actor. I argue, however, that while methodologically sound, this approach may lead us to miss important clues to the main underlying dynamics of cooperation. In research on crises and crisis management, the empirical hurdles to comparison have been marked. This has not primarily been caused by a level of analysis problem, but rather by a diverse range of research foci. Most crisis scholars tend to be rather pragmatic and apply their ideas and efforts wherever there is a case to be studied and analyzed. Pragmatism on the one hand and enthusiasm over exotic cases on the other has turned out to be the core of the problem. Crisis research tends to be very case focused, providing ample empirical detail and accuracy for individual cases, but makes comparative work cumbersome. And while there are notable exceptions, crisis research is riddled with project specific crisis definitions that make up the basis of case selection, and every researcher seems to have a favorite set of lenses through which the case is analyzed. While scholarly diversity is a sign of a healthy discipline, this inconsistency has discouraged large-n and comparative studies in the area of crisis management. Conceptually, I attempt to overcome the level of analysis problems of cooperation research identified earlier by viewing organizations as dynamic largegroup actors. On the one hand organizations are the implementers and real life producers of state polices. As such, organizations, public and private, are the limbs of the state corpus in the provision of services to the public, both in their response to crises internationally and domestically. On the other hand, organizations consist of people who interact in a group setting. As such, organizations can take on and be shaped by the psychological characteristics of individuals as they interact in this large-group setting. While the concepts presented by the three research traditions still warrant some translation, this conception of organizations as the unit of analysis makes all three perspectives applicable and relevant.

Conceptualizing Organizational Crisis Cooperation

17

Empirically, the new MITCRA dataset forms a bridge between the existing qualitative description crisis case studies and the need for standardized measures of variables that will hold up to robust quantitative testing. The consistency with which one crisis definition has been applied as the selection criteria for the cases coded, as well as the great spread of variables coded, has been integral in overcoming the diversity in approaches that has challenged prior crisis research. Organizations: The Actors How can a unified conceptualization of the relationship between individual and recursive overlapping collectives – such as factions, groups, and organizations – be achieved? One answer may be found in contemporary social theory where individual behavior is seen as enabled and constrained by relationships with multiple, overlapping collective identities. Factions, groups, and organizations exhibit structural properties such as norms, rules, and roles as well as constellations of status, power and resources. Each are structures (looking ‘in-ward’ to their component agents) and agents (looking ‘outward’ to wider socio-political arenas). In developing the capacity for collective action, these structures exert important influences on individual members (and smaller scale collective agent) components. (Stern and Sundelius 1994, 104–5)

Organizations in this study are viewed as dynamic large-scale groups. They are dynamic in the sense that one organization may serve a multitude of functions for a multitude of interested stakeholders and act in a number of settings, on a number of issues, under the influence of several competing factors. Still it seems that “[t]he basic elements of organizations have remained relatively constant through history” (Shafritz and Ott 2001, 2). For example, “[o]rganizations (or their important constituencies) have purposes (which may be explicit or implicit), attract participants, acquire and allocate resources to accomplish goals, use some form of structure to divide and coordinate activities, and rely on certain members to lead or manage others” (Shafritz and Ott 2001, 2). Even though these basic workings of organizations remain constant over time, [organizations’] purposes, structures, ways of doing things, and methods for coordination activities have always varied widely. The variations largely (but not exclusively) reflect an organization’s adaptation to its environment, because organizations are open systems that are influenced by and impact on the world around them … Organizations are inseparable parts of the society and the culture in which they exist and function. Human behavior – and thus also organizational

 ����������������������������������������������������������������������������� An even simpler, but compatible definition of organizations has been used by Shafritz and Ott (2001, 1), “[b]y organization, we mean a social unit with some particular purposes.”

18

Organizational Cooperation in Crises behavior – is influenced by culturally rooted beliefs, values, assumptions and behavioral norms that affect all aspects of organizations. (Shafritz and Ott 2001, 2)

Organizations in this study are also viewed as a form of large group capable of a bounded rationality. This implies both a deductively reasoned (analysis of pros, cons and consequences), and in some sense mechanical, way of functioning (through standard operation procedures), as well as a propensity to be influenced by non-rational (i.e., social or psychological) factors in its operations. The most basic reason for this bounded rationality is that organizations consist of people. As such, organizations are limited to the same extent as humans by bounded rationality, error in judgment and miscommunication. In some circumstances they are also constrained by what international relations scholars describe as ‘egotistic’ and ‘envious’ motives, and public administration scholars call ‘bureaucratic turf battles’. However, because organizations share many of the characteristics of the people who make up their workforce, they are also capable of making decisions based on what is perceived to be the costs and the benefits of a situation. They can imagine multiple outcomes and futures, they can be ‘generous’ (i.e., cooperate when there is no direct pay-off), and can share cultures, values and even social capital with other organizations they interact with. As the institutionalization of organizations show an organization can also be more than the sum of its parts (the people who work in it and the rules that govern the organizations conduct). Organizations can survive personnel turnover, changes in the environment, even the threats to or the elimination of its original purpose and function. These facts in no way undermine the basic understanding that organizations cannot act or interact with other organizations separate from the people that represent them. The study incorporates a wide range of organizational types: public organizations (e.g., agencies, departments of government, local governing bodies), private organizations (e.g., corporations, non-profit organizations and non-governmental organizations), and ad hoc organizations (such as victims’ organizations in disasters or ad hoc coalition such as trade workers, organizations formed in response to a labor market conflict). This study does not consider states a unit of analysis (one organization). If a state is involved in a crisis, the particular branches of government involved will be the unit of analysis (i.e., the foreign ministry of a country, the embassy of a country, or the air force of a particular state). In the empirical parts of this study, organizations as actors are represented by people (as groups of individuals) in decision-making positions and streetlevel bureaucrats (Lipsky 1980) that act on behalf of the organization where the rubber meets the road of inter-organizational cooperation. The type of broad unit of analysis that is suggested in this study prompts the question of when does the organization ends, so to speak, and the individual begin? In this study individuals are viewed as representatives of organizations until their interaction with other organizations is motivated primarily by personal benefit or loss. For example,

Conceptualizing Organizational Crisis Cooperation

19

when an individual is acting in the capacity or role that she is assigned in the organization, she is a representative of the organization. However, when what the individual does in the interaction is not linked to her function or role as a member or employee of an organization then this is considered personal behavior. While this study employs variables derived from psychology that pertain to individual and group perceptions, the selection of these variables has been done in such a way that the available case study material provides adequate coverage. As a result, it and it has been possible to code for these perceptual aspects without researching the cases further. Crises: The Setting In this study, crises are defined as situations where decision makers perceive that there is urgency and uncertainty, and a threat to basic values. Among scholars interested in crises, there have been two main approaches to the definition of crisis. One view relies on ‘objective’ definitions, while the other relies on actors’ subjective perceptions and makes use of ‘subjective’ definitions of the situation. ‘Objective’ scholars view crisis as an observed change or a turning point in the policy process. This has close ties to the notion of a shift in the balance of power in international relations (e.g., Jervis 1985; Morgenthau and Thompson 1985; Walt 1988; Waltz 1979) and to the notions of punctuated equilibrium in public administration (e.g., Baumgartner and Jones 1993; Eckhardt 1986; Kingdon 1995). Among crisis researchers this approach is best represented by Brecher and Wilkenfeldt’s (1989) study of foreign policy crises. ‘Subjective’ scholars consciously focus on the subjective perceptions of involved actors as the starting point of crises. This approach uses decision makers’ perception of the situation as the measure and indication of crisis (Rosenthal et al. 1989; Stern 1999; Sundelius et al. 1997). Researchers taking this approach argue that we cannot understand or study crises divorced from the perceptions of those who are actually experiencing and/or managing the situations.  ��������������������������������������������������������������������������������������� This behavior will not be the focus of the study as it would limit influencing factors to only individual level variables which exist primarily in people’s heads and often have to be identified through in-depth personal interviews.  ���������������������������������������������������������������������������������������� Of the five most frequently cited definitions of crises two can be said to be situation based (objective). These definitions are 1) A crisis 1. threatens high priority goals of the decision unit, 2. restricts the amount of time available before the decision is transformed, and 3. surprises the members of the decision unit by its occurrence (Holsti 1972). A crisis 1. threatens important values of the decision unit, 2. restricts the amount of time available before the decision is transformed, and 3. and surprises the members of the decision unit by its occurrence (Brecher, Wilkenfeld and Moser 1988). The three remaining definitions, however, are based on decision makers’ perceptions (subjective); 3) A crisis poses 1. a threat to basic values, with a simultaneous or subsequent awareness of 2. a finite time for response, and of 3. the high probability of involvement in military hostilities (Rosenthal, Charles and Hart 1989). A crisis poses 1. a serious threat to the basic structures or the

20

Organizational Cooperation in Crises

Based on George (1991) the definition of crisis used in this study is perception focused. It stipulates that crises involve (1) a threat to basic values, (2) urgency, and (3) uncertainty. The three aspects of this definition can be explained the following way: Basic values include a country, organization, or individual’s sense of the intrinsic principles or qualities that are necessary for that country, organization, or individual to exist. These might include, for instance, secure national borders, honesty in accounting practices, and individual autonomy. All crises also come with a sense of urgency; that is, a finite time in which decision makers are able to respond to the situation – a window of opportunity in which to act. Finally, crises are by nature uncertain. In crisis situations decision makers often have an incomplete understanding of the origin and risks of the crisis and an uncertain understanding of the impact that their actions will have on alleviating or exacerbating it. Each of these elements must come into play in the selected case. (Hermann et al. 2003, 6, italics added)

Crises present a particular interaction setting for organizations that differs from ‘normalcy’. In a normal policy setting, interacting organizations may display one or several crisis characteristics. What distinguishes a crisis is that decision makers in these organizations perceive all three components to be present in a situation they are facing. A perceived crisis situation may present a pressing need for cooperation while at the same time presenting special circumstances for cooperation. The crisis setting contains characteristics that may both facilitate and constrain cooperation among organizations. The multitude of challenges to cooperation in crises, coupled with the fact that it is often both urgent and important for organizations to interact effectively in these situations, make crises a particularly salient setting in which to study organizational cooperation. Cooperation: The Object of Study Cooperation is the dependent variable examined in this book. Three disciplines have devoted considerable time and effort to examining issues of cooperation and conflict and these disciplines have immediate bearing on the object of this study. International relations scholars interested in inter-state relations and security in the international system, intergovernmental institutions, ‘common-goods problems’, game-theory, and those scholars focusing on international political economy, fundamental values and norms of a social system which 2. under time pressure and 3. highly uncertain circumstances, necessitates making critical decisions (Sundelius, Stern and Bynander 1997). A decision-making crisis is a situation, deriving from a change in the external or internal environment of a collectivity, characterized by three necessary and sufficient perceptions on the part of the responsible decision makers: 1. a threat to core values, 2. urgency, 3. uncertainty (Stern 1999; Stern and Sundelius 2002).

Conceptualizing Organizational Crisis Cooperation

21

particularly trade, have a lot to contribute to the conceptualization of cooperation. Likewise, in public administration, scholars looking at inter-organizational relations, or organizational networks, particularly with regard to service provision, also focus on cooperation. Finally, psychological research that focuses on perception and stress in group decision-making, intra-group dynamics, like social cohesion, and individual attitudes and behavior, such as altruism and risk aversion, have laid some important ground work for understanding cooperative group behavior. While their conception of the unit of analysis differs what unites these three research traditions is an interest in the interaction among, or exchange between, different actors/units. Cooperation has been studied by scholars in these fields both as the dependent and as the independent variable. The research traditions differ, primarily, in their selection of the actor/units of analysis, how they view the relationship among the units of analysis, and in their conception of cooperation as primarily a behavior, goal, process, strategy, or a forum/institution. Cooperation is variously examined as attitudes/behavior of the organizations/actors, as the goal to be achieved through the actor interaction, as the means by which to reach some stated goal, and at other times as the institution in which interaction and exchanges take place. It is important to provide the theoretical overview below for two reasons. First, research on cooperation differs between international relations, public administration, and social psychology. It is important to present the definitions and operationalizations used in their original theoretical context, since these definitions and operationalizations later will be applied to two specific contexts; organizations as actors and crises as the interaction setting. These are actors and settings for which the theoretical concepts were not originally designed and, some may argue, to which they should not be applied. I will attempt to show, however, how the merging of these ideas and operationalization is not only appropriate but also the most fruitful way to get at the object of this study. Second, some of the assumptions and simplifications that scholars in these three fields make will necessarily be carried over into the operationalizations of this study. By illustrating where and why these simplifications were originally made, I hope that it will be easier to understand why the operationalizations of cooperation, at the two levels of analysis, in this study look the way they do. Cooperation in international relations The contribution by international relations (IR) scholars to our conception of cooperation as a dependent variable comes primarily from those interested in international security, intergovernmental relations, common-goods and collective action problems, and international trade. Some of the ways these IR scholars have conceptualized cooperation are as: regimes (e.g., Haggard and Simmons 1987; Krasner 1983; Young 1989), collective security mechanisms and collective action within the UN Security Council (Characky 2002), compliance with organizational rules and norms (Carruth 1985), economic sanctions as coercive cooperation (Martin 1992) and the use of trade dispute settlement mechanisms (Gowa

22

Organizational Cooperation in Crises

1995; Gowa and Mansfield 1993; Sherman and Eliasson 2006), trade openness (Rogowski 1989; Roy 1999), economic interaction (Deng 1998), reduction of tariffs (Copeland 1987), covenant relationships (Husted 1990), and integration within super-national structures like the EU (Switky 1995). Cooperation in public administration Research pertaining to inter-organizational relations, networks, international collaboration ventures, alliances or partnerships, in the field of public administration (PA), is directly relevant to our operationalization and understanding of cooperation in crises. Inter-organizational cooperation as a dependent variable has been characterized by these scholars, for example, as social integration within and across organizations, organizational and social connectivity (Wu 1989), interorganizational linkages (Patterson 1993), partnership and joint decision-making (Nakayama 1999), processes and procedures such as standardization and mutual adjustment (Johann 1983), interagency coordination and comprehensiveness of service delivery (O’Brien 1996), joint action (Ashman 1999), collaboration (Johnson 2003; Mandell 1999; Powell 1996), integration and coordination services (Isett 2001), consensus and degrees of collaboration (Scanlon 1990). Cooperation in social psychology Finally, psychological research focused on perception, stress, intra-group dynamics, risk aversion, and altruism contributes substantially to our conceptualization and understanding of cooperative group behavior. There is a multitude of ways in which psychological research has conceptualized cooperation. Psychologists who have looked at cooperation as the dependent variable have conceptualized it as, for instance, team collaboration (Date 1987), altruism (e.g., Einolf 2006; O’Gorman et al. 2005; Shinada et al. 2004), cohesion (Drescher et al. 1985), reciprocity (Hayashi et al. 1999), cooperative behavior and helping (Grzelak and Derlega 1982), prosocial behavior (Saroglou et al. 2005), social coordination (Galinsky et al. 2005), social capital and generalized reciprocity (Glaeser et al. 2002; Putnam 2000; Putnam et al. 1993), voluntary participation in community improving projects (Ryan et al. 2005), ‘irrational’ cooperation resisting the temptation to free-ride (Goldberg et al. 2005), trust (Raub 2004), generosity (Haley and Fessler 2005) in

 ����������������������������������������������������������������������������������� There is also quite some research done on the same issues, looking for instance at collective action problems (e.g., Petee 1986; Peterson 1992) and political alliance building (e.g., Choi 2001; Sager 1999) but outside of the international setting. This research can be said to fall into a general political science genre of studies looking at cooperation between political parties and citizens at a grassroots level (e.g., social movements and resource management regimes) (e.g., Thompson 1997). Cooperation in and of itself, has not been a central focus in mainstream political science research rather it has been the actors or the social issues at hand that have spurred an interest in cooperation. Therefore, we will not go into more depth about how these researchers conceptualize cooperation.

Conceptualizing Organizational Crisis Cooperation

23

economic ‘games’, and social capital in the form of voluntary cooperation in game theoretic models (Poulsen and Svendsen 2005). Conceptualizing Organizational Cooperation Even though coordination and cooperation have been main concerns of organization scholars across the twentieth century (Scott 2001, v-vi), organizational cooperation remains a phenomenon that has defied unified definition. One possible explanation for the seeming inability of scholars to capture and explain this type of interaction is the propensity to treat cooperation as a one-dimensional concept. While the three research traditions vary in how detailed they are in their conception and definition of cooperation, a trend that cuts across all three is a persistent focus on what influences cooperation (independent variables) rather than on the nature and concept of cooperation itself (as the dependent variable). Game theory provides an elaborate and prolific rational choice approach to the study of cooperation and conflict in international relations and psychology. Yet for all its concern with cooperation game theory provides little guidance as to an operational definition of the concept. The outcome of a game or an interaction in this line of research is often presented in binary terms; either the party decides to cooperate or it defects. There is little elaboration, however, of what real life operationalized ‘cooperation’ would entail in terms of expressions or behavior. Research on the narrower topic of cooperation in crisis situations has been pursued in a smaller set of subfields: foreign policy analysis, public policy analysis, and business administration. Like the study of cooperation in general,  ������������������������������������������������������������������������������������ Game-theorists with an interest in the psychology of strategic interaction focus on what effects individual-to-individual interaction has on the likelihood of cooperation (e.g., Goldberg, Markoczy and Zahn 2005; Haley and Fessler 2005; Kanner 2004). These scholars take a rational approach to the study of individual behavior. Social identity theorists look at inter-group social competition and discrimination (e.g., Reicher 2004; Rubin and Hewstone 2004; Tajfel and Turner 1979) and social cohesion and common goods scholars look at groups with the broader societal context (e.g., Sampson 1991; Wydick 1999). In general it can be said that this type of research often see social bonds (whom we feel close to) and individual identity (whom we see ourselves as) as key components of social coordination and of getting people to work together.  ��������������������������������������������������������������������������������������� Up until the end of the Cold War, theory and research on crisis management in general, and crisis cooperation in particular, was heavily influenced by its roots in international relations, security studies, and foreign policy analysis. Most of the work from this time deals with crisis cooperation is focused on foreign policy crises (e.g., Allison 1971; Brecher and Wilkenfeld 1989; Hermann 1972) and how to avoid war (e.g., George 1991; Holsti 1972). This is also the area of the crisis management scholarship where most of the theoretical meat lies, drawing heavily on international relations theory, game theoretical models and the negotiation/bargaining literature. Today, the importance of cognition, psychology in crisis decision-making, and framing are central to the study of crisis management (e.g., Stern 1999; Verbeek, 1994) and crisis cooperation (e.g., Kydd 2000; Maresca 2003; Miller

24

Organizational Cooperation in Crises

the study of crisis cooperation has focused more on causes of cooperation than on operationalizing the concept of cooperation. To the extent that scholars have looked at cooperation as the dependent variable, the prevalent focus has been on pathologies in cooperation (e.g., ‘t Hart et al. 1997) and the failure to cooperate in crisis situation (‘t Hart and Pijnenburg 1989; Gerodimos 2004), rather than on successful cases of cooperation. Hence, while the broader aim may have been to improve the management of crises and improve cooperation, much of the attention has been spent on what does not work, examples of worst practices, and the implications of poorly managed cases (Brandstrom and Kuipers 2003; Shiels 1991; Weick 1998). This leaves the question of defining cooperation in crisis open. In international relations research, cooperation has also been treated rather simplistically. Some would argue that the one of the main contributions of international relations theory to the study of cooperation is a consensus around a definition of cooperation. This consensus definition, originally introduced by Lindblom (1965), states that cooperation occurs “when actors adjust their behavior to the actual or anticipated preferences of others, through a process of policy coordination” (Lindblom 1965 quoted in Milner 1992, 467). This definition has subsequently been used by social psychologists doing research on this topic (Milner 1992). It is worth noting, as Milner (1992) does, that “[a]lthough a commonly accepted definition of cooperation seems to exist in the literature, this has not eliminated the problems of identifying events as cooperative or not. It has, however, reduced argument over the concept and allowed hypotheses about cooperation to be developed” (Milner 1992, 470). Turning to some of the international relations scholars who have explored cooperation empirically, the operationalizations of cooperation they use still leaves much to be desired. Martin (1992) defines cooperation as simply “any joint activity among states” (Martin 1992, 10), identifying the three sub-types of cooperation as ‘coincidence’, ‘coercion’, and ‘co-adjustment’ (1992, 17).10 Building off Young’s (1989) work on regimes Milner (1992, 469) is more explicit and asserts that cooperation can be tacit, negotiated, or imposed. Tacit cooperation takes place without communication and explicit agreement. The prisoners’ dilemma 2001) as are various organizational design factors (e.g., Brehmer 1991; Feridhanusetyawan 2001; Haney 1997; Harris 2000; Hatzenbichler 2001; Joyce 2000; Stern 1999).  ����������������������������������������������������������������������������� As examples she cites Parsons (1951), Deutsch et al. (1967), and Marwell and Schmitt (1975).  ������������������������������������������������������������������������������� Martin’s reason for using a broad definition is that she wishes to explore and distinguish between different types of cooperation problems as she develops her framework (Martin 1992). 10 ����������������������������������������������������������������������������� Martin’s types of ‘cooperation problems’ are really categories of games that are based on the equilibrium outcome of each game. Martin views these three types of cooperation situations as mutually exhaustive and exclusive, but ideal types. She recognizes that real-world cooperation often looks different, combining the dynamics of several games as reality often involves more than two actors in any one interaction.

Conceptualizing Organizational Crisis Cooperation

25

game epitomizes this type of cooperation. Negotiated cooperation emerges out of an explicit bargaining process. This type of explicit cooperation, consisting of negotiated policy coordination and agreements, is somewhat easier to identify than tacit cooperation. Milner describes imposed cooperation as a situation where “[t]he stronger party in the relationship can force the other side to alter its policies. If the stronger party also adjusts its own policies and attempts to realize mutual gains, cooperation has occurred” (Milner 1992, 469). That a strong actor or organization would be able to impose cooperation on another organization does not automatically fall within what we would think of as cooperative behavior, yet Milner argues that “as long as mutual policy coordination to realize joint gains occurs, then it is cooperation by our definition” (1992, 470).11 Cooperation in this study is viewed as an umbrella term for a range of organizational behavior extending from cooperative to competitive or conflictual. The work of a number of scholars suggests a simplistic theoretical treatment of cooperation as one-dimensional that does not correspond to empirical reality of this same phenomenon. Based on iterated game theoretical experiments, Axelrod and Keohane (1985), conclude that “cooperation can only take place in situations that contain a mixture of conflicting and complementary interests” (Axelrod and Keohane 1985, 226). Likewise, while reviewing organizational behavior and group decision-making, Stern and Sundelius (1994) point out that “[w]hile some scholars regard conflict dynamics (such as ‘bureaucratic politics’) as largely incompatible with premature concurrence-seeking (classic groupthink), cohesion and loyalty within the group make take a variety of forms in part depending on the co-existing conflict dynamics” (Stern and Sundelius 1994, 104). Referring specifically to the work of ‘t Hart (1991) they argue that “one of [his] major contributions … is his exploration of the relationships between cohesion and conflict as sources of groupthink tendencies” (Stern and Sundelius 1994, 104). The mix of cooperative and conflictual elements in group interactions is particularly salient in situations where the psychology of the group comes to play a key role. “Recent literature … suggests that manipulation of group decision processes may involve a subtle mix of cohesion and conflict” (Stern and Sundelius 1994, 104). I argue that, as with the study of decision-making (‘t Hart 1991, 250), operationalizing cooperation requires a multi-theoretical approach. Behavior (actions and statements) can be cooperative in three senses. Each of the three senses (ways of viewing cooperation) is more prominent in one of the three research traditions (international relations, public administration, or psychology), than in the others. The first sense of cooperation refers to actions which are combined, concerted, conjunct, conjunctive or united. This sense of cooperative behavior involves “the joint activity of two or more” (Synonym.com 2001–2007) 11 ��������������������������������������������������������������������������������������� It is also worth noting that in this view of cooperation, based on mutual gains, there is no conception of how these gains will be distributed among the actors or organizations. Fairness or equality, does not figure into the concept of cooperation in any predetermined way (Milner 1992).

26

Organizational Cooperation in Crises

organizations. The catchword here is joint as opposed to separate. The second sense of cooperative behavior refers to actions “done with or working with others for a common purpose or benefit” (Synonym.com 2001–2007). The lead words describing this type of cooperative behavior are collaborative, helpful, and synergetic. The third sense of cooperative is synonymous with accommodative, i.e., “willing to adjust to differences in order to obtain agreement” (Synonym.com 2001–2007). The signifying word in this sense of cooperative is noncompetitive, the opposite of competitive. Uncooperative behavior refers to either being unwilling to cooperate, or to be intentionally unaccommodating (disobliging) (Synonym. com 2001–2007).12 In this study, cooperative organizational behavior is behavior that aims toward a common goal or objective. This activity may be joint (first sense of cooperative) or it may be initiated and performed largely by one party (third sense). Cooperative behavior is also more or less facilitative and altruistic (second sense). In contrast, conflictual organizational behavior is behavior that aims toward divergent goals or objectives. This activity may also be joint or more unilateral (first sense of cooperative). Conflictual organizational behavior can be more or less confrontational and coercive (second sense of cooperative above). Operationalizations at Two Levels of Analysis: Cooperation as Behavior and Strategies There are two ways in which we can think of cooperation: cooperation as a behavior (the actual act of cooperating), and cooperation as a strategy (a way to get to the actual acts of cooperating). A strategy is viewed here as a likely way of acting, based on a roadmap of how to reach a goal one has in mind. A strategy provides strong incentives and rationales that influence and shape organizations’ behavior in interaction with others. Two ranges of cooperation indicators constructed below seek to capture the dependent variable, cooperation, in decision-making settings and across whole crisis cases. The first range, covering cooperative behavior in decision-situations, is more focused on output and the act of cooperating, whereas the second range, depicting cooperation strategies across whole cases, is more focused on how one gets there, i.e., to a point where one may or may not act cooperatively toward another organization in specific decision-situations. The strategies on the second scale may make certain actions in scale one more or less likely.13 Under the 12 ������������������������������������������������������������������������������������ Collaboration I understand as co-action, i.e., the act of working jointly. I do not include the reference to collaborating traitorously with an enemy here, i.e., quislingism. An action that is collaborative is an action that is accomplished by or through collaboration (Synonym.com 2001–2007). 13 �������������������������������������������������������������������������������� While it makes sense theoretically, it is not necessarily true that behavior in individual decision-situations lead to an overall strategy at the case level or the opposite, that an overall case cooperation strategy produces certain cooperative behavior at the

Conceptualizing Organizational Crisis Cooperation

27

umbrella term of ‘cooperation’ we can distinguish a number of indicators and practical expressions of cooperation (statements, behaviors, types of decisions). Cooperative Behavior at the Decision-occasion Level The first scale, focused on cooperative behavior in decision-situations, draws heavily on the Conflict and Mediation Event Observations (CAMEO) variable set. This set of cooperation and conflict operationalizations was developed by Gerner14 and Schrodt (e.g., Gerner et al. 2002) for the study of third party mediation in international disputes.15 The reason for using CAMEO in studying organizational crisis cooperation in decision-situations is three-fold. First, out of the limited number of quantitative coding frameworks on cooperation and conflict that exist, Gerner and Schrodt’s operationalization and variable identification for cooperative and conflictual behavior and statements lies closest to what this study aims to capture in decision-situations.16 Second, the repeated testing of the World Event Interaction Survey (WEIS) operationalizations that Gerner and Schrodt developed make the variable descriptions and overall operationalization uncommonly well defined and ‘calibrated’.17 Third, Gerner and Schrodt had the explicit ambition of decision-occasion level, is true. The relationship between what is termed cooperation strategies and behavior in individual decision occasions is really an empirical question that will be explored as part of this study. 14 ������������������������������������������������������������������������������� Deborah Gerner unfortunately died while this study was written, but I did have the opportunity before she passed away to share with her and Phil Schrodt in person what I was doing in this study and how I was drawing on their extensive work. Dr. Gerner was excited to hear about the coding project and my dissertation topic and we had a very fruitful discussion about quantitative coding in social science research together with a group of graduate students on a panel on methods in social science research at APSA in 2005. Her professionalism, her research tenacity and helpfulness to graduate students continue to inspire me as a scholar and participant in the academic community. 15 �������������������������������������������������������������������������������� The decision occasion variables are based on Appendix 1 in Gerner et al. (2002) and on Gerner and Schrodt (2003). 16 �������������������������������������������������������������������������� Other coding frameworks that have been examined and discarded in favor of CAMEO are the World Event Interaction Survey (WEIS), the Conflict and Peace Data Bank (COPDAB), and Behavioral Correlates of War (BCOW). 17 ������������������������������������������������������������������������� Gerner and Schrodt have worked with the Kansas Event Data Survey (KEDS), an automated event coding and data analysis program, since the early 1980’s. The operationalizations of cooperation and conflict in KEDS were based on McCelland’s World Event Interaction Survey (WEIS) coding framework. The KEDS coding and analysis of event data has been adopted and applied by many other researchers for other research projects (e.g., Goldstein and Pevehouse 1997; Huxtable 1997; Pevehouse and Goldstein 1999; Wood 1998). Over the years an number of problems with the operationalization and coding of cooperation and conflict behavior based on WEIS became apparent and in 2002 Gerner and Schrodt launched a new operationalization framework, the Conflict and Mediation Event Observations (CAMEO), and an improved automated coding program (TABARI) (Gerner et al. 2002).

28

Organizational Cooperation in Crises

including non-state actors which makes the CAMEO operationalizations better suited to studying organizational cooperation than exclusively state focused operationalization frameworks such as Behavioral Correlates of War (BCOW). The variables used in this study to observe cooperation behavior in decisionsituations is an adaptation of Gerner and Schrodt’s CAMEO variables (2003) with some adjustments for the context of organizational, rather than state, cooperation. The variables that have been placed in the range of decision-situation behavior (see Table 2.1) follow the main categories of cooperative and conflictual behavior and statements that Gerner and Schrodt have coded for. The range below does not, however, include the subcategory specifications that are included in the CAMEO-coding.18 For each decision-occasion established within a crisis case the following indicators are used to identify cooperative organizational behavior involved (through representatives that showed cooperative behavior). Table 2.1  Operationalization of cooperative behavior at the decision-occasion level of analysis ranging from strongest (yield) to weakest form of cooperation (structural violence) Yield Make/come to agreement Request/propose Decide to cooperate Express approval Consult/discuss Comment on Make demand Express disapproval Reject Threaten Reduce relations Use structural violence

• to give in, give up or to cede to another • agreements being struck, signed or made • suggest through active positive contact • decide to expand or intensify contact • expresses support or approval of another • meet to talk about what is happening • passive verbal communication • expresses coercive suggestion • indirect negative communication • active disapproval though direct communication • passive coercion through threat of action • active coercion through cutting contact points • use of force directed at another’s stakeholders

Yield Did one or more organizations represented in the decision units actually yield or concede to another actor? To yield in this context means to give way, give up, defer, give in, or cede something to someone else. To yield refers to an organization actually doing/performing the act of giving in or conceding, which is different 18 ����������������������������������������������������������������������������� The subcategories were coded for and included in the MITCRA-dataset but were not included in the study because of the large amount of missing data on these variables. There was simply not enough detail in the qualitative case studies to code for the more specific categorization of behavior that the subcategories represent.

Conceptualizing Organizational Crisis Cooperation

29

from simply stating that one gives in or gives up or makes future commitments/ agreements/promises to yield. Make/come to agreement This indicator refers to agreements being struck, signed or made between one organization and another. Agreements typically involve at least two parties and are reciprocal, but may also occasionally be used to refer to unilateral concessions (i.e., not reciprocal) or future commitments. Request/propose In the specific decision-situation, did one or more organizations request, propose, suggest or appeal to one or more outside actors? This indicator refers to a positive active contact where one organization is reaching out in an effort to engage another in a positive manner. Decide to cooperate This indicator refers specifically to a decision being made between two or more organizations to actively cooperate by initiating, resuming, improving, or expanding formal relations or relationships with each other. This can be seen as the act of approving, that is, to lend credibility and support to another by cooperating with that actor. Express approval This is the expression of approval of another organization, which can be said to be a weaker form of cooperation than deciding to cooperate. Expressing approval, in contrast to acting in a way that shows approval, is easier and less committal than the active doing referred to above (decide to cooperate). The key for this indicator is whether an organization expresses support or approval of or to another party in the crisis. Consult/discuss Were there any consultations/meetings between the organizations involved in this decision-occasion regarding the crisis? This indicator refers to the active communication between two or more organizations regarding the situation. Comment on This indicator refers to passive communication, i.e., indirect communication through commenting. The behavior involves an organization making a verbal statement concerning the crisis or situation to outside stakeholders such as the public or the media. Make demand Does one organization issue orders, commands, and/or decrees directed at another organization, rather than mere requests, suggestions, or appeals? This indicator

30

Organizational Cooperation in Crises

addresses more assertive suggestions where one organization is using power to persuade or force another organization to act a specific way. Express disapproval This indicator gets at indirect negative communication. Does one organization in the specific decision-situation criticize, denigrate or denounce another organization (in a verbal statement)? Reject Does an organization reject actions, statements, regulations and/or norms of another organization? This is the active form of disapproving of another organization through direct communication. Threaten This is an indicator of more intense conflictual behavior than making a demand. Here one organization verbally expresses threats or coercive warnings toward another in a specific decision-occasion. This is another passive form of communication which expresses considerable coercion through the threat of action. Reduce relations Here the organization actively pursues conflictual behavior by reducing normal or routine relations with another organization. When relations are reduced the possibilities of new cooperative action becomes more difficult as channels of communication between the organizations become inactive. Use structural violence This indicator refers to the most conflictual behavior we expect to see between organizations in decision-situation, i.e., an organization government-sponsored oppression and/or violence against the rights, property, or persons of another organization’s stakeholders.

Table 2.2  Indicators operationalizing cooperation at the case level of analysis Voluntary concessions

• giving something positive or giving up something valuable of own free will without being pressured by the other in order to help another Facilitating others’ goals • actively working to help another group reach its goals Facilitating in-group’s goals • actively working to help one’s own group reach its goals Coordinate • to determine and discuss whether one’s own plans and actions are in line with the policies and goals of others, and to work toward a policy adjustment Rally around the flag • a general concurrence seeking within a larger in-group, such as one’s country, and a lack of in-group criticism in the face of a perceived common threat Groupthink • excessive concurrence seeking leading a group to quickly reach a consensus around a decision, rather than critically examining many aspects of the decision problem New group syndrome • an unwillingness to take an active and critical role in preparing decisions stemming from being new members of the decision group; being unsure of one’s role in the group and of one’s relative competence compared to that of others in the group Display honesty • show through actions that one is being forthcoming with information and telling the truth Claim honesty • stating that one is being forthcoming with information and telling the truth Need for contact person • A lack of contact or communication between actors which raises the need for a liaison Credit-seeking • trying to receive credit for something widely regarded as a positive aspect of a situation, regardless of whether credit is due Stalling implementation • actively working to undermine a specific decision or policy by delaying its execution Content slippage • actively working to undermine a specific decision or policy by changing the intention or parts of the content of this decision or policy when implementing it Game of Old Maid • trying to avoid responsibility for a decision or a situation Blame-game • actively trying to avoid being blamed for negative aspects of a situation by spreading the blame around on people, places, and things Attributing fault to others • specifically trying to pin blame on a specific actor(s) Covert behavior • actively, but covertly, seeking information that can be used against another actor, actively, but covertly, working against another actor without a specific target Break agreement • intentionally and overtly failing to live up to a commitment or agreement made with another actor

32

Organizational Cooperation in Crises

Cooperation Strategies at the Case Level The operationalization of cooperation strategies across crisis cases that I have developed below (see Table 2.2) draws on a combination of cooperation conceptions in international relations, crisis research, and psychological research on pro-social behavior. In this study, a cooperation strategy is viewed as an abstract principle that guides the actions of one actor toward another. As such, the cooperative strategy acts as a roadmap for the actions of the organization in its interaction with others, making highly cooperative (or altruistic) or highly competitive behavior more or less likely. These abstract principles, or what I refer to as strategies, have been operationalized into a set of variables or indicators presented below (see Table 2.2). As part of a larger collective, members of an organization react to the surrounding environment and the evolving situation they have to manage. Likewise, these indicators follow how organizational representatives perceive and react to what others are doing, and how their perceptions of others’ actions affect their wants and goals. This operationalization not only captures how members of organizations may use the organization as a tool for exerting power. It also captures how organizations as structures limit individuals’ efforts to abuse power and reach their own goals. The different indicators of cooperation strategies are italicized in the running text to make it easier for the reader to follow. Voluntary Concessions and Cooperation in Harmony In psychology and sociology, cooperation and helping are often termed and viewed as forms of pro-social behavior (e.g., Staub 2003; Wispé 1972). The pro-social aspect is tied to the notion that cooperation and helping increase positive outcomes for other people (Grzelak and Derlega 1982, 2). Cooperative behavior, in this sense, is “behavior that maximizes both the individual’s and others’ interests whether the situation involves correspondent or noncorrespondent interests” (Grzelak and Derlega 1982, 3). This behavior can be either endocentrically motivated, i.e., acting for other’s benefit without external encouragement or reinforcement, or it can be exocentrically motivated, i.e., motivated by expected or actual rewards (Grzelak and Derlega 1982, 4). Altruism is an extraordinary form of helping behavior that squarely falls within the category of endocentrically motivated behavior.19 Altruism, or acting for others without encouragement or reinforcement, by its very nature tends to be unilateral 19 �������������������������������������������������������������������������������� There is considerable disagreement among psychologists and sociologists whether there is such a thing as ‘true altruism’. Even if it is theoretically possible to distinguish between endocentrically and exocentrically motivated behavior the line between these categories empirically is rather fluid. There is often a perceived or real personal benefit or interest in performing any type of cooperative behavior, altruism included, such as self-gratification, enhanced self-image or a belief in the rewards of setting an example for

Conceptualizing Organizational Crisis Cooperation

33

rather than joint cooperation (the first sense of cooperative presented earlier). There is, however, reciprocal cooperation that is also altruistic. “… there are many situations where both positive and negative reciprocal behavior is observed that cannot be explained in terms of strategic and far-sighted self-interest” (Diekmann 2004, 491). As an example of altruism and altruistic reciprocity, Diekmann describes how “a person acts favorably toward a stranger, and the stranger reciprocates the favor although it is unlikely that they will ever meet again. Also the likelihood that friends, colleagues, and relevant others will hear about the act of charity is zero” (Diekmann 2004, 491). And while social identify theory (e.g., Jost et al. 2004) suggests that attraction to the other party is an underlying cause of this type of cooperative behavior, Diekmann (2004) contends that this does not provide a solid explanation for altruistic reciprocity: “subjects who received a favor were strongly inclined to return the favor, while positive feelings toward the donor were of much less importance” (Diekmann 2004, 491). It therefore seems that altruistic actions introduce strong incentives for the receiving party to reciprocate. In both everyday observations and experimental research settings, Diekmann (2004) has shown that altruistic reciprocity exists and is something beyond what economists and game theorists have been able to account for other than in repeated interactions (Diekmann 2004, 487, 491).20 Poulsen and Svendsen (2005), for instance, have looked at how individuals’ own social preferences can generate social capital that supports cooperation. Their study demonstrated how individuals with a particular set of social preferences generate a maximum of social capital in any interaction where they know enough (above a critical threshold) about the other party’s preferences. In this study, when an organization acts cooperatively toward another in a crisis without any obvious exocentric reinforcement, I term this voluntary concession. In this situation an organization voluntarily makes the attainment of others’ goals its own goal, without any reward expected in return. The voluntary aspect comes from the fact that there are no obvious external pressures or encouragements that act to persuade or cajole the actor to help another. The concession part of this indicator of cooperation comes from the fact that the organization cares about what it is doing. It is no feat to cooperate on, or give up, something to which no value is others. Taking this into account some scholars argue that we as humans never really act in pure altruism. 20 ������������������������������������������������������������������������������� “Although there is a long tradition in sociology of research on reciprocity, a systematic theory leading to empirically testable predictions is still lacking. In economics and game theory, on the other hand, reciprocity was incorporated into rigorous models and has given rise to a more coherent theoretical perspective. However, the standard economic approach has the weakness that it cannot account for altruistic reciprocity or compliance to reciprocity norms in unrepeated interactions. Yet the field studies and experimental studies alike support the existence of a norm of reciprocity for a wide array of social activities and even among strangers” (Diekmann 2004, 487).

34

Organizational Cooperation in Crises

attached or which involves no personal cost (such as work or resource reductions). In this study, concessions imply that the organization is giving up something or doing something that, all things being equal, it would rather keep to itself or not do were it not for the fact that it chose to cooperate. Voluntary concessions, as indicators of cooperation, suggest that the organization is actively choosing out of its own motivation to help another actor without any expected reward. Facilitating the Attainment of Own and Others’ Goals when there is Discord Keohane (1984, 54) distinguishes between discord and harmony in the relationship between actors (states). When actors are in harmony, their policies automatically facilitate the attainment of others’ goals. In this state, the situation is, in essence, uncomplicated, apolitical, and does not warrant adjustment on anybody’s part (Keohane 1984, 53). It is in situations where there is discord, however, the goals and policies of the actors are not in line with each other and cooperation becomes necessary (Keohane 1984).21 Cooperation in these situations is a reaction to real or potential conflict. According to Keohane (1984), even if we expect situations of discord to be complicated and political, we should also expect actors to be able to cooperate to overcome policy and goal discrepancies. Cooperation, in Keohane’s view, requires “active attempts to adjust policies to meet the demands of others” (1984, 12). Cooperation for Keohane (1984) is about mutual adjustment, rather than simply a situation where common interests outweigh conflicting ones.22 Organizations can successfully cooperate by facilitating others’ goals in the process of facilitating the attainment of their own goals. A practical example of this is the resource sharing common to organizational networks (e.g., Child and Faulkner 1998; Das and Teng 1998; Provan and Milward 1991). In crises, particular resources or expertise are often short in supply and high in demand. The type of cooperation Keohane (1984) previously described occurs if the organizations pool their resources and adjust their use of resources on the basis of their own needs as well as the needs of other actors. This is a form of exocentrically motivated cooperation, where helping others is a part of achieving one’s own goals or where cooperation is expected to lead to some other perceived or real reward. In this study I will try to capture this reward based or utilitarian cooperation behavior by highlighting when actors actively facilitate the attainment of others’ goals and when they work to facilitate the attainment of their own organization’s goals.

21 ���������������������������������������������������������������������������������� If there is harmony among actors, Keohane asserts, cooperation is unnecessary and can even be harmful. 22 ������������������������������������������������������������������������� Since common interests can exist both with cooperation and with discord, cooperation is not merely a function of common interests.

Conceptualizing Organizational Crisis Cooperation

35

Coordination Coordination is a cooperative behavior that implies joint decision-making (Martin 1999) and communication to achieve congruence or similarity of policy formulation or action (Jennings 1998; O’Toole 1995). There are a number of ways in which coordination as an activity may take place. Coordination can be achieved by pre-existing plans for directing and standardizing how organizations will work (Hage et al. 1971). Keohane’s approach (1984), is through situational negotiation. For Keohane (1984), policy coordination does not necessarily involve negotiation or bargaining but often does. He argues that cooperation “requires that the actions of separate individuals or organizations – which are not in pre-existent harmony – be brought into conformity with one another through a process of negotiation … often referred to as policy coordination” (1984, 51, italics added). When negotiation and bargaining take place they are often accompanied by other actions aimed at inducing the other party to change their policy to match one’s own. Using coordination as an indicator of cooperation, I look to see if the actors make attempts to adjust their policies to each other’s objectives through communication and/or negotiation. Pathological Cooperation – Rally around the Flag, Groupthink and New Group Syndrome Social psychologists Staw, Sandelands, and Dutton (1981) view cooperation primarily in terms of group effects or group behavior. In their view, cooperation is something which leads to group cohesion, while conflict is something which leads to group disintegration. Forms of excessive group cooperation, such as rallyingaround the flag (Baker and Oneal 2001; Hetherington and Nelson 2003; Mueller 1970), groupthink (‘t Hart et al. 1997; Janis 1989), and new group syndrome (Stern 1997; Stern 1999) can be more or less pathological and counterproductive. Rallying around the flag is a general concurrence seeking within a larger ingroup, such as one’s country, and a lack of in-group criticism in the face of a perceived common threat. The concept was originally tested by Mueller (1970) in relation to presidential approval ratings. He ascribed this effect to “international crises and similar phenomena” which he predicted would “give a President a shortterm boost in popularity” (Mueller 1970, 20). Mueller (1970) also cites Polsby’s view on the grounds for this rallying behavior, stating that “Invariably, the popular response to a President during international crisis is favorable, regardless of the wisdom of the policies he pursues” (Polsby 1960, quoted in Mueller 1970, 21). Baker and Oneal (2001) summarize the two main arguments for why the public rallies around their leader in times of crises. The first explanation, the patriotism explanation, is that “when important political, economic, or strategic interests of the nation are at stake, the public will focus uncritically on and unite behind the commander-in-chief in a show of patriotic support” (Baker and Oneal 2001, 667).

36

Organizational Cooperation in Crises

In this argument rallying behavior is viewed as a reflex to a threat to core values, rather than a shift in opinions regarding the President and his policies. The second explanation, opinion leadership, asserts that the rallying behavior is based on public opinion formation limited by a lack of access to information or information that is to a great extent controlled by the President and his administration (Baker and Oneal 2001, 667–68).23 Mueller asserted that an important characteristic of situations that generate this type of behavior was that it must affect and “confront the nation as a whole” (1970, 21). He predicted that events that do not pose an outside threat to the group as a whole were just as likely to “exacerbate internal divisions as they are to soothe them” (Mueller 1970, 21).24 The idea that organizations may display the same type of rallying behavior as the public in crises is based on the perceptual crisis definition where decision makers perceive a threat to core values, urgency and uncertainty. Characterized by intensified uncritical support of a common leader and her policies as the crisis hits, this type of cooperation will be coded as rallying around the flag in the crisis cases. Groupthink is a similar form of strong concurrence seeking but one that pertains more directly to small decision-making groups. In cases of groupthink, a group quickly reaches a consensus around a decision rather than critically examine multiple aspects of the decision problem (Janis 1972, 1982). Groupthink has been described as “an excessive form of concurrence seeking among members of high prestige, tightly knit policy-making groups” and “[i]t is excessive to the extent that the group members have come to value the group (and their being part of it) higher than anything else” (‘t Hart 1991, 242). The high value attached to group membership causes the members “to strive for a quick and painless unanimity on the issues that the group has to confront … group members suppress personal doubts, silence dissenters, and follow the group leader’s suggestions” (‘t Hart 1991, 242).25 23 ����������������������������������������������������������������������������������� Potential critics in a fast unfolding crisis, where their access to information is limited, refrain from expressing criticism, say nothing or even offer “guarded support for the president’s policies for fear of looking foolish or unpatriotic” (Baker and Oneal 2001, 668). The media as a public watchdog also find it hard to provide a balanced or critical perspective on the President and his policies in these situations. The public, in turn, “is largely cut off from the cues it traditionally employs to develop an opinion and … the public is led to assume that there is a consensus among political leaders on the issue and to support the president” (Baker and Oneal 2001, 668). 24 �������������������������������������������������������������������������������� Focusing on Presidential approval ratings Mueller investigated and hypothesized about international crises as these rallying points where the nation as a whole was confronted, and domestic crises as points that were just as likely to fuel division and hence eliminate any approval spike for the President. 25 ������������������������������������������������������������������������������������ Manipulation of a group process, Stern and Sundelius point out, is in essence based in a conflict of preferences and may be hard empirically to distinguish from groupthink. “Manipulation entails the implementation of a hidden agenda by one or more group members, through the deliberate structuring of the process or the substantive information

Conceptualizing Organizational Crisis Cooperation

37

New group syndrome demonstrates an unwillingness on the part of actors within a group to take an active and critical role in preparing decisions due to the newness of the decision group. This type of behavior stems from actors being unsure of their competence or role in the group compared to that of others in the group (Stern 1997). “[N]ewly formed groups lack a group-specific sub-culture. Members’ uncertainty and insecurity seem to make such groups particularly susceptible to directive leadership from one or more assertive members, leading to premature concurrence-seeking” (Stern and Sundelius 1994, 103). “Such ad hoc decision units may also be more susceptible to … anticipatory compliance … More established groups will instead be strongly affected by their highly developed sub-culture and previous experience of success and failure” (Stern and Sundelius 1994, 103–4). Taken together, these three behaviors – rallying-around the flag and groupthink, and new group syndrome – illustrate highly cooperative reactions to threatening or uncertain environments that face a group. These strong forms of cooperation have the unintended consequence of leaving the group particularly vulnerable to the influence of strong individuals, and make it less likely that decisions and actions by the group will be based on a balanced deliberative process. Displaying Honesty and Claiming Honesty in the Interaction Predictability of action, keeping promises, and sticking to agreements made are core features of successful repeat interactions. These features are also key in establishing trust in interactions where the ability to hold others accountable is limited (Bolton et al. 2004). In international relations, the durability of agreements is directly linked to the ability of participants to monitor the behavior of others (Oye 1985, 15–7). In game theory terms, this is conceptualized as the degree to which a participant has access to information about another party in the interaction. To a great extent, the interaction dilemma operates under the assumption that perfect information on the others’ motives and preferences is lacking (Axelrod and Keohane 1985, 232–4; Poulsen and Svendsen 2005). Consequently, getting ‘costly signals’ is a way to build trust in repeated interactions (Raub 2004). International agreements (Axelrod and Keohane 1985, 234–6; Keohane 1984), arms reductions (Evangelista 1990; Nye 1984; Thee 1977), financial stability in international markets (Helleiner 1994; Ruggie 1982) and Internet trading (Bolton et al. 2004) base … While successful manipulation may manifest itself at the group level as premature concurrence-seeking, it is actually the result of a deviously engineered consensus which may exploit the relaxed atmosphere of a cohesive group or the natural tentativeness on the part of the members of a newly formed group. This kind of manipulation, however, implies intense, though latent, conflict. If resistance to the preferred course of action were not anticipated by the manipulator, manipulative tactics would not be necessary. However, the group level consequences of manipulation may be difficult to distinguish empirically from groupthink” (Stern and Sundelius 1994, 104).

38

Organizational Cooperation in Crises

depend on the degree of information openness between the actors and on evidence that the participants are indeed acting in ‘good faith’. Displaying honesty is a key attribute of a cooperating partner. Claiming honesty, even though it does not prove you are indeed acting in good faith, signals that you know of and recognize the value of this characteristic for others to cooperate with you. As a result, these two behaviors, displaying honesty and claiming honesty, serve as indicators of cooperative behavior in this study. Need for Contact Persons Hage, Aiken and Marrett (1971) view coordination as “the degree to which there are adequate linkages among organizational parts … so that – organizational objectives can be accomplished” (1971, 860, italics added). According to Hage, Aiken, and Marrett (1971) coordination is accomplished by plan or feedback. Feedback entails the spreading of new information in order to facilitate mutual adjustments of different parts (Hage et al. 1971).26 With this in mind achieving adequate communication and feedback requires contact between organizations. More hierarchical or militarily oriented structures tend to rely on planning to coordinate this process (Dynes and Aguirre 1979), whereas non-hierarchical entities in crises often bring new constellations of organizations together and rely on situational communication and feedback for coordination. An apparent need for contact points or persons in a crisis will be an indication that the organizations are working independently and that joint or coordinated action is not fully taking place. Resisting Coercive Cooperation – Implementation Slippage Returning to Keohane’s (1984) view of policy coordination, we find ways that discord may lead to conflictual or competitive behavior among actors. If attempts to adjust policies are not made when there is a misalignment of actors’ policies and goals, Keohane argues that discord between the actors rule. In this situation actors tend to regard each other’s policies as hindering the attainment of their own goals, and they hold each other responsible for these constraints. The resulting discord often leads actors to attempt to change other actors’ behavior or policies. When these attempts meet resistance, conflict develops between the actors (Keohane 1984, 52).27 26 ������������������������������������������������������������������������������� While the first is based on external control for adherence, the latter is more dependent on internal control in the form of social control and peer pressure. Hage, Aiken and Marrett (1971) argue that coordination based on feedback is more likely to occur when there is a great diversity of occupations and when the variety and uncertainty of the tasks increase. 27 ������������������������������������������������������������������������������ Cooperation, in turn, occurs when the attempts succeed in making the parties’ policies more compatible (Keohane 1984).

Conceptualizing Organizational Crisis Cooperation

39

This is essentially the same process described by scholars interested in bureaucratic politics (e.g., Honig 2001; McKeown 2000; Rosenthal et al. 1991). “The model of bureaucratic politics postulates that conflicts of interest and power games between different sections, departments, and agencies within a government administration are the most powerful determinants of policy choices” (‘t Hart 1991, 249). Preston and ‘t Hart (1999) describe the interaction between parties in bureau-political conflict simply as a “continuous ‘pulling and hauling’ and bargaining between (clusters of) actors” (Preston and ‘t Hart 1999, 55). They elaborate on the expressions this conflict may take in the implementation phase of a decision or policy process. In addition, they argue that decisions and policies are prone to both temporal and content slippage when actors are trying to change the behavior or policies of another. Temporal slippage refers to time delays in the implementation of decisions and policies, while content slippage represents modifications of the content of a policy or decision in the implementation process (Preston and ‘t Hart 1999, 55). These types of organizational behaviors demonstrate a struggle against or subversion of what these organizations perceive as coerced cooperation. Coercive cooperation (Martin 1992) is the mode of cooperation Keohane (1984) describes as the successful attempt by one actor to persuade, manipulate, or force another party to adapt its strategy. For the discontented and manipulated actor, the implementation phase gives her a chance to retaliate, either overtly or covertly, by stalling or undermining the intent of a decision or a policy direction with which she disagrees. Temporal and content implementation slippage will be two indicators of the more competitive behaviors found under the umbrella of cooperation. Credit-seeking In the event that a crisis turns out to be a political or financial opportunity, a certain amount of credit seeking may also take place. Symbolic actions play an important part as framing devices, especially if one seeks credit or wishes to put a more positive ‘spin’ on the perception of a crisis (e.g., ‘t Hart 1993; Hagström and Sundelius 2001). Crises often present distinct political opportunities. Showing empathy at the site of a catastrophe can generate substantial PR capital for leaders. The onslaught of VIP visits to a disaster site (Hagström and Sundelius 2001) is in part driven by this opportunity to show empathy, responsibility, and resolve (‘t Hart 1993). A leader who fails to spot these political opportunities or who fails to assume responsibility publicly at the right time may quickly see the end of her political career (Boin et al. 2005). For organizations, crises often triggers responses characteristic of bureaucratic politics, such as credit seeking behavior. “[I]n crisis situations government authorities and public agencies definitely do not lose interest in the ranking order of power and prestige. For crisis-relevant organizations, the actual moments of crisis are the very moments their continued existence may be at stake. Indeed, by definition, their rationale, legitimacy and even funding may derive form their

40

Organizational Cooperation in Crises

performance in critical situations” (Rosenthal et al. 1991, 223). For this reason “authorities and agencies involved in the process of crisis decision-making may coolly anticipate the re-allocation of personnel and budgetary resources in the aftermath of the crisis” (Rosenthal et al. 1991, 224). Organizations involved in a crisis may well anticipate that changes in the allocation and distribution of resources between organizations that take place in crises will remain in place after the situation stabilizes and eventually returns to normal (Rosenthal et al. 1991).28 Calculated Avoidance of Responsibility – Game of Old Maid, Blame-games, Attributing Fault to Others, Covert Behavior and Breaking Agreements Resorting to calculated avoidance of responsibility and blaming are as much a part of bureaucratic politics as seizing the opportunities of crises and accomplishing a favorable ‘fait accompli’. A majority of bureaucratic politics literature focuses on the competitive nature of public administration and management, with many scholars highlighting the importance of plurality of interests and competition (Allison 1971; Halperin 1974; McKeown 2000; Preston and ‘t Hart 1999, 52; Rosati 1981). When a group first forms to manage a particular crisis, the question of who owns the problem tends to still be open. In this initial ownership void, there are substantial opportunities for groups to either appropriate the problem and thereby get an opportunity to shape the management of the problem, or to ‘pass the buck’ and avoid taking responsibility (Staw et al. 1981). Passing the buck of responsibility in crises is similar to the card game Old Maid where the players try to avoid getting stuck with the odd queen at the end of the game.29 Another type of avoidance strategy involves framing the problem in a way that makes another organization the perceived natural lead agency (Staw et al. 1981). In cases of perceived policy fiascoes and in crisis situations that progress from bad to worse, sitting in a lead position can mean heavy political losses. It can also cause great personal stress as 28 ��������������������������������������������������������������������������������� It may also be that bureaucratic politics spring from the fact that “all parties concerned are convinced that they can make a positive contribution to the public cause. This situation fosters bureaus and bureaucrats insisting upon their interpretation of what would be the most effective, if not the only, way to avert threat” (Rosenthal et al. 1991, 225). Similarly, “a battle of the good Samaritans” may also take place if organizations with responsibility to provide services in a crisis, such as food distribution, receive an increase in service capacity during the crisis and finds that the demand for their services is lower than expected. Attempts to control and gain access to the demand side may then generate competitive organizational behavior, such as credit seeking in the crisis. Examples of this have been noted between organizations such as the Red Cross and the Salvation Army, in crises (e.g., Svedin 2001). 29 �������������������������������������������������������������������������������� The object of the game is to take cards from a neighboring player’s undisclosed hand, discarding pairs in your own hand, until one player runs out of cards. The player that is left with an odd queen (that has no matching king) is stuck with the ‘Old Maid’ and loses.

Conceptualizing Organizational Crisis Cooperation

41

it may, in the post-crisis phase, mean the end of a career and even the ‘death’ of discredited companies or agencies (Bovens and ‘t Hart 1996). An organization that perceives its own coping capacity to be low, and fears getting stuck with the management of a situation that it sees itself as ill equipped to handle, may become defensive and resort to coping mechanisms of avoidance or blaming (Brandstrom and Kuipers 2003; Hood 2002). In an attempt to redirect attention elsewhere, share the guilt, or to clarify a situation that the organization(s) facing the crisis thinks has been misperceived, the lead organization under pressure may criticize the actions of other actors. This organization may even act covertly to get information it would not otherwise be privy to in order to bolster its case or find ways of getting around the problems it is facing with other actors (Newlove et al. 2003, 126–8). If the organization cannot find ways of skirting, avoiding, or redirecting the criticism, they may simply not live up to the commitments made to other actors or break agreements made. Based on the characterizations and assertions about calculated avoidance of responsibility, five indicators of the more competitive sides of cooperation have been adopted. These indicators are game of Old Maid, blame-games, attributing fault to others, acting covertly, and breaking agreements. The Transboundary Crisis Management Data: Developing a Dataset to Look at Organizational Cooperation in Crises In order to examine organizational cooperation in crises empirically the author, together with Margaret Hermann and Bruce Dayton, has created a dataset that contains, among other things, the operationalizations of cooperative behavior and strategies presented in this book. The Transboundary Crisis Management (TCM) dataset now presents a first opportunity to systematically examine crisis cooperation across a large set of cases. This quantitative dataset draws on in-depth case studies of crises and the management of these crises, facilitating the testing of theoretical models and case study findings against data on crises from across the world, across time and at different levels of analysis. Against the backdrop of these empirical tests, we will be able to merge successful ideas into well-founded assertions about organizational cooperation and conflict in crises and set out to explore the effect of this on the management and outcomes of crises in future research. The case studies that the TCM dataset rests on have been designed according to the model of the structured focused comparison. This method, first developed by George (1979), has been elaborated and refined in applying it to a number of crisis research ventures. Two such projects, Crisis Management Baltic and Crisis Management Europe (Stern and Sundelius 2002),30 have produced case studies of 30 ������������������������������������������������������������������������� Details of the research project and the country volumes are available on , accessed 30 June 2006.

42

Organizational Cooperation in Crises

national crisis management in a large number of countries that have been made available through a set of country volumes (e.g., Sundelius et al. 1997; Stern and Nohrstedt 1999; Stern and Bynander 1998; Larsson et al. 2005; Hansén and Stern 2000; Bynander and Chmielewski 2002; Buus et al. 2005; Brändström and Malesic 2004; Bernhardsdottir and Svedin 2004; Porfiriev and Svedin 2002). A third, venture is the Transboundary Crisis Management Project31 at the Moynihan Institute of Global Affairs, through which a large number of case studies have been conducted as Masters of Arts capstone projects in public administration. Together, the body of case studies produced by these three projects comprises the empirical foundation from which the dataset used in this study has been generated. In practice, the application of the structured focused comparison design (George 1979; George and Bennett 2005) to the case studies means that the cases share a common crisis definition, as well as a definition of ‘occasions for decision’ and a procedure for identifying and reconstructing these occasions within a specific crisis. Following the original divisions and parameters of the case studies, the dataset is divided into four parts covering different empirical information from the case studies and coding different sets of variables (descriptive, decision-making variables, decision unit variables, and thematic variables). Consequently, even though the studies were developed in two separate research institutions, and were written and researched by scholars and practitioners across the world, all studies follow the same key focus, method and structure. The dataset contains empirical data on 331 variables and is a quantification of crisis characteristics and crisis management variables that have been coded in 71 thick description crisis case studies. Within these case studies, a total of 333 observation points have been identified at the occasion for decision level of analysis.32 A subset of these observations pertain to organizational cooperation as the production of the dataset was geared toward maximizing the research value of this dataset for scholars interested in crises and crisis management. The crisis cases vary greatly in terms of the types of crises studied (floods, riots, poweroutages, train crashes etc.), the time period when the crisis occurred, the geographic location, and consequently the cases vary a great deal in terms of organizational settings. The variables that have been coded include (broadly speaking) crisis characteristics, organizational factors and characteristics, and individual or group behavior and statements. These types of variables stem from the cognitiveinstitutional approach within which the case study design was originally developed (see Stern 1999; Stern and Sundelius 2002). A large number of other variables identified in research on organizational cooperation and crisis management are coded in the dataset. The variables in the quantitative template were selected based on a state-of-the-art survey of the field 31 ������������������������������������������������� Details of the research project are available on , accessed 30 June 2006. 32 �������������������������������������������������������������������������������� These numbers reflect the status of the dataset at the time of the study. Since January 2007 more cases have been coded and added to the dataset.

Conceptualizing Organizational Crisis Cooperation

43

of crisis management research. Articles giving an overview of various research strands in the field were used as an initial variable map (Stern and Sundelius 2002; Hagan et al. 2001; Hermann 2001; Hermann et al. 2001). More specialized articles and publications by prominent scholars in the field were then surveyed for variables that have more recently surfaced on the research agenda of scholars, that have received less empirical testing, and that have a narrower area of application (e.g., ‘t Hart et al. 1997; ‘t Hart et al. 1993; Svedin 2001; Newlove et al. 2000; Hood 2002). The aim has been to make the variables and variable measures as detailed as the empirical material in the case studies would allow. CATPCA: A Way of Empirically Examining Cooperation in Crises In order to empirically examine organizational cooperation in crises I will apply Categorical Prinicipal Component Analysis (CATPCA) to data in the newly created Transboundary Crisis Management (TCM) dataset which contains the indicators of cooperative behavior and strategies presented above. CATPCA is a quantitative technique very similar to the more commonly used Principal Component Analysis (PCA). Principal component analysis (either CATPCA for categorical data or PCA for interval data) is often used as an exploratory technique aimed at reducing data and finding patterns among variables assumed to measure aspects of a common underlying concept (Linting et al. 2006b; Meulman et al. n.d.). This method of analysis finds common empirical patterns between variables, sorting common features into so-called ‘components’ and assigning a value, or factor score, to each variable. This score indicates the strength of the connection between individual variables and identified components. CATPCA can also be used to assign specific values to each observation in the dataset, or object scores, based on the component solution. One of the advantages of using principal component analysis (either CATPCA or PCA) is that the technique can handle a large number of variables (in this study referred to as indicators). For the purpose of this study, this means that the multitude of indicators involved in a complex concept like cooperation can be coded and included in the analysis. Principal component analysis can even be used in situations where the number of indicators exceeds the number of observations (Manisera et al. 2005, 53). As a result, the traditional guideline that a solid statistical examination requires ten times more observations than the number of variables entered does not apply to this exploratory technique. As previously stated, CATPCA is the specific type of principal component analysis used in this study, and has a number of additional advantages over PCA. The main advantage is that CATPCA allows for the analysis of categorical variables, common to social science research and the measure of the dependent variable in this

44

Organizational Cooperation in Crises

study.33 In addition, CATPCA is also capable of handling multiple measures (such as numerical, nominal and ordinal variables)34 simultaneously without requiring recoding prior to analysis. By using CATPCA in SPSS, individual variables can be assigned a specific scaling, something that ordinary PCA or factor analysis does not allow. CATPCA also does not rely on assumptions about the distribution of the data, making it possible to consider non-linear relations. The underlying workings of CATPCA (and principal component analysis in general) consist of an analysis of variance. The assumption is that the observations of the variables included have some common variance, as well as both individual variance for the variable and variance for the individual observation (so-called type 1 error).35 What this analysis does is show clusters in the common variance in the sample (identifying components) and reports how much of the individual variable variance can be predicted or accounted for by these clusters (showing components scores of each variable on each factor). The object score calculated for the individual observations in each variable show where the observation would fall based on the cluster solution produced. Even though CATPCA is a data exploration and reduction approach, it is still capable of testing the stability of the model it presents. Preliminary tests using bootstrapping have proved successful in assessing the stability of both linear and non-linear CATPCA (Linting et al. 2006a). This type of model stability test of CATPCA is still under development and is not yet widely accessible, but it will be possible to apply these tests in the future. The object scores calculated for the CATPCA component solutions will be used as interval measures of the components of cooperation. This new interval measure will be used in Chapter 6 to link cooperative behaviors and strategies using correlation analysis. As outlined in Chapter 7, correlation and regression analyses will provide the empirical links between the identified sub-types in the dependent variable and variations in the crisis setting.

33 ���������������������������������������������������������������������������� While some of the variables are Likert-scale type variables with categories representing high, medium, low and often are treated as an interval scale (using PCA for its analysis) this is not conceptually right (Manisera et al. 2005). They point out that treating multinomial variables as interval scales assumes that the distance between the categories is equally spaced and in essence that category six would be two times ‘stronger’ than category three. Using CATPCA avoids these assumptions while retaining the order of the categories. 34 ���������������������������������������������������������������������� Continuous variables can be included in CATPCA provided that they are transformed into interval measures. 35 ������������������������������������������������������������������������������������� The variance shared by the variables can be assumed at the outset to be all variance making it equal to 1. The principal component analysis then shows the amount of variance that could be explained by the model for each variable as less than 1 (for example .93 or .62, corresponding to 93 percent and 62 percent of the individual variable variance for each).

Chapter 3

Crisis Cooperation in Light of the Three Traditions: Case Illustrations In this chapter we examine three case samples, from the Transboundary Crisis Management (TMC) dataset that illustrate how the cooperation indicators identified by the three research traditions manifest in crises. The case samples show that each perspective – international relations, organizational research, and social psychology – contributes to and improves our understanding of crisis cooperation. The three research traditions share a common interest in the interaction, or exchange between different actors. Where they differ is primarily in their selection of actors or units of analysis, their view of the relationship between the units of analysis, and their conception of cooperation (behavior, goal, process, strategy, or a forum/institution). Not surprisingly, each of the three research traditions view cooperation through a particular analytical lens: psychology primarily examines the attitudes and behaviors of the organizations or actors; international relations scholars examine the goal to be achieved through actor interaction; and organizational scholars examine the means by which an institutional interaction or exchange achieves a stated goal. In the following sections below, research traditions will be briefly recapitulated in terms of their main conceptualizations of cooperation. Following each tradition a case study example then demonstrates what these conceptualizations look like in a real crisis case. The case examples will discuss the general interaction patterns between key actors in the case, and what cooperation looked like in both decisionsituations and across the crisis as a whole. A comparison of the cases will underline the relevance of each tradition and the importance of drawing on all three in order to more fully understand organizational cooperation in a crisis interaction setting. The indicators of cooperation that are used in the study will be highlighted in the case examples using italics. An International Relations Perspective on Cooperation The first perspective we draw on in our conceptualization of crisis cooperation, come from the field of international relations (IR). Scholars researching cooperation in this field view cooperation primarily as a goal that may be achieved through the interaction of actors. To some extent they also look at cooperation as embodied in institutions where interaction and exchange between actors take place in a peaceful

46

Organizational Cooperation in Crises

way. International relations scholars have conceptualized cooperation as: collective security mechanisms, participation in alliances and international organizations, collective actions among sovereign actors, compliance with institutional norms and regimes, non-violent sanctions as a form of coercive cooperation, non-violent dispute settlement (such as mediation, negotiation, or adjudication), trade and economic interactions, reducing barriers to interaction (such as trade tariffs or quotas), and integration within super-national institutions. The case example of NATO and the Kosovo crisis, discussed below, is meant to illustrate how conceptions of cooperation stemming from international relations research can enhance our own conceptualization of cooperation in crises. It should be noted that the case was only in part an international relations crisis, in that it involved a international military alliance objecting to the actions taken by a sovereign government against an area and people within that sovereign nations’ borders. The crisis is taking place in the context of the war in Former Yugoslavia and brings to light the international community’s reaction to what was perceived as another attempt by Serbian leader Milošević to ethnically cleanse part of the region. Case Example 1: NATO and the Kosovo Crisis General Description of the Interaction in the Case The interaction in this case study was primarily between NATO, the Kosovo Liberation Army (KLA), and the Yugoslav security forces in Kosovo. NATO did not perceive the conflict between the Yugoslavian forces and the KLA as a legitimate domestic conflict because the Belgrade government had recently used the Yugoslavian forces to engage in domestic ethnic cleansing. “In order to stop the conflict and in order to put pressure on Belgrade to abide to the UNSC 1119 resolution, NATO used the strategy ‘diplomacy backed by the use of force’ where NATO was providing leverage to the diplomats in negotiations with the parties in the conflict” (Stepanovic 2003, 2). Four months into this diplomatic strategy, the conflict in Kosovo was threatening to spread into neighboring countries. In response to this perceived threat, NATO decided to show a warning signal to the Yugoslav leadership by displaying a show of force and authorizing military presence on the ground in Kosovo and Macedonia. This initial threat of military force by NATO was followed by a strategic dance of increases and decreases in Yugoslav army presence in Kosovo. Likewise, a series of increases and decreases in threats by NATO were made in response to Yugoslav actions and violations of UN resolution 1119. Approximately nine months after the first NATO military engagement in Kosovo and Macedonia, NATO initiated a bombing campaign to achieve two overarching goals: to contribute to the  ������������������������������������������������������������������ This case example is based on the case study by Stepanovic (2003).

Crisis Cooperation in Light of the Three Traditions

47

international community’s efforts of achieving a peaceful resolution to the crisis, and to promote stability and security in the region. Across the Kosovo conflict, NATO played a number of different, although not necessarily commensurable, roles. It acted as chief negotiator between the parties (the Yugoslavian government and the KLA), it was the adjudicator/enforcer which confronted the parties with an ultimatum when negotiations stalled, and, finally, it was the enforcer/executioner of the international community’s will as it engaged in the airstrike campaign to resolve the crisis. Ultimately, the military campaign resulted in a great number of, both Yugoslavian but also Kosovar casualties, and displaced Kosovar refugees. The Kosovo crisis also set a new international norm for humanitarian interventions in domestic conflicts (Stepanovic 2003, 2–3). Cooperation and Conflict in Decision-situations While the actors in this Kosovo case did engage in a degree of a cooperative behavior, the majority of the actions in specific decision-situations were, in fact, competitive or conflictual. The most frequent indicator of cooperation in decisionsituations in this case is that the actors consult each other on what is happening. This is not a particularly strong form of cooperation in a decision-situation; they are meeting and talking to one another, but not doing much more. The second most frequent indicator in this case is that an actor requests, proposes or suggests that another actor does something in a decision-situation. This is a stronger form of cooperation, but it is far less frequent than actors simply meeting and consulting with one another. Only one out of the five decision-situations analyzed in the case was primarily cooperative, showing NATO consulting, approving, agreeing and yielding in the decision-situation. This situation required NATO to formulate a response to the Yugoslav government’s acceptance of Organization of Security and Cooperation in Europe (OSCE) observers on the ground in Kosovo. NATO had authorized airstrikes and a military build up in the region the same day that the Yugoslavian government agreed to have the OSCE observers come in and verify the withdrawal of 10,000 troops from Kosovo. NATO was not keen on having other multinational organizations on the ground, but received open recognition from the negotiator Richard Holbrook that NATO’s pressure was working on changing the Yugoslav leadership’s mind on its presence in Kosovo and encouraged them to keep the pressure up. In response to Yugoslavia’s verified withdrawal of the troops, NATO voted to suspend the airstrikes. The other decision-situations, however, displayed several indicators of competition and conflict. The two most commonly featured indicators of conflict in this case were the actors expressing disapproval of something someone else did or said in decision-situations or threatening another actor in a specific decisionsituation. In one decision-situation NATO, consulted, requested, disapproved, threatened and finally went so far as to reduce normal relations with the other actor (the Yugoslavian government). This decision-situation arose when the US

48

Organizational Cooperation in Crises

representative of the international community effort presented an ultimatum at the peace talk in Rambouillet, France were rejected by the Yugoslav leadership and the KLA. The rejection of the ultimatum resulted in an intensified Yugoslavian presence in Kosovo, fleeing refugees and the OSCE leaving Kosovo. US negotiator Richard Holebrook made a final attempt to persuade Milošević to stop the government’s offensive in Kosovo or they would face imminent NATO air strikes. When Milošević refused to comply, NATO initiated the air strikes. “NATO leaders at that point firmly decided they had to ‘do something’ about Milošević’s repression and atrocities, teach him a lesson for failing to keep earlier agreements, and move to uphold their own credibility after many threats and no action” (Stepanovic 2003, 12). Cooperation and Conflict across the Case The actors involved in this case displayed a range of indicators of both cooperative and conflictive behavior across the crisis. On the more cooperative end of the spectrum, NATO, as an actor, facilitated the achievement of its own group’s goals (helping their own group and group members). NATO and other actors also coordinated their actions in order to achieve common objectives or better results. Actors also facilitated the attainment of others’ goals, which is a more altruistic form of cooperation that does not immediately or exclusively benefit the actor itself (NATO, in this case). There were also voluntary concessions made by actors in this crisis. This is a positive first step that signals that one party is cooperatively inclined and willing to meet others half-way. Toward the more competitive or conflictive end of the spectrum of indicators we see actors in this case seeking credit for that which is perceived as going well in the crisis and for actions perceived as valuable in the management of the crisis. We also see actors breaking agreements they have made. This may be something caused by conflict as well as something that causes additional conflict in relation to other actors. In the Kosovo crisis actors also attributed fault in the conflict to other actors. This is a form of blame-game, where an actor attributing fault to others is not only trying to escape blame herself, but is openly challenging the actions or intentions of others. By faulting others, the actor is challenging other actors to either own up to their part in what went badly or to publicly show that they, in fact, did nothing wrong. In the Kosovo case, NATO faulted Milošević and his supporters in both the Yugoslav government and military for the violence that was erupting in Kosovo. NATO argued that “the same people, who were greatly responsible for the violence and ethnic cleansing in Bosnia, … were again trying, in [an] already seen scenario[,] to do the same things against [the] Albanian civilian population in Kosovo” (Stepanovic 2003, 14). We did not, however, see actors display some of the behaviors or make the kinds of statements that are common indicators of cooperation and conflict as seen by organization scholars and by social psychologists. We did do not see the types of bureaucratic competitive and conflictive behavior that is indicated by

Crisis Cooperation in Light of the Three Traditions

49

content slippage when decisions and policies were implemented. We also did not see stalling tactics in the implementation of decisions across the crisis, what organizational scholars would consider representative of conflictive organizational behavior. We also did not see some of the more psychologically driven cooperative behaviors such as groupthink and rallying around the flag. There was not a pronounced blame-game with regard to whom or what had caused the crisis. We also did not see the kinds of persuasion and image management that belongs to the more psychological and political understandings of crisis cooperation, behaviors such as displaying honesty toward others or actors publicly claiming that they are being honest and forthcoming about the crisis and their actions in regards to it. An Organizational Perspective on Cooperation The second perspective that contributes to our conceptualization of cooperation is that presented by organization scholars. From an organization research point of view, cooperation has been envisioned as the means by which organizations reach stated goals, or as the institution within which interaction and exchanges between organizations take place. Inter-organizational cooperation has been cast in terms of the following: organizational connectivity, the degree of social integration within and across organizations, inter-organizational linkages, partnerships, joint decision-making, alliances and task-forces, cooperative processes and procedures (such as standardization and mutual adjustment), inter-organizational coordination, and service delivery networks, joint action, collaboration, and consensus. The second case example on the Korean financial crisis illustrates contributions by organizational scholars to our conceptualization of cooperation in crises. This national financial crisis appeared on the surface to be the result of a run on foreign investment capital. At a deeper level, however, it was shown to be the result of foundational economic problems including the rapid and condensed growth of the Korean economy, the creation of conglomerates, over-investment, corruption, and negative balance of trade. It was not a crisis only for the Korean Department of Finance as an organization or major Korean financial institutions; it was a crisis that required cooperation between several actors, both nationally and internationally. Case Example 2: The Korean Financial Crisis General Description of Interaction across the Case In 1997 the Asian currency devalued several Asian currencies causing a shortage of foreign currency reserves and increased the national debt to the extent that South Korea was unable to make payments on its national debt. Domestic prices  ������������������������������������������������������������ This case example is based on the case study by Park (2003).

50

Organizational Cooperation in Crises

soared rapidly and several large Korean corporations accumulated bad debts and went bankrupt. Facing Presidential elections, the party that had ruled during forty years of rapid economic growth, suddenly struggled to maintain its position at first without accepting any outside help in the financial crisis. The crisis was initially managed by the Korean President, the Minister of the Treasury, the Chief of the Korean Central bank. With the situation spiraling out of control, the government eventually asked the International Monetary Fund (IMF) for help, and ultimately accepted a bail out and reform package. The resulting cooperation took place between a much wider range of actors: notably the Korean government, the South Korean Central Bank, several large Korean corporations, the Korean press, national political parties, foreign creditors (mainly banks), the IMF, and the United States. All of these actors worked to bring the Korean financial crisis to a halt and bring about the needed changes to restore confidence in the Korean market (Park 2003, 1–3). Cooperation and Conflict in Decision-situations In the two first decision-situations in this case example, the Korean government displayed mainly conflictive or less cooperative behavior. These situations raised questions regarding whether or not the government actually perceived a financial crisis, as well as whether the government should take action once the crisis worsened. Ultimately, the government simply commented on the situation publicly, met and consulted on the issue, and then rejected the perception of a crisis and any need for further action. The first situation was a collective failure to recognize the problem and the impending crisis at hand. Armed with information on the currency crisis facing Thailand, as well as the increasingly unfavorable economic indicators in the South Korean economy, the government consulted on the situation and proceeded to publicly reject the idea that anything like the crisis in Thailand could ever happen in Korea. In the second decision-situation, the Minister of the Treasury commented publicly on the situation and admitted that South Korea was now in an economic crisis. The government then met and consulted as a group, rejecting the suggestion that the government needed to act with regard to the crisis. The government publicly stated that the Korean economy was still fundamentally strong and that the economy would bounce back from the crisis on its own (Park 2003, 15–7). The third decision-situation was one of the most intense in the Korean financial crisis. This situation emerged when the South Korean government was faced with the decision of whether or not to accept the bailout offer made by the IMF. Here we find several indicators of cooperative behavior, with two indicators of more competitive or conflictive behavior. In this decision-situation, the government commented on the situation by publicly announcing a number of policies that were being put in place to deal with the crisis. The situation, however, continued to deteriorate and the government met and consulted as a group on whether to accept bailout loans from the IMF. The government decided to cooperate with the

Crisis Cooperation in Light of the Three Traditions

51

IMF, accept the offer, and approved the bailout package. The government then made an agreement with the IMF and requested $55 billion USD in emergency funds. The sitting President of South Korea, who was initially somewhat shocked by how the economy was spiraling out of control, started to take an active role in the management of the crisis. He demanded further action and expressed his disapproval of the team that had been handling the crisis up to this point by replacing them and denying the new team members a veto position (Park 2003, 17–9). Cooperation and Conflict across the Case Across the Korean financial crisis, the South Korean government displayed three indicators of cooperation. First, they displayed groupthink, a type of pathological cooperative group behavior. Indications that the government as a body was expressing groupthink behavior were most evident in the first half of the crisis. The government initially met several times and, despite overwhelming evidence to the contrary, came to the conclusion that an economic crisis could not happen in South Korea. The perception of invulnerability they displayed, characteristic of groupthink behavior, may very well have come from the ruling party having successfully governed through an economic miracle growth of forty years. This previous groupthink behavior may have influenced the government publicly stating the denial that went into their group consensus perception of the situation, i.e., that the government would actually need to step in and do something about it. The government resisted accepting outside help from the IMF for so long was, in part, to facilitate the goal the ruling party had of winning the upcoming Presidential elections. Accepting the crisis as a fact and that they needed help would not strengthen their political position. By holding out, the government was trying to facilitate its own group’s goals. Finally, the government started to coordinate its policies with the IMF when it accepted the bailout package and a set of structural economic reforms that the IMF would oversee. Many of the practical solutions to the paralysis of the Korean economy, such as rolling over short term debts for companies and the government, also involved close coordination between the Korean government, the IMF, the USA, and major international bank creditors. There are a great number of cooperative and competitive or conflictive behaviors and statements that do not come across in this case. Among them are cooperative indicators that we saw in the previous case. For example, the facilitation of others’ goals and voluntary concessions were clearly present in the Kosovo crisis while notably absent here. However, we also did not see some of the conflictive indicators that the first case example showed, such as, the breaking of agreements, attributing fault to others, and credit-seeking for that which is perceived as going well. This lack of conflictive indicators is due, in part, to the fact that unlike the previous case the relationship between the key actors in this case was not an adversarial one. Consequently, we do not see some of the conflictive behaviors

52

Organizational Cooperation in Crises

that an IR perspective on cooperation would have us look for. It is also true that there were fewer things during the course of this case that were perceived of as going well. This clearly limited the number of things that the Korean government could seek credit for. The Korean government in this case could have worked to facilitate the attainment of others’ goals, like helping large corporations in their negotiations with creditors, but they were first looking out for themselves and achieving their own goals. When it was clear that the government would have to accept the IMF bailout, then it worked to get loans for major corporations, but more as an afterthought or part of getting the government’s economic plan back on track, rather than a genuine interest in the goals of others. A Psychological Perspective on Cooperation The contribution of the third tradition of cooperation research, coming out of social psychology, is a view of cooperation as the perceptions, attitudes, and behaviors of the organization. Psychological research on cooperation has focused on perception, stress, intra-group dynamics, risk aversion, and altruism in cooperative group behavior. Psychological conceptualizations of cooperation include: team collaboration, altruism, cohesion, reciprocity, helping behavior, pro-social behavior, social coordination, social capital and generalized reciprocity, voluntary participation in community improving projects, resisting free-riding, trust, generosity, and social capital. The case introduced below shows how this third tradition contributes to our conceptualization and understanding of cooperation in an actual crisis case. While the Red River flood disaster was not triggered by a psychological event per se (it was triggered by spring run off in the plains of the US and Canada) it illustrates the psychological underpinnings of the cooperation and conflict that took place among the organizations managing the crisis. Case Example 3: The Red River Flood General Description of Interaction across the Case This crisis involved actors in the USA and Canada at all three levels of government, and included the military, non-governmental organizations, and the press. The key interactions described in the case study took place between municipal governments in Manitoba, the provincial Emergency Management Organization (MEMO), the provincial Premier, the Canadian Armed Forces, and the Canadian Federal Government. While the crisis started out with a municipally led decentralized emergency response, as the flood waters increased and several states of emergency  ��������������������������������������������������������������� This case example is based on the case study by Svedin (1998c).

Crisis Cooperation in Light of the Three Traditions

53

were declared, the response in Manitoba shifted to one of centralized command and control. This shifted MEMO’s role in the crisis from one of coordinating and supporting local governments to commanding the municipalities and leading other organizations that would subsequently become involved (Svedin 1998, 43–7). The crisis was a highpoint in Canadian disaster experience in terms of its display of cooperation between the key managing agency MEMO, a multitude of organizations including: the media, volunteers and non-governmental organizations, the Canadian military, social services, the police, and the public (Svedin 1998, 46–50, 54–9). There were also clear expressions of competition and conflict at points in the crisis that became more pronounced as decision-making power was shifted from one level or organization to another. This also surfaced as diametrically opposed organizational management structures attempted to collaborate and support one another (Svedin 1998, 44–5). The sheer multitude of actors involved, even counting only the ones on the Canadian side of the Red River flood, coupled with the fragmentation of responsibility, caused aspects of the immediate flood management and recovery and reconstruction phases of the crisis to be prolonged and plagued with inefficiencies. A major success in the response to the flood was the deployment of the Canadian military to aid in a civilian emergency. This peaceful domestic use of the military domestically marked a major step beyond a national trauma that had haunted civil-military relations in Canada since the highly politicized quelling of separatist demonstrations in Quebec in 1970 (Svedin 1998, 28–30). Cooperation and Conflict in Decision-situations Individual decision-situations in the Red River flood case primarily see key actors commenting on the situation publicly, consulting with one another, and requesting things from other actors. Very few competitive and conflictive behaviors were displayed in key decision-situations during the crisis. The decision-situation that displayed the most cooperative behavior arose as the Manitoba Emergency Management Organization considered whether or not ask for help from the Canadian military. In this situation, the actors commented on the situation prior to the crisis by informing the Department of National Defence that a spring flood in the province may require military assistance. This occurred two months before the flood event. MEMO, who had monitored the flooding situation, consulted with City officials and approved a plan to cooperate with the military. MEMO then asked the province Premier to officially request the assistance of the military in managing the flood (Svedin 1998, 28–30). One of the least cooperative decision-situations, where the key actors actually decided to reduce normal relations with another actor, emerged as these actors considered when and how to pull the plug on support in the recovery phase of the crisis. MEMO, together with members from the social services Trauma Team, were stationed in trailers in two communities to provide support to local residents,

54

Organizational Cooperation in Crises

many of whom were still homeless. This support was planned to be in place for six weeks, but ended up remaining in place eight months. Toward the end of this period, MEMO and the social services were asking themselves, and consulting each other, when to pull out, knowing that these communities were most likely unable to survive on their own. Finally, team members stationed in these communities decided that returning to their own lives, and the way they had lived them before the flood, had to take precedence over the continued support for these community residents. MEMO then reduced what had become normal relations with these hardest hit towns and pulled their trailers out of their communities (Svedin 1998, 40). Cooperation and Conflict across the Case Across this crisis we find five largely cooperative behaviors and statements. Rallying around the flag, facilitating in-group’s goals, facilitating others’ goals, claiming honesty and actors coordinating with others. Seeing the impact of the flood on TV rallied administrative actors, decision makers, local residents, and volunteers around the common goal of working against the flood. This rallying effect extended to other recently flooded communities in other parts of the country and to the Canadian public at large. As a result, the Red Cross received more donations and offers to volunteer than ever previously registered in a Canadian disaster. Until the capital of the province was seriously threatened by the flood, provincial level decision makers were also very empathetic to the plight of smaller rural communities and rallied to support them. There was also a general disposition among decision makers in this case to help others and their own to achieve the goals they set out in the flood. Several organizations that do not generally cooperate, even in disasters, were helping each other reach common and individual group goals. For example, the Mennonite community took over sandbagging efforts from the Canadian Military. The military helped the police enforce security in evacuated areas, and the media facilitated emergency organization goals of broadcasting, printing safety information, and updates to the public. There were great efforts to meet and coordinate efforts between the numerous actors with authority to make decisions in the case. This was particularly evident in the early part of the flood when the decentralized decision-making structure required a coordination process between municipal governments. Later in the case, coordination took place primarily between political and operational decision-making groups regarding priorities, financing, and reconstruction. This case differs slightly from both of the previous case examples presented in this chapter. One of the most salient differences is the lack of competitive or conflictive behaviors and statements that stretched across the crisis. To the extent that conflict arose, it was in isolated incidents that were quickly addressed or that subsided largely by themselves. With regard to conflict indicators across the case, the Red River flood case is similar to that of the Korean Financial crisis, but different from the NATO intervention in Kosovo case. Like in the Korean case, the

Crisis Cooperation in Light of the Three Traditions

55

Red River Flooding case did not demonstrate the kind of adversarial relationship between the key decision makers in the crisis that the NATO intervention in Kosovo case had shown. This case, by its very nature, is different from the kind of actor confrontation situations and actors overcoming constraints to cooperation in a competitive environment that much of the international relations research focuses on. The Red River case also differs from the Korean financial crisis case in that is displays far more cooperative behaviors among the actors in the case. Helping behavior as a social phenomenon and a pro-social individual characteristic, as well as rallying around the flag and come to each others rescue as a psychological reaction to a perceived outside threat, are quite unique features to cooperation in this case. This illustrates that some forms of cooperation, and the reasons behind cooperative behavior and statements, would be missed in our understanding of crisis cooperation if we did not also utilize a psychological perspective. Each of the three research traditions that this study draws on enable us to more fully understand different aspects of crisis cooperation. As displayed in the case studies, the relationship between the actors in the case, the degree to which cooperation is shaped by structural and organizational characteristics, and the effect psychology has on cooperation work together to shape crisis cooperation. By drawing on these diverse research traditions we are able to capture the wide range of cooperative behaviors and statements that organizations actually display when interacting across time under crisis conditions.

This page has been left blank intentionally

Chapter 4

Organizational Behavior in Decision-Situations This chapter presents the first set of results from a categorical principal component analysis (CATPCA) of the MICRA data. This quantitative analysis provides an empirical map of underlying patterns and commonalities at the decision-situation level of analysis. The results reveal five types of cooperative organizational behavior. The first type contains a broad set of cooperative indicators while the second consists of primarily conflict oriented indicators. The third type involves uncommitted action, the fourth deals with decisive action such as yielding or rejecting, and the fifth behavior type presents value-based judgments towards other actors. Indicators Included An initial survey of the correlation between the indicators showed that the data does not have problems of extreme multicollinearity (< .9) or singularity (= .0). That is to say, there are no perfectly correlated indicators and none of the indicators completely lack correlation with other indicators. Consequently, all indicators that were specified in the previous chapter as the operationalization of cooperative behavior were included in the CATPCA analysis. Descriptive statistics of the indicators are presented in the table below (Table 4.1).

58

Organizational Cooperation in Crises

Table 4.1  Descriptive statistics for the decision-occasion level variables N

Demand Disapprove Reject Threaten Reduce relations Violence/Coercion Comment Consult Approval Decide to cooperate Request/Propose Make agreement Yield

Valid

Missing

295 295 294 294 292 295 268 294 299 302 300 296 294

31 31 32 32 34 31 58 32 27 24 26 30 32

Standard deviation .35684 .37879 .31621 .24015 .24092 .21298 .49350 .49168 .33735 .42055 .46978 .40017 .19820

Extraction Method The optimal scaling of the indicators included in the decision-occasion level CATPCA was identified by running and comparing the four different scaling options. The comparison revealed that the percentage of variance accounted for  ������������������������������������������������������������������������������ Because the scaling of the variables in CATPCA is not related to the original measurement scaling of the variables, the appropriate CATPCA scaling is found by simply running the different scaling options and comparing the percentage of variance accounted for (taking into account the restrictiveness desired and degree of interpretability). Multiple nominal scaling will always produce the closes fit to the data, but is not always the most desirable since it is unrestricted and sometimes is difficult to interpret. Using multiple nominal scaling and the fitting technique that CATPCA then applies, multiple correspondence analysis, also do not allow the transformed variables to be rotated (using ordinary PCA). This, again, makes interpretation more difficult. As a consequence, after comparing the different scaling options, it may be better to choose a scaling that produces a less optimal fit but that is more interpretable.  ���������������������������������������������������������������������������� CATPCA offers a number of scaling options for the variables included in the analysis. The scaling of the variables in CATPCA is unrelated to the original measurement scaling of the variables (i.e., categorical, ordinal or interval measured variables). The CATPCA procedure simply finds the best fit to the data, treating each observation as a separate category when it transforms the variables into standardized estimations. The optional scaling comes with increasing degrees of restrictions of how the procedure fits the data. Multiple nominal scaling is the least restrictive scaling. With this option CATPCA

Organizational Behavior in Decision-Situations

59

by the different scalings was very similar. Consequently, going with the most restrictive scaling option, numerical scaling, was preferable. In addition to the options of scaling, the options of treating missing values passively or actively were compared before running the analysis. This comparison showed that the active treatment of missing values did not add any information to the model. The missing values at the decision-occasion level of analysis were therefore treated passively in the CATPCA analysis. Five components were extracted based on the Scree plot criterion, Eigenvalues > 1, and on the criterion of variance accounted for by the solution. The CATPCA produces a scatter plot (of category points in the centroids of the associated objects) in a two dimensional space (a so-called Centroid Model representation) where clusters can be visually identified (Meulman et al. n.d.). The second least restrictive scaling is single nominal scaling. This type of transformation requires CATPCA to place the category points on a line going through the origin when it tries to find the best fit to the data. Ordinal scaling and numerical scalings are progressively more restrictive. With the ordinal scaling of the variables requires CATPCA to place the category points on a line (through the origin) but also to keep the assigned numbers of the categories in their original order as it goes through the process of fitting the data. Finally, numerical scaling requires CATPCA to place the category points on a line, keeping the original order of the numbers representing each category, and to keep the distance between each of the category numbers is equal, while finding the optimal fit to the data. The transformations that require CATPCA to fit the category points on a line through the origin produce Vector Models, where each line (vector) represents a variable’s fit through the data. The Spline options that CATPCA offers were not applied here as they are primarily appropriate when the test includes a very large number of variables (Linting et al. 2006b). The thirteen variables included in this test were not considered to be such a large number.  ���������������������������������������������������������������������������������� The multiple nominal scaling produces only a slightly better fit to the data than the other options, with the variance accounted being 33.6 percent for multinomial (total Eigenvalue 4.369) and 33.54 percent (total Eigenvalue 4.364) for the other scaling options. Since the difference in fit between the multinomial scaling (unrestricted) and the other scalings (restrictive) is only 0.06 percent it is reasonable to proceed with one of the restrictive scalings for the occasion for decision level of analysis.  ������������������������������������������������������������������������������ When missing values are treated passively observations with missing values on all variables are treated as supplementary variables and, as such, are excluded from the calculation of the optimal fit.  ������������������������������������������������������������������������������ When missing values are added to the model (imputed) as an extra category the value of additional information can be assessed by examining the category point plots for each variable. If the points for the imputed category of missing data frequently fall outside of or are clearly separated from the other category points it may be worth keeping the extra category so that the missing data information can be examined to patterns between these ‘missing cases’ can be pin-pointed.  ������������������������������������������������������������������������������� The CATPCA solution fitted to the occasion for decision data used the variable principle normalization. Furthermore, ofdrcd, the unique record number for each occasion for decision, was included in the variable list as a labeling variable for the object scores.  ��������������������������������������������������������������������������������� “In both linear and nonlinear PCA, the researcher has to decide on the number of components to be retained in the solution. One of the most popular criteria used to make

60

Organizational Cooperation in Crises

solution selected accounts for approximately fifty-four percent of the total variance in the data at the occasion for decision level of analysis. The model summary below (Table 4.2) specifies the Eigenvalues and the variance accounted for by each component (type of cooperation) in the selected five-component solution.

that decision is the scree criterion” (Linting et al. 2006b, 17). In order to evaluate the scree criterion a scree plot (line graph) of the components (x-axis) and the associated eigenvalues (y-axis) can be requested in SPSS. Ideally, “such a plot shows a break, or an ‘elbow’, between a steep slope for the components that contribute considerably to the VAF [variance accounted for], and a faint slope (‘scree’) for the components that contribute only very little to the VAF. The location of the elbow gives an indication of the optimal number of components to be included in the solution” (Linting et al. 2006b, 17). However, the scree plot does not always produce such a clear break, which is the reason why the visual Scree plot criterion often is combined with a numerical comparison of the variance accounted for in the models. The scree plot of the occasion for decision level data shows one ‘elbow’ or breakpoint where the slope of the line flattens out at component 3 and another at component 5 where the line starts then dropping drastically again. Since it is not entirely clear from the scree plot which of the two cut-off points was most appropriate, I examined the specific percentages of variance accounted for and the eigenvalues that the different solutions provide. Based on the scree plot ‘elbow’ (indicating a cut-off at 3 components and one at 5 components) and the overall variance explained by the components (largest for the 5 component solution) it made the most sense to specify five components in the analysis. “In linear PCA, the eigenvalues and the component loadings are obtained from the correlation matrix between the variables. Because this correlation matrix is fixed, the components are nested, meaning that the first p components of a p+1-dimensional solution are the same as the components of a p-dimension solution. A nonlinear PCA with all of the variables treated numerically equals a linear PCA, and therefore the solution in such a case is also nested. However, when not all variables are treated numerically – as is usually the case -, the correlation matrix is not fixed, but has to be estimated, because it is based on optimally quantified variables. Given the optimal quantification requirements, nonlinear PCA maximizes the first p eigenvalues of the correlation matrix. Therefore, the components in nonlinear PCA are not nested, because optimal quantification depends on the dimensionality p that is chosen (i.e., how many eigenvalues are maximized). Therefore, when nonlinear PCA is applied, it is advisable to look at solutions with p-1 and p+1 dimensions as well, when p is the chosen number of components” (Linting et al. 2006b, 17).  ���������������������������������������������������������������������������������� The percentage of variance accounted for in a model can be calculated by dividing the total Eigenvalue of the model by the number of variables included in the model. Under certain conditions, however, this procedure does not provide an exact measure of the percentage of variance accounted for.

Organizational Behavior in Decision-Situations

61

Table 4.2  Eigenvalues and the variance accounted for by each component in the component solution for the decision-occasion level data Model summary – 5 components Component 1 2 3 4 5 Total

Cronbach’s alpha .603 .570 .313 .257 .238 .954

Variance accounted for Total (Eigenvalue)

% of variance

2.254 2.108 1.406 1.311 1.281 8.360

12.275 11.480 7.657 7.139 6.976 54.460

Component Loadings after Rotation The principal component analysis provides a score (component loading) for each variable on each of the five underlying patterns of cooperation (components). In loose terms it can be said that these scores tell us how strongly associated each indicator is with the pattern that has been identified. Because we have asked the CATPCA to score each indicator on each component, we now need to determine for which cooperative behaviors (components) it makes sense to analyze the indicators’ contribution. A variable will have scores on all components but there is a cut-off level for significant scores. We also need to take note of whether the score is positive or negative. If the score is negative, it may still be significant, but it means that the absence of the particular variable in a cooperative behavior is a robust (significant) finding. The component loadings for each indicator in the rotated solution are displayed in the following table (Table 4.3). The < .4 criterion was used as the cut-off value for significant indicator loadings and indicators that scored higher than .4 were

 ��������������������������������������������������������������������������� By saving the transposed variables and their component loadings the CATPCA solution can be rotated in the regular PCA procedure in SPSS. The CATPCA procedure in SPSS (version 14) does not allow for rotation, however, the option of saving the transposed variables makes rotation possible using the more elaborated PCA procedure. I have chosen a so-called orthogonal (uncorrelated) rotation, the Varimax rotation, because the variable observations become standardized as CATPCA produces the optimal solution for the variables in the model. If I would have used a rotation technique that allowed for correlation between the components (oblique rotation), such as Direct Oblimin, the optimal fit created by CATPCA would have been lost.

62

Organizational Cooperation in Crises

then included in the analysis of the components.10 Indicator scores lower than .4 have been shaded out (since they will not be included in the analysis) to make the table more reader friendly (Table 4.3).11 Table 4.3  Varimax rotated solution of transposed decision-occasion level variables Rotated component matrix Component

Demand Disapprove

1

2

3

4

5

.627 .605

.069 -.120

-.095 .198

.088 -.063

-.152 .425

Reject

.317

-.321

.102

.507

-.160

Threaten

.673

.003

.155

-.016

.084

Reduce relations

.783

.034

-.052

-.003

.091

Violence/Coercion

.272

-.102

-.215

.045

.512

Comment

.175

-.016

.735

-.122

-.109

Consult

-.078

.076

.702

.184

.179

Approval

.135

.653

-.209

-.130

-.047

Decide to cooperate

-.067

.763

.175

.048

.125

Request/Propose

-.065

.171

.181

-.024

.716

Make agreement Yield

.005 -.004

.574 .151

.338 .095

.430 .677

-.079 -.096

Note: Rotation method: Varimax with Kaiser Normalization. Rotation converged in 6 iterations.

Discussion of the Decision-occasion Level Components The categorical principal component analysis tells us that there are five types of cooperative behavior at the decision-occasion level of the crises examined. In the 10 ������������������������������������������������������������������������������������� This criterion for assessing the significance of a component loading (the variable’s relative contribution to the component) is based on Steven’s critical values table (Field 2005, 637–8). Component loadings of .4 explain approximately 16 percent of the variance in a variable. 11 ���������������������������������������������������������������������������������������� I used the rotated solution of the model as the basis of the analysis as it is slightly easier to interpret than the unrotated version. In the rotated solution the transposed variables are reconfigured in such as way that they load most strongly onto one component (leaving the components uncorrelated) and thereby simplifying the interpretation while keeping the most important relations between the variables in the model.

Organizational Behavior in Decision-Situations

63

sections below each component, and the indicators that load onto the component, are discussed in more detail. Behavior 1 – Fight Reduce relations Threaten Demand Disapprove Four indicators make up this component, which clearly illustrates strong conflictual behavior. I term this decision-occasion component fight. The fact that this component has the largest number of significant indicator loadings is partly an effect CATPCA maximizing the fit on the first component in any solution. The indicators reduce relations and threaten, both load strongly onto the component. Threaten, shows the decision unit verbally expressing a threat of action or a coercive warning toward another actor, and reduce relations represents the decision unit actually moving into action and reducing what would be normal or routine relations with another actor. The fourth indicator in the component is demand which represents decision units issuing orders, commands, or decrees toward another party. This type of behavior differs from a request or an appeal in that is it backed by a confrontational demand asserting the right of the decision unit in requesting the other actor to comply. Disapprove has the weakest of the significant loadings, indicating that the variance it adds to this component is not neatly clustered but rather is spread out. This indicator illustrates that the decision unit expresses disapproval, objections, or complaints toward another actor. Now that we know what indicators make up the type of decision-situation behavior I have termed fight, it may be helpful to go back to the qualitative crisis case studies that the data in the TCM dataset has been extracted from. The case studies can provide a narrative example of what a strong example of this type ‘cooperative’ behavior looks like in real life. The narrative example (Table 4.4) describes a particular occasion for decision that contains all or many of the indicators (reduce relations, threaten, demand, disapprove) that make up the component fight.12

12 �������������������������������������������������������������������������������������� Other case examples of fight as a ‘cooperative’ behavior in decision-situation can be found in; occasion for decision (OFD) 184 How should the US deal with the fragmentation of Iranian authority and who can it negotiate with to solve the crisis? (Reagan 2003); OFD 181 Should the US consider developing military offensive options to expel Iraqi forces from Kuwait? (Arana 2004); OFD 196 Should Ukraine seek international support? (Piskalyuk 2004); OFD 325 How should Vajpayee government respond to potential threat

Organizational Cooperation in Crises

64

Table 4.4  Case example of fight Case 13

Carter and the Iran Hostage Crisis (Reagan 2003)

OFD 183 2. Iranian authorities approve of seizing the embassy.   What strategy should the US follow? “The Carter Administration developed a strategy to compel the Iranians to release the hostages. The objective was to get the hostages released, without resorting to acts of war or undercutting the nation’s standing in the world. The Carter Administration did not want to capitulate to Iranian demands nor employ military force. Ultimately the President chose a strategy of negotiation while simultaneously imposing economic sanctions. The purpose was to isolate Iran from the rest of the world and apply pressure through economic means.” (Reagan 2003, 32) The strategy that the President finally presented to his council at Camp David included the following: “Isolate Iran from the rest of the world. Get other friendly nations to condemn Iranian actions. Get the Shah [of Iran] to leave [the US] as soon as practicable. Impose an economic embargo on Iran and get allies to also support it. Get the UN to impose sanctions on Iran. Mine Iranian harbors and threaten stronger punishments if the hostages are punished. Prepare for military action against Iran.” (Reagan 2003, 32)

Behavior 2 – Agree Decide to cooperate Approval Make agreement The second component shows positive action and generally cooperative behavior. I term this the second type of cooperative behavior at the decision-occasion level of analysis agree. The three indicators with significant loadings are decide to cooperate, approval, and make agreement. These represent decisions to cooperate. The strongest loading indicator, decide to cooperate, represents the decision unit’s decision to initiate, resume, improve, or expand formal relations with another actor. The second indicator making up this component, approval, signifies the decision unit to democratic India? (Pathak 2004); OFD 326 How should New Dehli pressure Islamabad to act against jihad groups supported by military government of Pakistan? (Pathak 2004).

Organizational Behavior in Decision-Situations

65

expressing approval of or support for another party in the crisis. The third and final indicator of this component, make agreement, represents agreements being made between the decision unit and another actor(s). This type of agreement usually refers to reciprocal agreements between at least two parties, but may occasionally take the form of unilateral concessions or future commitments that are not reciprocal. Table 4.5  Case example of agree Case 15

Daewoo Group’s Crisis Settlement (Kim 2004)

OFD 190 2. Persuading Financial Institutions “The Korean government worried about the massive redemption of bonds issued by Daewoo companies at the same time. The Korean government was seeking a way to extend the expiration of the bonds issued by Daewoo and providing new loans for Daewoo. Actually, the twenty-one banks involved no longer wanted to lend money to Daewoo, even though they may not get back the money lent to Daewoo before. The Korean Government, however, had different plans for the banks. The [Financial Supervisory Commission] FSC convened the meeting of banks’ chairmen for dealing with Daewoo problems on July 30, 1999 after Daewoo’s financial crisis. At the meeting, government officials explained the principles for the use of public funds to support Daewoo and requested the extension of the expiration of the bonds. The principles were as follows: i. All banks have to help Daewoo for the national economy in this situation. ii. If banks have a problem of liquidity, the government will help their difficulties at will. Controversy in the banking community was growing in the wake of an agreement to provide four trillion won (US$ 4 billion) in new loans to the financially troubled Daewoo Group. While major Daewoo Group creditors, mostly banks, argued that an immediate injection of new capital was necessary for the survival of the nation’s secondlargest conglomerate, others, including investment and trust company strongly opposed the idea. One bank chairman murmured ‘We were infused with public funds last year. So we can’t invest to ailing business groups like Daewoo. But there is no way except following the government’s decision.’ All chairmen from the banks who attended the meeting agreed with the government’s proposal to support Daewoo. However, trust and investment companies said that it was not fair for them to take on most of the burden of providing new loans simply because they possessed seventy-seven percent of commercial papers (CPs) and corporate bonds issued by Daewoo Group. Also they argued that ‘We can not afford to provide another package to Daewoo Group without undermining our existing business clients. It is simply unfair that we have to take on this burden while other major creditor banks are almost free of this.’ In this situation, the Financial Supervisory Commission (FSC) persuaded financial institutions who refused the guidelines of the government saying that ‘rules are rules and investment and trust companies should respect the financial authority’s guidelines.’ Finally, sixty-nine creditors decided to allow the rollover of the short-term loans for another six months and provide new loans. As a result of these trials, the massive redemption of bonds issued by Daewoo companies had passed without having any serious impact on the financial market, at least superficially. In fact, the chaos that had been predicted by many market observers did not happen. The government intervened to prevent any serious market disruption triggered by massive redemptions of bonds issued by the Daewoo group.” (Kim 2004, 15–7)

66

Organizational Cooperation in Crises

Being familiar with the three indicators that make up the type of decision-situation behavior I have termed agree, we now turn to the qualitative crisis case studies and look at what a strong example of this type cooperative behavior looks like. The decision-occasion described previously (Table 4.5) is a strong example of agree, containing all or many of the indicators (decide to cooperate, approval, make agreement) that make up the component.13 Behavior 3 – Talk Comment Consult When compared with the previous behavior, this component shows a form of cooperation that is both narrow and weak. This component demonstrates the decision unit and other actors just talking, I therefore I term the component talk. Comment, as an indicator, signifies the decision unit making some form of verbal statement concerning the crisis toward an outside audience of stakeholders. The second indicator that makes up component 3, consult, simply refers to any consultation or meeting taking place between the actors regarding the crisis. Now that we know the variables make up talk as a decision-situation cooperation behavior, we once again turn back to the qualitative crisis case studies that the TCM dataset builds off and look at what a clear illustration of this type behavior of looks like.14 The excerpt included in Table 4.6 is a decision-occasion that represents a strong example of what talk as a component looks like in an actual crisis case.15

13 ������������������������������������������������������������������������������ Other case examples of agree as a cooperative behavior in decision-situations can be found in occasion for decision (OFD) 180 What should the US diplomatic and military response be to the Iraqi invasion of Kuwait? (Arana 2004); OFD 69 How will the government react to the continuing disorder and international criticism? (Como 2003); OFD 19 President asks Board of Directors to hire new interim Executive Director (Hoeschele 2003). 14 ��������������������������������������������������������������������������������� A clear illustration, or strong example, is an illustration where the particular occasion for decision contains positive observations on both the variables that make up the component. 15 ������������������������������������������������������������������������������������ Other examples of talk as a cooperative behavior in decision-situation can be found in occasion for decision (OFD) 4 How to deal with power vacuum left by retiring executive director? (Hoeschele 2003); OFD 67 How should the Albanian Government react to the warnings put out by the IMF and the World Bank? (Como 2003); OFD 35 How should the district administration return and rehabilitate the riot victims? (Gajbhiye 2003); OFD 42 Should the cabinet take action as the crisis worsened?; OFD 43 Should the cabinet accept the IMF proposal for reviving the Korean economy?; OFD 45 Should the cabinet negotiate with global creditors? (Park 2003).

Organizational Behavior in Decision-Situations

67

Table 4.6  Case example of talk Case 3

Accident at Three Mile Island (March 2003)

OFD 15 Decision Occasion 1: At 7:00 a.m. on March 28, 1979, TMI supervisors declare a site emergency. At 7:25 a.m. this action is upgraded to a general emergency. By 7:45 federal and state government officials have been notified of the emergency. How should the government respond? “The first occasion for decision occurred on day one (Wednesday, March 28th) of the crisis after a general emergency had been declared at the TMI [Three Mile Island] plant. Federal, state, and local government representatives functioned as a coalition of autonomous actors in this early stage of the incident.” (March 2003, 1–2) “Federal officials were the first on the scene at the nuclear plant. A total of five NRC [Nuclear Regulatory Commission] inspectors was dispatched early on March 28th and arrived at TMI shortly after 10:00 a.m. They spend most of the day surveying the conditions at the plant, attempting to establish communications with NRC regional and national headquarters, and assisting the plant operators in bringing the maligned reactor under control. State and local officials located offsite were forced to communicate with the NRC personnel through woefully inadequate telephone lines. The inability to communicate effectively between government officials at all levels made this coalition a very loose and ineffective one.” (March 2003, 12) “The resulting decision outcome was one of fragmented symbolic action. Lacking accurate information, the NRC inspectors at TMI and those at regional and national headquarters could not agree on how to bring the reactor under control and whether efforts on the first day of the crisis did in fact make the reactor safe. Therefore, the NRC could not make confident recommendations to state and local authorities as to whether an evacuation was needed. State and local government faced a similar situation. Information was so hard to obtain that a decision on the necessity of evacuation could not be made.” (March 2003, 1–2)

Behavior 4 – Negotiate Reject Make agreement Yield The fourth component represents a definitive action to either reject or yield to another party. I call this component negotiate. Two of the three indicators that make up this component, reject and yield, are almost diametrical opposites. So the component displays conflictual behavior, with the decision unit rejecting the actions, statements, regulations, or norms of another actor. It also, however, shows cooperative behavior where the decision unit yielding or conceding to another actor. It should be noted that yielding is different from making promises or commitments to future action in that yielding entails present action of a strong cooperative nature. Together the two indicators

68

Organizational Cooperation in Crises

present a definitive action, either conflictual or cooperative, where another actor is either rejected or submitted to. Table 4.7  Case example of negotiate Case 77

Breakdown of Pastrana’s Peace Process (Santander and Matos 2003)

In this case, the President has clearly stated in front of millions of Colombians that he is prepared to take the demilitarized zone (DMZ) by force and he thereby has stated a strong ultimatum to the FARC guerillas, with whom the peace process is rapidly falling apart. This is the set-up of the decision occasion that exemplifies Component 4 – Negotiate. OFD 304 Should President Pastrana allow the participation of the UN Special Advisor and extend the deadline 48 more hours? “After such a flash of bravery by the President of Colombia, his desire to use the force was curtailed by the International Community that had played a decisive role in facilitating and accompanying the process. It was obvious that after three years of governing a large sector of the national territory, the FARC was not going to give up so easily its main accomplishment in the negotiation: therefore, it drew upon intermediation of the International Community, the United Nations and the Episcopal Conference. In order to continue to strengthen his international image, Pastrana [the Colombian President] accepted the intervention of the United Nations Delegate, James Lemoyne, the Ambassador of France, Daniel Parfait, and the President of the Episcopal Conference, Monsignor Alberto Giraldo. The President extended the deadline for 48 hours more for the DMZ to be cleared, making January 12 as the new date and, afterwards, due to the intervention by the delegates, he announced January 14 as the last deadline. They [the President and the international representatives] had continuous meetings and the task of the facilitating commission was tireless, but always with the firm conviction and having as its main objective the protection of the civilian population. Finally, they managed the FARC to commit themselves to new agreements, but pointed out that first they wanted to hear the political parties’ opinion concerning the process and the country’s status. Upon this new proposal, the President announced at midnight on January 12 again before the news media, that he did not agree with the conditions imposed by the FARC, and that therefore, on January 14 at 9.30 p.m. the army would retake control. Despite his assuredness, President Pastrana authorized the delegates to meet once more with the representatives of the guerilla group in San Vicente del Caguan. Thus, five hours before the expiration of the final deadline, the delegates announced that the President reiterated the safeguards for the continuation of the negotiations, that the FARC accepted the safeguards and that both parties had declared their determination to adopt the points included in the San Francisco de la Sombra Agreement. Taking the final position was a complicated process; several actors were involved, but the power of the future of the situation was concentrated in one sole person: President Pastrana, the political position he had assured again prevented him from showing himself weak before the enemy, if he changed his initial position, he could be subject to new requirements from his counterpart [FARC].” (Santander and Matos 2003, 14)

Organizational Behavior in Decision-Situations

69

The third indicator that makes up this component, make agreement, represents agreements being made between the decision unit and another actor(s). Agreement are usually reciprocal and between at least two parties, but may occasionally take the form of unilateral concessions or future commitments. In Table 4.7 we take a look at what this component, negotiate, looks like in an actual crisis case. The excerpt included in Table 4.7 is a strong example16 of negotiate, including most or all of the indicators (reject, make agreement, yield) that make up the component.17 Behavior 5 – Manipulate Request/Propose Violence/Coercion Disapprove In terms of analysis, the fifth component is perhaps the analytically most challenging. The component entails requests together with structural violence or coercion and disapproval. I term this component manipulate. The indicator request/propose is a cooperative behavior in decision-situations where the decision unit proposes, suggests or appeals to one or more actors. The second indicator making up this component, structural violence/coercion, indicator is the most conflictual behavior at the decision-occasion level of analysis. It represents the decision unit using government-sponsored oppression and or violence against civilians, their rights, or their property. Lastly, disapproval reveals the decision unit expressing objections and or complains toward another actor in the crisis. Having determined what three indicators make up manipulate, we turn to the actual case studies to find an illustration of this component in action. The excerpt included in Table 4.8 is a decision-occasion that represents a good example of negotiate, as it contains several or all of the indicators (request/propose, violence/ coercion, disapprove) that make up the component.18 16 ���������������������������������������������������������������������������������� A strong example of this component is an example in which the particular occasion for decision contains positive observations on all or many of the variables included in the component. 17 ���������������������������������������������������������������������������������� Other case examples of negotiate as a cooperative behavior in decision-situations can be found in occasion for decision (OFD) 218 How should NATO respond to Yugoslav government’s acceptance of OSCE verifiers on the ground? (Stepanovic 2003b); OFD 116 Hazelwood violates Exxon’s alcohol policy; why does he flee Alaska? (Watson 2003). 18 ��������������������������������������������������������������������������������� Other examples of manipulate as a cooperative behavior in decision-situation can be found in OFD (occasion for decision) 183 How should the US respond to the Iranian appeal of the embassy seizure? (Reagan 2003); OFD 43 Should the cabinet accept the IMF proposal for reviving the Korean economy? (Park 2003); OFD 224 Believing the standoff could continue, FBI officials present a plan to end the standoff with Attorney General Janet Reno, which is based on the use of CS gas to drive the Branch Davidians from the

70

Organizational Cooperation in Crises

Table 4.8  Case example of manipulate Case 77 Breakdown of Pastrana’s Peace Process (Santander and Matos 2003) After a third extension of the deadline for FARC in the negotiation of the peace process occasion for decision four appears, illustrating OFD Component 5 – Manipulate. OFD 66 Should Pastrana declare the end of the Peace Process after the hijack[ing]? “After the extension of the given deadline for the DMZ [demilitarized zone], a new hope appeared in the Administration’s negotiation team, Pastrana’s [the Colombian President’s] new objective was to enforce a new timetable to propose the possibility of a ceasefire. On Febuary 20, 2002, however, the panorama changed completely, when the members of the Military Forces informed the President that an Aires Company airplane had been hijacked; the President immediately changed his policies. The President’s search of information to confirm FARC’s responsibility in this incident was the first sign that the course of events would suffer a shift. Additionally, and in line with his approach, he asked the Armed Forces for evidence and proofs of outrages committed in the DMZ. Consequently, the President, in an autonomous way, assumed the responsibility for whatever direction the negotiations would take, pointing out empathically to his generals and ministers that: ‘I will solve this problem. There is no one or anything that can change what I am thinking. And as it is I who has to take the decision, I wish to tell you that today I will break the process with the FARC.’ In view of such determination there was no step back, no matter what the guerilla fighters did or said, he had already made his decision regarding the facts and, for that, he acted as a predominant leader who acted in accordance with the political and legal principles of his country. The measures were well proven; as he ordered to immediately reverse the political status that had been granted to the FARC guerilla to negotiate. Furthermore, he requested that the orders to capture the heads of the subversive organization be reactivated. His [the President’s] aim and objective are well defined. His purpose was to end with the DMZ and for this he appealed to the legal rules that not only gave him security, but also credibility before the national and international opinion. Furthermore, he was also supported by the security corps making use of the right that, in accordance with the Colombian Constitution, appointed the State as the sole and legal holder of the monopoly of power.” (Santander and Matos 2003, 15–6)

Indicators that Cut across More than One Component A number of indicators appeared with significant scores in more than one component in the unrotated solution. As opposed to the competitive or conflictive indicators, these indicators are all among the more facilitative under the umbrella of cooperation. Three indicators of cooperative behavior, comment, consult, and approve, cut across three components. One indicator, request/propose, cut across two components. compound. Should the plan be approved? (Koester 2003); OFD 115 How to deal with severe weather hampering cleanup? (Watson 2003).

Organizational Behavior in Decision-Situations

71

What this cross-cutting feature indicates is that these indicators are particularly salient in cooperative organization behavior in crises. In other words, cooperating in crises often involves representatives of organizations acknowledging what has happened in the crisis by publicly commenting on it. In connection with this, we can note that expressing approval of another party involved in the crisis is a common feature of crisis cooperation in decision-occasions. This means that one way of showing cooperative attitude is praising the actions and work of a collaborating partner to an external audience (the public, the media or other organizations). Crisis cooperation in specific decision-situations also frequently takes the form of actors involved in the crisis setting up meetings to talk about what is happening. In summary, the three most prominent characteristics of crisis cooperation in decision-situations demonstrate the organizations talking about the crisis to each other and outside audiences. The last indicator, which is somewhat less prominent than the first three, shows an actor taking a positive action by presenting a proposal, suggesting a way forward, or making a specific request for help to another party. The behavior displayed in these cross-cutting indicators show organizations engaging in tempered actions and expressing both positive and collaborative attitudes rather than conflictual manners. Having established that organizations display five types of cooperative behavior in crisis decision-situations, the next step is to empirically examine what cooperative strategies organizations pursue during crises.

This page has been left blank intentionally

Chapter 5

Cooperative Strategies across Crises Chapter 5 puts forward the second set of results of the empirical quantitative study. The case level of analysis the CATPCA identifies four underlying patterns. The first pattern is related to structural and organizational variables. The second represents psychological reactions to the crisis characteristics. The third case level pattern represents actors being earnest and responsible, and the fourth illustrates cooperation that takes place in situations of perceived management success. Together these patterns make up four complex cooperation strategies that organizations pursue across a crisis case in its interactions with other actors. Variables Included A preliminary scan of the indicators selected at the case level of analysis for the CATPCA reveals that two of the indicators, new group syndrome and game of Old Maid, have missing correlations across all other variables. These indicators therefore have to be excluded before the principal component analysis can be conducted. This exploratory statistical model builds on the assumptions that there is shared variance among the indicators (Field 2005, 641). The indicators included in the analysis and their descriptive statistics are displayed in Table 5.1. Extraction Method The CATPCA of the case level cooperation variables was conducted using nominal scaling of the variables. Missing data at the case level of analysis was  ������������������������������������������������������������������������ When approaching the component analysis using CATPCA for the case level measures of cooperation we again have to make decisions about the most appropriate scaling of the variables. Because the scaling of the variables is not related to the original measurement scaling of the variables the appropriate scaling is identified by running and comparing the different scaling options (with regard to the variance accounted for). Multiple nominal scaling for all variables, as we would expect because it is the least restrictive scaling option, provide the best fit in the variance accounted for. The graphical display of the variable scores using this scaling shows that there is a nonlinear relationship between the case level variables included in the model. The ability to reveal non-linear relations in the data is one of the advantages of CATPCA compared to regular PCA. The second best fit (having compared all the scaling options) is single nominal scaling of the variables with a principal variable normalization. When comparing extractions with two

74

Organizational Cooperation in Crises

actively treated as an extra category and the solution was rotated using the uncorrelated Varimax rotation. Table 5.1  Descriptive statistics of cooperation at the case level N Valid

Missing

Standard deviation

Rally around the flag Groupthink

20 21

46 45

.51299 .48305

New group syndrome

17

49

.00000

Blame-game

21

45

.49761

Game of Old Maid

17

49

.00000

Credit seeking behavior

44

22

.52223

Facil. in-group’s goals

45

21

.58344

Facil. other’s goals

45

21

.61381

Honesty and self-disclosure in interaction

22

33

.42893

Claim self to be honest

38

28

.36945

Work to coordinate efforts

59

7

.39280

Make voluntary concessions

58

8

.47343

Need for contact person(s)

61

5

.24959

Break agreements

61

5

.34036

Stalling tactics in decision implementation

63

3

.24580

Content slippage in decision implementation

63

3

.27248

Attribution of fault to other people Evidence of behind the scenes (covert) behavior

64 61

2 5

.80672 .40150

components the change in the total Eigenvalue is 0.665 (6.939–6,274 = 0.665), representing a 4.2 percent change in the total variance accounted for (from approximately 43 percent to 39 percent of the variance in the data). Taking into account the desire for restrictiveness and interpretability, the loss in variance accounted for with the nominal scaling is acceptable and the CATPCA can be conducted with single nominal scaling and Varimax rotation.  �������������������������������������������������������������������������� There are two ways of dealing with missing values in CATPCA, passively or actively. Both options were evaluated before proceeding with the analysis. When missing values were imputed by creating an extra category the fit of the data proved better than when missing data was treated passively. Furthermore, the variable plots tests showed that the extra category of missing values often end up quite separate from the other categories, indicating that it may worth saving this category separately to explore common traits between the missing data categories. The CATPCA solution fitted to the case data used variable principle normalization. Furthermore, Caseid, the unique record number for each crisis case, was included in the variable list as a labeling variable for the object scores.

Cooperative Strategies across Crises

75

The identification of the number of components at the case level was based on the Scree plot criterion, Eigenvalues > 1, and Steven’s recommendation. The first two criteria showed that the model has a good statistical fit for three, four and five components, respectively. However, in the model with five components only two of the indicator loadings meet the .4 criteria of significance in the last component. As a result, the model with four components was selected. The CATPCA solution selected, accounts for approximately sixty-eight percent of the total variance in the data at the case level of analysis. The individual contribution of each component to the overall variance accounted for is displayed in Table 5.2.

 ���������������������������������������������������������������������������� There are three ways of determining the appropriate number of components to extract in CATPCA. The first two are based on statistical considerations and the third is based on interpretability. Determining the number of components that is used to analyze the data is based on a joint assessment of these three criteria. The statistically based criteria are Kaiser’s criterion of keeping components whose eigenvalues are larger than 1, and the Scree plot ‘elbow’ criterion. Kaisers criterion tends to be the most accurate when the number of variables in the analysis is less than thirty (Field 2005, 633). Consequently, Kaiser’s criterion was therefore appropriate for this principal component analysis. In order to determine how many components should be extracted I examine a Scree plot with the maximum number of components extracted (where the number of components equals the number of variables included). The Scree plot illustrated two ‘elbows’ where the line seemed to flatten out. This indicated that there were two natural cut-off points to the number of components.  ���������������������������������������������������������������������������� The recommend minimum number of loadings onto a variable is three (with a < +0.4 or > -0.4 loading). The .4 criteria is based on Stevens’ recommendation cited in Field (2005, 659).  ���������������������������������������������������������������������������������� The first cut-off point is at three components and the second cut-off point is at five components. In order to determine which solution gives the best fit I examined the Eigenvalues for solutions with three, four, and five components. The examination of the models showed that all three solutions give the components eigenvalues larger than 1. All three models also accounted for a large part of the variance in the data, and the components in the tree models fit rather well together judged on the Cronbach’s Alpha. All three models therefore work well, based on the statistical criteria, in terms of representing the data. Because I could choose either model as a good statistical fit, I used the third criteria for component selection. Both the model with 4 and with 5 components provide valuable insights into the clustering of the variables when we look at the component loadings, however, in the model with 5 components only two loadings meet the .4 criteria of significance. The recommend minimum number of loadings onto a variable is three (with a < +0.4 or > -0.4 loading). The .4 criteria is based on Stevens’ recommendation cited in Field (2005, 659). Therefore the model with 4 components, which makes sense both statistically and from an interpretation point of view, was selected for analysis.

Organizational Cooperation in Crises

76

Table 5.2  Summary of the cooperation components extracted at the case level Model summary – 4 components Components

Variance accounted for

Cronbach’s alpha

1 2

0.841 0.656

% of variance

Total (Eigenvalue)

4.735 2.597

29.592 16.233

3

0.486

1.836

11.473

4 Total

0.415 .968*

1.636 10.804

10.225 67.522

Note: * Total Cronbach’s Alpha is based on the total Eigenvalue.

Table 5.3  Varimax rotated CATPCA solution of case level variables Rotated component matrix(a) Component 1

2

3

4

Rally around the flag Groupthink

-.156 -.161

.937 0.932

-.009 -.036

-.088 -.039

Blame-game

-.035

.935

.021

-.110

Credit seeking behavior

-.125

-.024

-.365

.544

Facil. in-group’s goals

.087

-.114

.066

.836

Facil. other’s goals Honesty and self-disclosure in interaction Claim self to be honest

.270

-.284

.359

.596

.211

-.046

.725

-.072

-.084

.052

.728

.078

Work to coordinate efforts

.759

-.176

.139

-.064

Make voluntary concessions

.665

.098

.076

.233

Need for contact person(s)

.731

-.082

-.414

-.051

Break agreements Stalling tactics in decision implement. Content slippage in decision implement Attribution of fault to other people

.616

-.285

.113

-.437

.656

-.158

-.247

.216

.886

-.215

-.123

.097

.413

.072

-.754

.022

.627

.001

-.003

.318

Evidence of behind the scenes (covert) behavior

Note: Rotation Method: Varimax with Kaiser Normalization. Rotation converged in 5 iterations.

Cooperative Strategies across Crises

77

Component Loadings after Rotation Looking at how the different variables load onto the four components in the rotated solution, we find that component one consists of eight positively loaded indicators. Component two consists of only three positively loaded indicators. Component three is made up of two positively and two negatively loaded indicators, and component four has three positive and one negative indicator. Component scores lower than .4 are not included in the analysis, and have been shaded out to make the table easier to read (Table 5.3). In the analysis of the components below I use this rotated solution, which is easier to interpret than an unrotated solution. Discussion of the Strategies across Crises The categorical principal component analysis of the case level indicators of cooperation strategies yielded four components. The components and the variables that load onto them are discussed more in-depth in the sections below. Strategy 1 – Bureaucratic Politics Content slippage in decision implementation Work to coordinate efforts Need for contact person(s) Make voluntary concessions Stalling tactics in decision implementation Evidence of behind the scenes (covert) behavior Break agreement Attribution of fault to other people This first case level component illustrates a decision implementation situation (both in terms of content and the time it takes to implement decisions) that is riddled with problems. And while many of which seem structural in nature, others can be considered bureau-political. I term this first component bureaucratic politics. Looking at the indicator loadings, the strongest loading variable is content slippage in decision implementation. This indicator shows that the substantive implementation of decisions is not what the decision makers had in mind. If this subversion of the implementation process is intentional, it means that decisions are purposefully being watered down or corrupted in their intended function by those who are charged with applying them. Content slippage is then a sign that the  ���������������������������������������������������������������������������������� The comparatively large number of variables loading onto component 1 is partly an effect of CATPCA maximizing the fit on the first component in the solution.  ������������������������������������������������������������������������������������ The shaded loadings do not meet the .4 criteria and are not include in the analysis.

78

Organizational Cooperation in Crises

policies that decision makers are agreeing on are being resisted by their own streetlevel bureaucrats or operational partners (organizations) engaged to carry out the policies. In this type of situation stalling tactics in decision implementation also makes sense. This indicator is also present in component one, but the connection this indicator has to the component is somewhat weaker. This indicates that it does not fit as well into the overall concept as some of the other indicators such as content slippage. Stalling tactics in decision implementation implies attempts to subvert the outcome of a decision process by means of delaying its completion. As in the case with the intentional subversion of decision content in the implementation phase, purposefully delaying the implementation of decision can be interpreted as policy-makers and policies meeting resistance from the bureaucracy. However, decisions gone awry in the implementation phase are not necessarily the result of intentional corruption, however. Content slippage can result from of a failure to communicate the purpose of the decision and its contents properly. In this case we have something similar to the ‘whisper game’ where the content of a message gradually gets distorted as it is passed on from one actor to the next or down the chain of command. This is certainly be true if the organizations involved are dealing with a complex crisis with many interdependencies such as technological crises with multiple connections and tight couplings (e.g., Perrow 1984), which are naturally difficult to communicate in simple messages. This is also true when the management of the crisis engages organizations that operate according to radically different logics (e.g., technical versus political logics, see for instance Newlove et al. 2003, 132–4). Furthermore, content slippage may occur in situations where the implementation structure is a loosely connected network or where several organizations have to coordinate their actions during the implementation phase with little in the way of support structures. This brings us to the next three indictors that load onto component one; working to coordinate efforts, needing a contact person(s), and making voluntary concessions. It seems with these indicators in mind, that the way we should interpret the content slippage and the lower loading of stalling tactics in implementation is more toward the side of unintentional subversions of decision execution. The actors are sincerely working to coordinate crisis management efforts yet there is a need to establish contact persons to relieve the parties’ failure to communicate. Both these variables indicate that there are multiple organizations whose work needs to be coordinated and that the parties involved must actively work to make this happen. The need for contact persons, together with the making of voluntary concessions, demonstrate a lack of connecting points, of joining structures, if you will, that would facilitate communication between the parts of organizations or between separate organizations. Voluntary concessions indicate that whatever it is that the organizations agree to, or do for each other, is not mandated by rules or obligation, rather the actions take place on a voluntary basis. This results in parties interacting outside of formal arrangement or in ad hoc responses to a lack of regulated interaction.

Cooperative Strategies across Crises

79

Table 5.4  Case example of bureaucratic politics Case 24 The Federal Raid on Ruby Ridge (Puterbauch 2002) “3.5.3. Interaction with Affected Stakeholders. Law enforcement interacted with neighbors and family of the Weavers in an effort to obtain information or to persuade Weaver to surrender. However, the only consideration for stakeholders outside of the law enforcement community was shown by the USMS. Citizens and sympathizers of the Weavers were allowed to demonstrate away from the site where enforcement agents had set up their headquarters. Their comments and feelings expressed were ignored. The public outpouring of sympathy may have enhanced the feeling of danger and pressure that the law enforcement agents perceived, and may also have generated or increased feelings of anger by agents against Weaver, who was to blame for the situation. Public pressure and the loss of public confidence in law enforcement agents increased as a result of the events, and led to investigations of law enforcement conduct by the Department of Justice, and by the U.S. Senate. Issues of Policy Coordination Need for Coordination. The events at Ruby Ridge involved the BATF, the USAO, the USMS and the FBI, as well as the state police, state government and the National Guard. Coordination was effected through personal communication in meetings and briefings between the various groups. While decisions were made and consensus was achieved, there were large gaps in the coordination effort. The FBI failed to debrief the marshals who had been at the initial shooting where Marshal Degan was killed. The misunderstanding over the approval of the rules of engagement is another sign of failed coordination. Quite often, agency heads made decisions without consulting individuals who had first-hand knowledge of events. When Stakeholders’ Definition of the Problem Differ. The differences in this case are not so much in the definition of the problem, but rather in the appropriate approach. All participants agreed that Weaver needed to be arrested, and that danger to Weaver’s family and law enforcement agents, as well as further loss of life should be avoided, and reinforcement of Weaver by sympathizers should be prevented. The USAO’s and FBI’s primary concern was for law enforcement agents, the USMS primary concern was for Weaver’s children. Thus the USMS favored a slow approach, with emphasis on negotiation or deception, while the USAO and FBI emphasized a tactical approach and the use of force. The initial FBI operations plan lacked any provision for negotiations. Such provisions were only included at the insistence of FBI headquarters. Each of the groups was content to leave decision-making to the grouping charge and made no vigorous attempts to change the views and approaches of the decision makers. 3.6.3. Conditions Favoring Coordination. All groups present were part of the Department of Justice. Within each organization, chains of command were established and enforced. This setting facilitated interaction amongst the groups and the smooth passing of responsibility from one group to another. The fact that a law enforcement agent – Marshal Degan – had been killed, also added to the common purpose of all organizations. These conditions persisted in part through the trial of Randy Weaver. However, during subsequent investigations, blaming and blame shifting amongst organizations prevailed.” (Puterbauch 2002, 32–3)

80

Organizational Cooperation in Crises

The remaining three indicators that load positively onto the component are covert behavior, the breaking of agreements, and attribution of fault to other parties. In these we find a tendency to take advantage of the lack of communication between the organizations as well as a lack of commitment to joint action. Evidence of covert behavior illustrates the degree to which actor(s) seek information or make back-room deals behind the scenes in order to protect their own in-group. This type of behavior does not necessarily indicate bad intent. It may be that the channels of information-sharing between organizations is so poor that the actors are left with little choice but to use back-channels to gain insight into what the other actors are doing. However, taken together with agreements between actors being broken and fault being attributed to groups or parties other than ones own, covert behavior to protect ones own group can seem rather opportunistic. Having now elaborated on these indictors, we can return to the crisis case studies to get an idea of what this type of cooperative strategy looks like in the empirics. The excerpt included in Table 5.4 is from a case that constitutes a strong example of bureaucratic politics and illustrates what this component looks like in an actual case. As mentioned earlier, a strong example is one where the particular case contains positive observations on all or most of the variables included in the component. Strategy 2 – Concurrence Seeking Rally around the flag Blame-game Groupthink This component is strongly in-group oriented with the parties fighting to disassociate themselves from the negative events that are occurring. I call this component concurrence seeking. The first indicator loading onto this component is rallying around the flag. This indicator represents a psychologically driven phenomenon in which a group of people pull together and come to each other’s aid by showing support in the face of a perceived common outside threat. The perceived outside threat and the constructed in-group create strong cohesive group dynamics that are often very hard to challenge from the outside. Groupthink is a particular problem-framing and decision-making dynamic that is triggered by the above mentioned strong ingroup cohesion in the face of a perceived threat, and is generated when members of a group would rather see strong, quick agreement on a threatening problem rather than risk the group cohesion being challenged. The felt need to keep the group  ����������������������������������������������������������������������������� Other case examples of bureaucratic politics as an overall case strategy for cooperation can be found in Case 44 – NATO and the Kosovo Conflict Feburary 23, 1998 – March 24, 1999 (Stepanovic 2003a) and Case 77 – Breakdown of Pastrana’s Peace Process (Santander and Matos 2003).

Cooperative Strategies across Crises

81

together by quickly agreeing on a shared view prevails over critical assessment of the basis on which the threatening problem is being evaluated and decisions are being made. Both of the variables, rally around the flag and groupthink, are based on the perception of an outside threat that creates strong in-group and out-group distinctions and concurrence seeking within groups. The third indicator in this component, blame-game, echoes the sense of threat and the protection of an in-group. Blame-game shows the attempt to shift blame for things that are perceived as going wrong in a situation, to other people, groups, or organizations. In one sense this type of focusing of the blame is a natural way to maintain group cohesion. It removes blame that could otherwise split the group and places it with an out-group that the in-group members can then project their fear and aggression on, or distance themselves from. This type of easy and safe allocation of blame on an outside group is consistent with groupthink dynamics in which group-consensus is more important than critically assessing information about the situation at hand. In anther sense, however, blame-game present an attractive tool for shifting the focus based on political or more calculated reasons, such as when there is fault within a group that need to be addressed. Blame-game is attractive precisely because they draw on a heightened attention to in-group and out-group divisions already salient in threat situations (such as crises perceived as going from bad to worse). Having reviewed the indicators make up the case cooperation strategy I have termed concurrence seeking, we turn to the crisis case studies to examine what this cooperation strategy looks like empirically. The following excerpt (Table 5.5) is from a case that provides a strong example of concurrence seeking.10

 ��������������������������������������������������������������������������������� A strong example empirical example contains positive observations on all or most of the variables included in the component. 10 ���������������������������������������������������������������������������������� Another strong case example of concurrence seeking as an overall case cooperation strategy can be found in Case 25 The 1981 Air Traffic Controller’s Strike (Hyatt 2003).

82

Organizational Cooperation in Crises

Table 5.5  Case example of concurrence seeking Case 1

The Small Town YMCA (Hoeschele 2003)

“In June of 2000, Smalltown real estate broker Georgette G. was asked to act as the president of the Board of Directors of the local YMCA. When Georgette took over as the Board president, however, the YMCA was in the midst of serious financial troubles and dwindling membership. Just how bad things were, Georgette had no idea. As she delved more deeply into the troubles of the organization, she uncovered problems relating to self-dealing between the organization and its trustees, lack of financial reporting and accountability, failures in expenditure monitoring procedures and even criminal activities. She realized that the YMCA was in crisis. The organization’s continued illegal and unethical activities threatened the YMCA’s basic values as stated in the organization’s constitution (revised 5/11/84). In fact, its very existence is threatened by financial insolvency, creating a sense of urgency, and uncertainty, among staff and Board members.” (Hoeschele 2003, 2) “Were value conflicts present? Throughout this case, Georgette’s values revolved specifically around saving the YMCA and allowing the community to continue to benefit from the programs it offered. She was relatively new to the YMCA Board, so had a more objective view of the organization and what had to be done to end the crisis. Other Board members, especially those that had been on the Board for a long time, felt that they would rather see the YMCA close than allow ‘those women’ to take over. One male YMCA board member stated it this way, ‘I have this terrible fear that, one Thursday night I’ll come into the Y for Men’s Basketball and there’ll be a bunch of women in the gym doing macramé.’ For Georgette, merger would equal success in effectively ending the crisis. For the rest of the Board, and for Dave McInnis, merger actually equaled failure – a personal failure on the part of the Board to keep the organization afloat. For these individuals, winning meant continuing as a single organization, not necessarily coming out of financial crisis. Other problems with the proposed merger revolve around the perception of the YWCA as a feminist organization. The Smalltown YWCA was the county’s largest provider of child care and aide to victims of domestic violence. Members of the local YWCA chapter looked at themselves solely in these terms, and certainly never considered themselves to be militant feminists. They were, however, known to be advocates for women’s reproductive rights, especially in terms of abortion rights. A number of YMCA Board members were staunch Catholics – one, in fact, ran an extremely conservative adoption agency that actively advocated against abortion. Though the Smalltown YWCA chapter didn’t engage in specific pro-choice activities, the proposed merger was seen as a threat to those YMCA Board members who actively espoused the pro-life agenda.

Cooperative Strategies across Crises

83

Table 5.5  continued Culture issues between the two organizations were also never addressed adequately. The YMCA had only allowed women in their work-out facility in the past 20 years or so, and allowed women to serve on the Board more recently than that. There are still members of the YMCA that hearken back to the days when they could walk around the weight room naked, undisturbed by women in their midst. The YMCA Board was also bothered by the fact that they YWCA is allowed to exclude men from full membership in both the organization as a whole and, in particular, at the Board level. At the same time, the YWCA worried that ‘those idiot men’ would come into their organization and mess things up. Conversations at the YWCA revolved around women being bullied into making decisions by the men on their Board. These conversations got back to the YMCA and increased concern on the part of YMCA Board members that the two organizations were too different to work effectively together. How did policymakers cope with value conflict? Georgette was blind to values conflict throughout the crisis. Since she didn’t really understand the situation from the perspective of other Board members, she was never successful in addressing their objections before the vote. For her, it was obvious that the YMCA was in financial crisis and that the only way to end the crisis was through merger. She couldn’t conceive of other members of the Board seeing the situation differently. It was only after the vote that she began to analyze it in terms of values. Dave McInnis, on the other hand, had an excellent handle on the values of each of the Board members that would eventually vote no on the question of merger. He understood who on the YMCA Board would be easily influenced with the ‘anti-feminist’ argument, who would be swayed by the ‘pro-choice’ card and who would be likely to prefer ‘death’ to merger. He played his cards well and successfully influenced the Board to vote against merger. Short-term vs. Long-term Interests: In this case, Georgette felt that merger with the YWCA would address both short-term needs (help in escaping a financial crisis that was spiraling out of control) and long-term interests (establishing a combined organization that would grow in strength and size as it eliminated competition). For her, admitting managerial defeat in the short term was a small price to pay for ending the financial crisis and putting both the YWCA and the YMCA on a strong footing for growth. For other Board members, and for Dave McInnis in particular, the short term benefit of ending fiscal problems came with too large a price tag: allowing the women to ‘win’ and admitting that they had not been able to sustain their organization. Who Resolves Conflict Can Matter: Dave saw clearly that by undermining Georgette personally, by attacking her motives and bringing up her friendship with Laura, he would be able to influence the final vote. He was also intimidating enough to keep other members of the Board of Directors from actively lobbying for the merger. Three other members of the Board of Directors voted in favor of the proposed merger, but none of them had been willing to speak out in favor of the merger publicly for fear of Dave and his slander machine. If these other three Board members had advocated more vocally in favor of the merger, it is possible that the Board may have voted yes.” (Hoeschele 2003, 15–17)

84

Organizational Cooperation in Crises

Table 5.5  continued “Politico-Bureaucratic Cooperation and Conflict When Stakeholders’ Definitions of the Problem Differ: In this case, stakeholders looked at the crisis as a threat to the very existence of the YMCA. The difference was this: those in favor felt that it was only through merger that the YMCA could continue to exist, while those against felt that the merger itself threatened the existence of the organization. For the former, the merger talks were the answer to the crisis, for the latter, the merger talks were the crisis! With Georgette as the only really vocal merger proponent in the single group decision unit (the Board of Directors), she became the target of those who were most opposed to the merger. Pressures to Stick Together or to Blame Others: Dave McInnis used Georgette’s very vocal support of the merger to rally those against it. This is where she was most vulnerable – she was close friends with the YWCA Board president, she was a woman and she wanted to sell the YMCA out. Dave was able to convince most of the rest of the Board that they had to fight Georgette, and that if they just worked together they could easily pull the YMCA out of it’s financial troubles. Dave’s standing in the community as a large business owner and community scion also helped him use pressure to convince Board members to vote against the merger.” (Hoeschele 2003, 20) “For the most part, though, it mattered little what others in the community thought about the proposed merger. Since the majority of the YMCA Board failed to consider the organization to be in crisis at all, they refused to listen to the opinions of others outside the YMCA community that disagreed with their own.” (Hoeschele 2003, 21)

Strategy 3 – Signal Trustworthiness Claim self to be honest Honesty and self-disclosure in interaction Absence of a need for contact person Absence of an attribution of fault to other people Component three centers on organizations facing up to responsibilities and making honest efforts to interact constructively. I call this cooperation strategy signal trustworthiness. One of the two indicators that load positively onto the component, claiming self to be honest, entails the actors explicitly referring to themselves as being honest and forthcoming with information sharing. There is a certain amount of ambiguity that surrounds self-claimed honesty, especially when explicit references are being made to it. It raises questions of whether or not the actor’s own honesty is somehow being called into question and thereby triggers an explicit confirmation or rebuttal (saying that that actor is in fact being honest and forthcoming). It also raises the question whether the explicit reference to the own group being honest and forthcoming is an encouragement to be forthcoming or an implicit accusation

Cooperative Strategies across Crises

85

of dishonesty directed at the parties the actor is interacting with. It may also be that statements where actors express or claim that they themselves are being honest and forthcoming are simply expressions meant to underline that the interaction is working well, placing official importance on the honesty that is in fact already taking place in the interaction. Without placing this behavior in its proper context it is hard to make out whether such explicit references to honesty are made in good or bad faith. The next indicator in the component, honesty and self-disclosure in interaction, does, however, make the more benign interpretation of claims to be honest fit better. This second indicator entails the actors showing that they are honest in the group setting by actually disclosing information they may have wanted to keep to themselves (had they not been honest and forthcoming). Following the dictum that actions speak louder than words, this indicator confirms that actors are in fact making earnest efforts at communicating important information in the interaction and are generally trying to be constructive in the interaction. The next two indicators that load significantly, but negatively, onto the component actually also support the general idea of the component. The negative loading of a need for contact person(s) due to parties’ failure to communicate should be interpreted as the need for contact persons not coming into question/ being related to the underlying concept uniting the component. If the parties state that they are being honest, and they are in fact being forthcoming with information, there will be no thought of appointing a contact person, as there will be no perceived communication failures. Furthermore, the absence of attributions of fault to other people reflect that in a situation where the parties are being honest and forthcoming, they are in fact assuming responsibility and trying to constructively handle the situation by interacting with each other. This leaves little need to allocate fault and little room to designate a scapegoat if all involved know what is going on and why. We can find good examples11 of this cooperation strategy, signaling trustworthiness, in the case studies that the dataset is built on. In Table 5.6, an excerpt from a representative case illustrates the third case cooperation strategy signaling trustworthiness.12

11 ����������������������������������������������������������������������������������� A good example is a case that contains positive observations on all or most of the variables included in the component (except the negatively loaded variables). 12 ����������������������������������������������������������������������������� Another strong case example of signaling trustworthiness as the overall case cooperation strategy can be found in Case 6 Y2K Crisis in Korea (Shin 2003).

86

Organizational Cooperation in Crises

Table 5.6  Case example of signaling trustworthiness Case 5

The 1997 Hanbo Scandal (Yum 2003)

“Hanbo Steel, a number 14th business conglomerate, collapsed under $6bn in debts – first bankruptcy of a leading Korean conglomerate in a decade, as banks dishonored Hanbo Steel’s check on January 13th. Hanbo Group CEO, Jung Tae-soo continued to maintain the large-scale project of constructing steel mill that required enormous amounts of money while he was on trial related to the slush fund scandal of the former Presidents and the Suso Scandal (involving the development of housing sites). There had been every kind of speculation about his miraculous capability to mobilize capital and there continue to be rumors of a close relationship to politics including President’s second son Kim Hyun-chol. The Hanbo Scandal became a major political issue in that President’s second son Hyun-chol has been suspected of being the background to the illegal loan and the next presidential election was schedule on December 18th. And Korean financial market became seized with panic because rumors of the imminent collapse of other big businesses were circulating and the credit ratings of Korean banks and businesses on the international market dropped sharply. But the investigation to Hanbo Scandal dragged even though President Kim Young-sam ordered ‘a thorough investigation without any regard for sanctuary,’ and the prosecution seemed to wrap up the investigation after indicting Hanbo Group CEO, Hanbo finance manager, four lawmakers and three bankers on February 19th. The people were cynically watching the Hanbo investigation because they believed that the government had already set a limit as to how far the investigation would go. Consequently, economic crisis was deepened and the people’s anger accelerated day by day. After all, President and the prosecution couldn’t but to re-investigate into Hanbo Scandal on March 21st, and finally President’s second son Kim Hyun-chol was arrested on May 15th. But President Kim already lost his intellectual-moral-cultural hegemony.” (Yum 2003, 3–4) “Failure in coping properly with Hanbo Scandal resulted in leadership collapse, lost confidence in the government from inside and outside, and accelerated political, economical and social unrest. And finally Korea could not help asking IMF bailout package.” (Yum 2003, 5) “6.2. The Initial Definition; a through investigation except Hyun-chol” ODM 1–1 (Jan. 27); The President orders “a through investigation without any regard for sanctuary.” OCM 1–2 (Jan. 27); A high-ranking official of the Presidential Office emphasizes that there is absolutely no connection between the President, his family or his office. Only the President virtually defined the situation at his will and influenced how others in the ruling circle view the situation. And he probably thought with ease that he could manage to control the Hanbo incident and protect his lovely son by mobilizing the Prosecution and the ruling party.

Cooperative Strategies across Crises

87

Table 5.6  continued “On January 27, the President ordered ‘a through investigation without any regard for sanctuary.’ But same day a high-ranking official of the Presidential Office emphasized that there was absolutely no connection between the President, his family or his office. The combination of these two remarks was interpreted as ‘a through investigation except Kim Hyun-chol.’ And this was ‘a de facto guideline’ for the investigation into the Hanbo collapse and initial framing identified by predominant leader.” (Yum 2003, 32–3) “And not many people, despite of President order, believed that the prosecution can reveal the truth on Hanbo Scandal in relation to President’s son. Instead they regarded a high-ranking official’s remark as a limitation of investigation because the prosecution, filled with Kim’s confidents, was not a dependant actor from the President. Not many people really believe that the results of the investigation under the President’s order will really expose the true financial transactions of Hanbo and who they implicate. The government has promised that it will investigate not only political circles, but also associates of the President, without any regard for sanctuary. However, past experience tells us that such an investigation will only result in a whitewash. Such disillusion comes from the doubt whether the prosecution can truly conduct an impartial investigation. … The recent reshuffle of a key member of the prosecution team, who was replaced by a person very close to the President.” (Yum 2003, 34–5) “President Kim’s self-righteousness did not accept any feedback from other actors. And Initial frame to protect Hyun-chol by mobilizing the Prosecution continued as followings. ODM 2; Choosing a Scapegoat 2–1; CID summons politicians mentioned by Jung Tae-soo (February 6) 2–2; CID arrests Hong In-kil as a scapegoat (February 11) ODM 3; Issuing a “Paper Indulgence” for Hyun-chol 3–1; Hyun-chol files three libel suits (February 17) 3–2; The result of the Hanbo investigation is announced (February 19) 3–3; CID fails to establish the connection (February 21) ODM 4; Trying a Political Settlement 4–1; President apologize to the nation (February 25) 4–2; full-scale cabinet and ruling party reshuffle (March 5) After all of the efforts, President Kim failed to restore the people’s trust. And Prosecutor General and CID Chief were suddenly replaced on March 21 (ODM 5–1). Finally on May 15, Kim Hyun-chol was arrested on suspicion of taking 6.6 billion won in bribes from six businessmen and avoiding gift and income tax on the payments.” (Yum 2003, 36)

88

Organizational Cooperation in Crises

Strategy 4 – Success-based Helping Behavior Facilitating attainment of in-group’s goals Facilitating attainment of others’ goals Credit seeking behavior Absence of breaking agreements Component four reflects helping behavior in a crisis that is perceived as being, at least in part, successfully managed. This fourth and last component I call successbased helping behavior. The strongest loading indicator in the component is facilitating the attainment of the in-groups’ goals, meaning that the persons involved in managing the crisis are making an effort to help other members of the group to reach their goals. The second indicator, facilitating the attainment of others’ goals, see the same people making efforts to help other outside groups reach their goals. In other words, these variables show decision makers being constructive in helping both those of their own group and outside parties reach their goals in the management of the crisis. With these two variables in mind, this component can be viewed as positively cooperative. That being said, the component does come with a degree of selfishness as decision makers also seek to take credit for the things in the crisis that are viewed as going well or being well-managed. A crisis may see few or many actors seeking credit for or claiming that they played a significant role in bringing about the successful management of the crisis, indicating both high and low levels of competition for the glory of a job well done. The final indicator loading significantly, but negatively, onto this component is breaking agreements. The absence of any agreements broken in the case again highlights that this is a primarily cooperative component where everyone wants to be seen as contributing to the good and successful management of the crisis. Facilitating the attainment of goals indicates that this help in not necessarily done through formal agreement by parties and, as a consequence, breaking or not breaking agreements does not come into play in the concept that underlines this component. For this final component, success-based helping behavior, we can also find good illustrations13 of how this strategy plays out in a real case. The excerpt included in Table 5.7 is from a case that constitutes a strong example of successbased helping behavior.14 13 ���������������������������������������������������������������������������������������� A good illustration is a crisis case that contains positive observations on all or most of the variables included in the component (except the negatively loaded variable). 14 ����������������������������������������������������������������������������� Other case examples that illustrate success-based helping as an overall case cooperation strategy can be found in Case 20 The 1988 Yellowstone Forest Fires (Hawkins 2003); Case 44 NATO and the Kosovo Conflict (Stepanovic 2003); Case 27 Daewoo Group’s Crisis Settlement (Kim 2004); Case 43 The Love Canal Crisis of 1978 (Horrell 2003).

Cooperative Strategies across Crises

89

Table 5.7  Case example success-based helping Case 4

Communal Riots in Malegaon (Gajbhiye 2003)

“The bureau political relation played a very crucial role in Malegaon riots. The very basis of the riots as it later turned out to be was the political causes underlying the events which eventually led to riots. The local politicians, mainly the local MLA and the leader of the opposition at the Malegaon blamed it to the apathy and antagonistic attitude of police towards Muslim minorities which according to them led to the crisis. The revenue administration looked at the problem from a very different angle. It perceived the riot as a result of the underlying rivalry between the two political ideologies which were trying to win over the Muslims of Malegaon on their side. The politicians, least concerned about the outcome of their political actions, had nurtured communal passions of the people for quite some time. The higher level political leaders such as the guardian Minister, the Dy. CM [Deputy County Magistrate] and the CM [County Magistrate] were worried because the riots could generate a criticism of the govt. from the opposition. In spite of all these differing perception, there was a basic understanding and a feeling that the riot should be quelled immediately and should not spread to other areas. In this regard, there was a perfect bureau political cooperation of little conflict in this regard arose when the political leaders belonging to ruling party demanded arrest of the leader of the opposition at Malegaon. But, the govt. thoughtfully avoided this step in the midst of the riots as it would have worsened the situation and would have led the further riots in other parts of the region. As the crisis occurred, there were wide spread repercussions of Malegaon riots./…/ Many people worried as the reports of burning, looting, arson and riots came in from the adjoining districts of Dhule, and Jalgaon. The riots had effect on the nearby contagious areas. The govt. directed the DMs and SPs of all the districts to be on alert and double up their efforts to maintain peace. So as a precaution, a high alert was sounded throughout the state. The DMs of various districts and police officials exchanged information and prepared themselves for any kind of exigency. This prevented the occurrence of any riots in other parts especially in Bhiwandi and Mumbai, the two cities notoriously known for communal disturbances in the past.” (Gajbhiye 2003) “The response of the decision units was quite professional and calculated, well thought of. Although the riots came as a surprise and caught the administration unaware, the authorities geared up and assessed the intensity of the environment causing and fueling the riots.” (Gajbhiye 2003) “It is worthwhile to note here that the state govt. had totally depended upon the district administration (of which now, the Divisional Commissioner had became apart) to handle the situation on their own and there was no directions from above. This independence in handling the situation encouraged the authorities to decide the matter in unison. There was a kind of perfect coalition of police, revenue administration and other agencies such as Municipal Council, the military and the media, which united directed their efforts to restore peace. There was a perfect synchronicity of views and ideas among the policy makers here which helped quell the riots early. There was no dissenting voice as the legal situation and natural leadership lied with the DM who took up the major role of facilitator, a mentor and director and a coordinator.” (Gajbhiye 2003)

90

Organizational Cooperation in Crises

Variables that Run across More than One Component In the unrotated CATPCA solution four indicators reported significantly scores on more than one component. 15 These indicators can, in a sense, be seen as most generic to cooperation at the case level of analysis. The indicators that appeared with significant scores multiple times were blame-game, facilitating others’ goals, need for contact person(s), and attributing fault to other people. Blame-game and attributing fault (to other people) are part of the political game that often surrounds crises. These two indicators are about establishing who is at fault for causing the crisis or for aggravating the conditions of the crisis by not doing what they are supposed to do (not fulfilling their duties, not working with others etc.). These indicators are also concerned with the desire to avoid being pinned with blame and responsibility for the occurrence of the crisis or parts of its management (regardless of whether the criticism is justified or not). This is the essence of ‘passing the buck’; i.e., trying to pass on the blame or the cost of assuming responsibility to another actor. The political art of raising issues of blame and pinning someone else with the responsibility can also be used opportunistically in what has been termed the “politicization” of crises and policy failures (e.g., Brandstrom and Kuipers 2003). This active use of the attribution of fault and the initiation of a blame-game (Hood 2002) is closely linked to agenda setting and the windows-of-opportunity (Kingdon 1995) that crises can help generate.16 In these cases the politicization of an issue or a problem involves amplifying the aspect of threat in a situation and adding a sense of urgency to a problem that has hitherto been off the political agenda. The actors that are in the business of politicizing in order to bring about change that was not possible before then capitalize on the uncertainty of the situation and work to allocate blame or responsibility with an actor that they see as a political opponent or competitor. Attributing fault to other actors is also linked to the third cross-cutting variable, need for contact person(s). While attributing fault to others involved in the management of a crisis is rather uncooperative, it may be that the fault that is perceived is due to a lack of communication or coordination between the actors. 15 ������������������������������������������������������������������������������������ I have chosen a so-called orthogonal (uncorrelated) rotation, the Varimax rotation, because the variable observations become standardized as CATPCA produces the optimal solution for the variables in the model. If I would have used a rotation technique that allowed for correlation between the components (oblique rotation), such as Direct Oblimin, the optimal fit created by CATPCA would have been lost. By saving the transposed variables and their component loadings the CATPCA solution can be rotated in the regular PCA procedure in SPSS. The CATPCA procedure in SPSS (version 14) does not allow for rotation, however, the option of saving the transposed variables makes rotation possible using the more elaborated PCA procedure. 16 ����������������������������������������������������������������������������������� For illustrative case studies of this see Brändström (2000, 2001), Buus and Olsson (2006), Bynander (2003), and Hopkins (2000).

Cooperative Strategies across Crises

91

The need for a contact person due to the actors’ failure to communicate is fertile ground for important issues falling between chairs and organizations inadvertently working at cross purposes. This may later be cause for the attribution of fault or even the rise of blame-games among decision makers. Facilitating the attainment of others’ goals is really the only one of the crosscutting indicators that is truly cooperative in nature. Likewise, it is the most cooperative of the indicators as they have been placed on the spectrum of case level cooperation. The indicator shows actors working actively to coordinate and act in a way that will aid the goals of other decision makers and organizations involved in the management of the crisis, rather than own goals. This is encouraging news for the overall picture of the more widespread characteristics of crisis cooperation. While there are several characteristics that have suggested that assuming responsibility and identifying fault are likely to be handled with competitive political motives in mind, and that failures to communicate are common and often generate a need for contact persons, there is also an indication that unselfish cooperation to achieve goals is also a general feature.

This page has been left blank intentionally

Chapter 6

Linking Behavior and Strategies This chapter examines the links between organizational cooperation behavior and strategies in crises, with particular focus on two empirical questions: what significant empirical relationships are there between the behaviors and the strategies, and how are organizational cooperation behaviors and strategies identified linked to the characteristics of crisis? Both questions can be answered using correlation analysis. The latter question, focusing on what the relationship is between these types of cooperation and the characteristics of crisis, will also be examined using a linear regression model test. The Empirical Relationship between the Cooperation Variables With regard to the relationship between cooperation behavior at the decisionoccasion level of analysis and cooperation strategies at the case level of analysis, we start out by looking at any correlations between individual indicators. Using a correlations analysis with a Kendall’s tau_b coefficient and a significance cutoff of .05 we find a number of correlations between individual decision-occasion indicators and individual case level indicators. The empirical relationships between the indicators are presented below (Table 6.1).

 ���������������������������������������������������������������������������������� The choice of correlation coefficient was Spearman R or Kendall’s tau and both of these models assume that the variables are measured at least an ordinal scale. “Kendall’s tau is equivalent to the Spearman R statistic with regard to the underlying assumptions. It is also comparable in terms of its statistical power. However, Spearman R and Kendall tau are usually not identical in magnitude because their underlying logic, as well as their computational formulas are very different … More importantly, Kendall tau and Spearman R imply different interpretations: While Spearman R can be thought of as the regular Pearson product-moment correlation coefficient as computed from ranks, Kendall tau rather represents a probability. Specifically, it is the difference between the probability that the observed data are in the same order for the two variables versus the probability that the observed data are in different orders for the two variables. Two different variants of tau are computed, usually called tau_a and tau_b. These measures differ only with regard as to how tied ranks are handled. In most cases these values will be fairly similar” (Statsoft 1984–2008).

Organizational Cooperation in Crises

94

Credit seeking Facilitat. ingroups’ goals Facilitat. others’ goals Display honesty Claim honesty Coordinate efforts Voluntary concession Need for contact person Break agreements Stalling tactics Content slippage in implement. Attribute fault to others Covert behavior

Comment

Blame-game

Consult

Groupthink

Approval

Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N

Decide to cooperate

Rally-around the flag

Request/ propose

 

Agreement

 

Yield make

Table 6.1  Correlations between decision-occasion level indicators and case level indicators

.105 .349 80 -.195 .074 85 .178 .116 79 -.005 .936 214 -.106 .135 189 -.043 .532 202 .051 .624 93 -.071 .352 171 .056 .355 276 .146* .018 266 .029 .622 287 .118* .049 278 -.044 .461 287 .155** .008 290 .108 .053 294 .027 .645 281

.053 .637 81 -.110 .312 86 .071 .524 81 .052 .433 216 -.063 .371 192 -.073 .282 204 .151 .146 94 -.110 .148 173 .018 .766 277 .113 .064 268 .002 .995 289 .046 .444 279 -.104 .079 288 .112 .055 292 .031 .575 296 .004 .952 283

-.117 .291 83 -.010 .924 88 -.017 .877 83 -.033 .618 216 .053 .444 193 .033 .628 205 .049 .632 97 .228** .003 176 .013 .834 279 -.025 .684 270 -.060 .305 290 .042 .478 283 .017 .773 292 -.062 .292 294 .038 .496 297 -.022 .707 285

-.082 .451 85 .067 .528 90 -.011 .921 85 .038 .571 216 -.112 .107 195 .033 .621 206 .039 .700 99 .021 .779 178 -.020 .739 280 -.025 .682 272 -.078 .184 290 .013 .831 284 .095 .103 294 -.041 .477 295 .006 .919 298 .100 .086 286

-.082 .455 84 .069 .515 89 .094 .393 83 -.072 .279 216 -.150* .032 191 .001 .984 204 -.147 .153 96 .065 .395 175 .025 .673 278 .116 .057 268 -.043 .466 289 .162** .007 282 .045 .438 292 -.017 .772 292 -.065 .241 297 .025 .670 283

-.025 .824 82 -.088 .413 87 .045 .686 81 -.098 .146 213 -.120 .088 188 -.061 .369 201 .039 .709 93 .197* .010 171 .089 .142 275 .036 .555 266 .070 .239 286 .092 .127 278 .044 .454 287 .107 .070 289 .127* .023 294 .102 .084 280

.056 .629 75 -.235* .037 80 .018 .875 78 -.089 .221 180 -.113 .132 169 -.204** .004 186 .142 .178 91 .193* .019 150 .136* .036 241 .111 .093 231 .131* .038 251 .047 .461 250 .056 .370 261 .114 .071 254 .075 .201 264 -.024 .702 246

Linking Behavior and Strategies

95

Blame-game Credit seeking Facilitat. ingroups’ goals Facilitat. others’ goals Display honesty Claim honesty Coordinate efforts Voluntary concession Need for contact person Break agreements Stalling tactics Content slippage in implement Attribute fault to others Covert behavior

Violence/ coercion

Groupthink

Reduce relations

Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N Corr. Coeff. Sig. N

Threaten

Rally-around the flag

Reject

 

Disapprove

 

Demand

Table 6.1  continued

-.030 .786 81 -.228* .036 86 .025 .825 80 -.029 .662 214 .091 .195 189 -.117 .085 202 .078 .453 94 .070 .361 172 -.011 .854 276 .118 .054 266 .086 .144 287 .008 .897 279 .007 .904 288 .173** .003 290 .009 .874 294 -.057 .332 281

-.051 .649 82 -.074 .494 87 .154 .169 81 .076 .260 213 -.021 .762 189 -.147* .031 202 -.175 .092 94 -.027 .728 171 -.106 .078 275 .155* .012 265 .150* .012 286 .256** .007 278 .116 .050 288 .186** .002 289 .293** .004 294 .011 .853 280

.052 .645 80 -.027 .803 85 .167 .140 79 -.095 .157 213 -.101 .155 188 -.143* .037 201 -.154 .141 92 -.076 .320 170 -.038 .533 275 .006 .924 265 -.032 .594 286 .158** .009 277 -.024 .686 287 .179** .002 289 .054 .333 294 .010 .864 280

-.038 .736 79 -.144 .190 84 .031 .782 79 .063 .345 213 .017 .811 189 -.102 .135 202 -.109 .297 93 -.096 .212 171 -.025 .673 276 .097 .113 266 .142* .017 287 .230** .009 278 -.054 .360 286 .230** .003 290 .149** .008 294 -.064 .278 281

.018 .870 79 -.133 .226 84 -.159 .162 78 -.003 .965 212 -.059 .408 188 -.040 .557 201 -.066 .530 92 .012 .874 170 .074 .222 274 .138* .025 264 .156** .009 285 .132* .028 276 -.049 .405 285 .252** .006 288 -.048 .396 292 .054 .364 279

.000 1.000 80 .034 .758 85 .126 .264 80 .041 .536 214 .083 .240 190 -.030 .663 203 -.095 .357 94 .081 .291 172 -.082 .173 277 .224** .006 267 .108 .069 288 .033 .577 279 -.040 .502 287 .085 .147 291 .085 .125 295 -.001 .982 282

Note: *** = Correlation is significant at the 0.001 level (2-tailed); ** = Correlation is significant at the 0.01 level (2-tailed); * = Correlation is significant at the 0.05 level (2-tailed).

96

Organizational Cooperation in Crises

Cooperation Indicator Correlation Analysis The correlation analysis shows the relationship between two indicators and whether or not this relationship is significant. This type of test does not, however, indicate direction, i.e., which of the indicators is causing the other to be present. Furthermore, the correlation analysis does not tell us whether or not a third indicator, in fact, is causing the connection between the two indicators (such as type of crisis, the number of actors, or the actors’ relative power). With this in mind, all the assertions below concerning the relationship between cooperation at the two levels of analysis are, statistically speaking, supported by the correlations displayed. From an analytical point of view, however, all of these assertions do not seem equally plausible. Based on how the concepts are coded in the dataset and the meaning that the variables carry, I posit a number of hypotheses about the relationship between cooperative behavior at occasions for decision and the overall cooperative strategies that span across the crisis case. While the correlations can be used as a point of departure for thinking about causal relationships, the correlations themselves do not indicate or prove any causation. Yield The two alternative ways of thinking about the correlated items are: A. If the organization yields in the decision-situation, it is likely to display voluntary concessions, break agreements, and display content slippage in the implementation across a crisis case. B. If an organization displays voluntary concessions, breaks agreements, and engages in content slippage in the implementation across a crisis case, it is likely to yield in specific decision-situations. My interpretation is that the following way of thinking about these correlations is the most fruitful: A. If the organization yields in the decision-situation, it is likely to display voluntary concessions, break agreements, and display content slippage in the implementation across a crisis case. This situation is reminiscent of manipulation. The organization is being forced to give in, in the specific decision-situation, and will act quite slippery across the case by granting some voluntary concessions, but also breaking agreements and undermining the content of decisions in their implementation phase. The  �������������������������������������������������������������������������������� i.e., a decision is watered down or corrupted content-wise in the course of its implementation.

Linking Behavior and Strategies

97

relationship between the decision-situation indicator and case indicators could also be the reverse, of course, but it is unclear why the organization would yield in decision-situations if it followed an overall strategy that involved corruption and breaking agreements. Request/propose The two alternative ways of thinking about these correlated items are: A. If the organization makes a request or a proposal in the decisionsituation, it is likely to claim honesty across the crisis case. B. If the organization claims honesty, it is likely to request or propose things to others in individual decision-situations. My argument is that the following interpretation of the correlations is the most fruitful: A. If the organization makes a request or a proposal in the decision-situation, then it will claim honesty across the crisis case. If an organization actually brings proposals and requests to others when facing particular decisions, it may well feel that it has been as forthcoming and open about its needs and planned actions as anyone could be. This would explain why, as an overall strategy, the organization would choose to claim honesty when the management of the crisis is perceived as going poorly. If the proposal or request was made in good faith, the organization may perceive that its suggestions were ignored or rejected and that the resulting mismanagement was something they could not affect. Another possibility, perhaps less credible, is that the organization claiming to be honest has participated in the decision process with the intention of doing little, if anything, but hoping to get away with a perception that it has done what it can. In this situation the organization might well propose or make requests that it knows are likely to be rejected. This way the organization looks like it is doing all it can, is contributing to the decision-making process and the management of the situation, when it in fact is stalling to avoid any real action or commitment. Approve The two views one can take on these correlations are represented by the assertion based on the decision-situation indicator and the two assertions based on the case strategies:

 ���������������������������������������������������������������������������� This may also involve suggesting or appealing to one or more outside actors.

98

Organizational Cooperation in Crises

A. If the organization expresses approval in the decision-situation, it will not facilitate in-group’s goals but rather break agreements across the crisis case. B1. If the organization facilitates the attainment of the in-group’s goals across the crisis, it will not express approval in decision-situations. B2. If the organization breaks agreements across the crisis case, it will express approval in the decision-situation I argue that the following interpretation of the correlation is the most productive: B1. If the organization facilitates attainment of the in-group’s goals across the crisis, it will not express approval in decision-situations. B2. If the organization breaks agreements across the crisis case, it will express approval in the decision-situation. Approval of another may imply a hierarchical relationship between two parties, but it does not do so with necessity. What approval does signal is a certain distance from the other party. For example, if you approve of something that somebody else did you are indeed giving them positive reinforcement. However, you are not conducting the activity of praising together with this other party. Approving of somebody else’s actions therefore implies a distance between the organizations in this specific context. Consequently, if the overall cooperation strategy in a case is to facilitate the ingroup’s goals, then the organization is unlikely to express approval of another as it is primarily focused on its own group and activities. If, however, the overall strategy involves breaking agreements, an action which also signals some distance between the organization and the other actor, then the organization may well express approval of actions by others that fall in line with what the organization wants to have happen. Consult The two ways we may interpret the correlated items are: A. If the organization consults with another in a specific decision-situation, it will claim honesty and attribute faults to others across the crisis case. B. If the organization claims honesty or attributes faults to others across a crisis case, it will consult with others in decision-situations.

 ���������������������������������������������������������������������� i.e., express support or approval of or to another party in the crisis

Linking Behavior and Strategies

99

In my view the following way of thinking about the correlation is the most fertile: A. If the organization consults with another in a specific decision-situation, it will claim honesty and attribute faults to others across the crisis case. Actions such as claiming honesty and attributing fault to others are more likely to appear in situations where the crisis is perceived as being managed poorly or having become exacerbated through somebody’s negative actions or inaction. If the organization consults with others as specific decision problems come up, it is more likely to feel that it has been trying to fix things (find out what the problem looks like, who has what, what the alternatives are) and hence claim honesty when the crisis is perceived as being poorly managed. Furthermore, if the organization consults with others it may also feel that part or all of the fault lies with the parties it consulted and others become an easy target. Consulting with others can enable the spread of blame, preferably to many actors, in cases of management failure. For this reason it is more logical that the behavior precede the strategy, rather than the behavior following the strategy. Some of the case level strategy indicators, such as attributing fault to others, are more likely to appear later on in a crisis but may appear early on if the crisis is perceived as being caused by negligence. In this latter situation a blame-game may commence at the outset of the crisis and become the core threat to actors over the course of the crisis. Claiming to be honest and attributing fault to others are likely strategies to allocate blame. Comment Alternative ways of thinking about the correlated items are: A. If the organization comments on the crisis in the decision-situation, it will not display groupthink or facilitate others’ goals across the crisis case. It will, however, claim honesty, work to coordinate efforts, and display a need for contact persons. B1. If the organization displays groupthink or facilitates the attainment of others’ goals across the crisis it will not comment on the crisis in decisionsituations.

100

Organizational Cooperation in Crises

B2. If the organization either claims honesty, works to coordinate efforts, or if there is a need for contact persons across the crisis case, it will comment on the crisis in decision-situations. The correlations between the indicators at the decision-occasion level and case level are most gainfully interpreted the following way: B1. If the organization displays groupthink or facilitates the attainment of others’ goals across the crisis, it will not comment on the crisis in decisionsituations. B2. If the organization either claims honesty, works to coordinate efforts, or if there is a need for contact persons across the crisis case, it will comment on the crisis in decision-situations. This correlation seems to represent a situation where the organization is taken aback by what has happened. The two strategic cooperation indicators that appear at the case level bear witness to this. Where the organization pursues groupthink or facilitates the attainment of other’s goals we do not see the organization simply commenting on the crisis in decision-situations. These two cooperation strategy alternatives signal a strong reaction, a pulling together and coming to each others’ support in the face of a situation that is likely perceived as quite threatening or dangerous. In this situation the organization and other actors are not standing idly by and watching, as commenting on a crisis implies, but rather find themselves in the thick of it. Consequently, the organization in this situation would be unlikely to just comment on what is happening. In the other cooperation strategy alternative, we see the organization claiming honesty, working to coordinate efforts and needing contact persons. This also suggests a management team taken by surprise and, perhaps as a consequence, displaying disjoint management. In this situation the organization may not have a lot of information on what is really going on, either in terms of the scope of the crisis or what other actors are doing. The actors are in the midst of organizing a response and are having trouble connecting to each other. The organization may choose to only comment on the crisis in a decisionsituation due to the shortage of information about what is going on. In both these situations, the behavior in specific decision-situation scenarios follows the overall management and cooperation strategy across the case.

 ����������������������������������������������������������������������������������� i.e., group(s) focus on maintaining cohesion and loyalty by striving for unanimity and closing itself off from information that differs from what the group wants to do.  ������������������������������������������������������������������������������������� i.e., makes a verbal statement concerning the crisis to outside stakeholders such as the public, the media, etc.

Linking Behavior and Strategies

101

Demand The two ways of thinking about the correlated items are as follows: A. If the organization demands something from others in decision-situations, it will not display groupthink but will, however, display content slippage in the implementation of decisions across the crisis case. B1. If the organization displays groupthink across the crisis case, it will not make demands on others in decision-situations. B2. If the organization displays content slippage in the implementation of decisions across the crisis case, it will make demands on others in decisionsituations. The most fruitful way of thinking about the correlation I argue is: A. If the organization demands something from others in decisionsituations, it will not display groupthink but will display content slippage in the implementation of decisions across the crisis case. If the organization makes demands on others in specific decision-situations there is either an adversarial relationship between those involved in the decision or the group as such is open enough to conflict that expressing demands is accepted in the group. In this latter case, the risk of groupthink across the crisis is very low. Dealing with conflict by concurrence seeking and stifling dissent are hallmarks of groupthink, and making demands in such a decision-situation would violate the silent ethic of the group. Furthermore, if this making of demands in specific decision-situations is an expression of an adversarial relationship between the organization and another actor(s), then it is reasonable that content slippage will take place with regards to decisions that did not go as planned or demanded. With regard to these correlations, it seems more likely that the behavior in the decisionsituations precedes the overall strategy in the case. Disapprove The two ways that this correlation can be expressed are:

 ������������������������������������������������������������������������������������ i.e., issues orders, commands, and/or decrees directed at another party rather than mere requests, suggestions, or appeals.  ������������������������������������������������������������������������������������� i.e., group(s) focused on maintaining cohesion and loyalty by striving for unanimity and closing itself off from information that differs from what the group wants to do.  �������������������������������������������������������������������������������� i.e., a decision is watered down or corrupted content-wise in the course of its implementation.

102

Organizational Cooperation in Crises

A. If the organization disapproves of others in decision-situations, it will not facilitate the attainments of others goals. It will, however, make voluntary concessions, need contact persons, break agreements, display content slippage in implementation, and attribute fault to others across the crisis case. B1. If the organization facilitates the attainment of others’ goals, it will not express disapproval of others in decision-situations. B2. If the organization makes voluntary concessions, needs contact persons, breaks agreements, displays content slippage in implementation, and attributes fault to others across the crisis case, it will express disapproval of others in decision-situations. The following interpretation of the correlation I think is the most productive: B1. If the organization facilitates the attainment of others’ goals, it will not express disapproval of others in decision-situations. B2. If the organization makes voluntary concessions, needs contact persons, breaks agreements, displays content slippage10 in implementation, and attributes fault to others across the crisis case, it will express disapproval of others in decision-situations. The two different cooperation strategies suggested by the case level indicators in this correlation suggest that both strategies generate the disapproving behavior11 in the specific decision-situation. In the first instance, where the organization has an overall strategy of facilitating the attainment of others’ goals, it is unlikely that the organization would, as part of the decision-making process, openly criticize another actor that it may be trying to help. Similarly, a situation where the organization expresses disapproval of an actor or their actions in a decision-situation and then has a highly cooperative strategy across the case facilitating other actors’ goals seems unlikely. Looking at the second strand of case indicators that correlate with expressing disapproval in decision-situations, we find a highly conflictual strategy where there is a fair amount of independence between the parties (indicated by voluntary concessions). This situation makes it more likely that there would be conflicting views on how to act among the parties and makes it possible for the organization to express its disapproval in decision-situations.

10 �������������������������������������������������������������������������������� i.e., a decision is watered down or corrupted content-wise in the course of its implementation. 11 ��������������������������������������������������������������������������������� i.e., the organization expresses disapproval, objection and/or complaints toward another actor.

Linking Behavior and Strategies

103

Reject The alternative ways of thinking about the correlated items are: A. If the organization rejects others in decision-situations, it will not facilitate the attainment of others’ goals across the crisis case. It will, however, break agreements and display content slippage in the implementation of decisions across the crisis case. B1. If the organization facilitates the attainment of others’ goals across the crisis case, it will not reject others in decision-situations. B2. If the organization breaks agreements and displays content slippage of decision implementation across the crisis case, it will reject others in decision-situations. The following I argue is the most fruitful way of thinking about the correlations: A. If the organization rejects others in decision-situations, it will not facilitate the attainment of others’ goals across the crisis case. It will, however, break agreements and display content slippage in the implementation of decisions across the crisis case. Explicitly rejecting another organization or party in a decision-situation is a quite severe form of conflictual behavior. And while breaking agreements and pursuing implementation slippage12 can be seen as less severe, they are indicative of the same fundamental conflictual relationship. As a result the behavior in this situation precedes the case strategy. If the organization is facilitating the attainment of others’ goals as a strategy across the case, this organization is unlikely to openly reject the actor it is trying to help in a decision-situation. Similarly, if the organization openly rejects another actor in a specific decision-situation, it is unlikely to adopt a strategy that involves facilitating the attainment of others’ goals across the crisis. With regard to rejection behavior and facilitating others’ goals, the correlation seems equally probable in both directions. Threaten The two directions in which we may express the correlations are: A. If the organization threatens others in decision-situations, it will need contact persons, break agreements, display content slippage in the implementation of decisions, and attribute faults to others across the crisis case. 12 �������������������������������������������������������������������������������� i.e., a decision is watered down or corrupted content-wise in the course of its implementation.

104

Organizational Cooperation in Crises

B. If the organization needs contact persons, breaks agreements, displays content slippage in the implementation of decisions, and attributes faults to others across the crisis case, then it will threaten others in decisionsituations. The most fertile way of looking at the correlations I argue is the following: A. If the organization threatens others in decision-situations, it will need contact persons, break agreements, display content slippage in the implementation of decisions, and attribute faults to others across the crisis case. As in the case with reject, explicitly threatening13 another actor in a decisionsituation is a more openly conflictual action than either of the strategy indicators in the correlation. This type of situation symbolizes underlying conflict between the organization and another actor. If the organization threatens another actor in a decision-situation, it would follow that the overall cooperation strategy in the case would include a need for contact persons (due to poor communication with other parties), the organization breaking agreements made, undermining the decision implementation,14 and attributing faults to others. Even if the reverse relationship could also be true, it does not necessarily follow that the organization would threaten others even if it broke agreements and tried to undermine the decision implementation process across the case. Reduce relations The following are two ways of thinking about the correlated items: A. If the organization reduces normal relations with others in decisionsituations, then it will make voluntary concessions, need contact persons, break agreements, and display content slippage in the implementation of decisions across the crisis case. B. If the organization makes voluntary concessions, needs contact persons, breaks agreements, and displays content slippage in the implementation of decisions across the crisis case, then it will reduce normal relations in decision-situations. I argue that the following way of thinking about the correlation is the most productive:

13 ��������������������������������������������������������������������������� i.e., verbally expresses threats or coercive warnings toward another actor. 14 �������������������������������������������������������������������������������� i.e., a decision is watered down or corrupted content-wise in the course of its implementation.

Linking Behavior and Strategies

105

A. If the organization reduces normal relations with others in decisionsituations, then it will make voluntary concessions, need contact persons, break agreements, and display content slippage in the implementation of decisions across the crisis case. This decision-situation indicator illustrates the organization making a choice to reduce what would be normal relations15 with another organization. This type of decision is likely to come out of an escalating conflict. In a situation where normal relations between organizations are reduced, it is more likely that there will be a need for contact persons. Consequently, this is a situation where the organization is taking a step back and acting more unilaterally and independently. In this scenario, it is more likely that any cooperative acts would be voluntary, rather than obligatory or contractual. The breaking of agreements (a further show of mistrust and conflict) further illustrates that the organization is not viewing itself as bound by the relations it had with the other organization. The organization is also showing that it has withdrawn its willingness to cooperate with the prior partner organization as it undermines decisions it did not like by compromising the intent of the decision in the implementation phase.16 Violence/coercion The two alternative ways of thinking about the correlated items are: A. If the organization uses violence or coercion in decision-situations, it will make voluntary concessions. B. If the organization makes voluntary concessions across the crisis case, it will use violence or coercion in decision-situations. The most plausible way of thinking about the correlation I argue is: A. If the organization uses violence or coercion in decision-situations, it will make voluntary concessions. Clearly the most extreme of the behavior indicator types, this situation illustrates how organizations get their way (achieve their goals) through coercion and violence.17 As a result, if the organization grants anything at all to other parties it is a voluntary concession. This situation sees a highly dominant and independent organization forcing others to comply with the decision the organization has in 15 ���������������������������������������������������������� i.e., reduce normal, routine relations with another actor. 16 �������������������������������������������������������������������������������� i.e., a decision is watered down or corrupted content-wise in the course of its implementation. 17 ������������������������������������������������������������������������������������ i.e., use government-sponsored oppression, and/or violence against civilians, their rights, or property.

106

Organizational Cooperation in Crises

mind. The organization may grant concessions during the crisis to those that have been forced if the organization so chooses. This situation of strong and centralized power may be one of martial law in crisis situations, situations of presidential decision-making in matters of national security, or a situation of state to state confrontations in international crises. Correlations between Cooperation Behavior and Strategies We now turn to look at the correlational relationships between the five identified types of cooperation behavior and the four overarching cooperation strategies. The component scores18 generated by the CATPCA procedure,19 used to identify the behaviors (see Chapter 4) and strategies (see Chapter 5), are used as continuous indicator measures of each behavior and strategy. The indicators included are continuous and the sample is quite large20 and a Pearson’s correlation coefficient is therefore used to analyze the relationship between cooperation behaviors and strategies.

18 ������������������������������������������������������������������������� The object scores show the relationships between the objects (individual observations of a variable) and the components. 19 �������������������������������������������������������������������������� CATPCA is a quantitative technique very similar to the more commonly used Principal Component Analysis (PCA). Principal component analysis (either CATPCA for categorical data or PCA for interval data) is often used as an exploratory technique aimed at reducing data and finding patterns among variables assumed to measure aspects of a common underlying concept (Linting et al. 2006b; Meulman et al. n.d.). Principal component analysis finds common empirical patterns between variables, sorting these common features into so-called ‘components’ and assigns a value (factor score) to each variable showing how strongly the connection is between each variable and each component. CATPCA can also be used to assign specific values to each observation in the dataset (object scores) based on the component solution. 20 ������������������������������������������������������������������������������������� The Pearson’s coefficient is sensitive to a number of characteristics of the data. A straight linear correlation required that variables are at least interval scale measurements. One assumption of the model is that the residuals are normally distributed, which through scatterplots and a normality test the variables in this test have been confirmed to be, and that the residual values are the same for all independent variables. This assumption is less problematic if the sample is larger than 50. Most researchers then argue that more serious biases are unlikely and that if the sample is larger than 100 there is no concern for biases even if the residual assumptions are not met (Statsoft 1984–2008).

Linking Behavior and Strategies

107

Table 6.2  Descriptive statistics for the correlation analysis of types of cooperation behavior and cooperation strategies

Fight Agree Talk Negotiate Manipulate Bureaucratic politics Concurrence seeking Signal trustworthiness Success-based helping

Mean

Std. Deviation

N

-0.1169 0.0111 0.0790 0.0375 0.0347 0.0152 0.0239 0.0209 -0.0067

1.48297 1.15564 1.56471 1.33455 1.66485 1.09073 0.93812 1.00382 1.05070

315 315 315 315 315 321 321 321 321

Cooperation Strategies and Specific Situation Behaviors The correlation results (Table 6.3) suggest a number of relations between the specific organizational behaviors and different cooperation strategies. Conceptually, even if a strategy is assumed to precede and to direct subsequent behavior, it is worth pointing out that the correlations do not assume nor indicate any such causal direction. The connection between strategies and behavior can be formulated into the following expectations: Bureaucratic politics If the organization pursues a strategy of bureaucratic politics, the organization is likely to just talk in individual decision-situations. It is unlikely to either fight or agree with others in decision-situations. Concurrence seeking If the organization pursues concurrence seeking as a cooperation strategy across the case, it will agree and negotiate in individual decision-situations, but not fight. Signal trustworthiness If the organization pursues a strategy of signaling trustworthiness, it will fight in individual decision-situations. It will not, however, negotiate or manipulate in these situations.

108

Organizational Cooperation in Crises

Table 6.3  Correlations between types of cooperation behavior and cooperation strategies Case cooperation strategies Decision-situation Bureaucratic Threat Signal Successcooperation behavior politics polarization trustworthiness based helping Pearson Corr. Fight Sig. (2tailed) N Pearson Corr. Agree Sig. (2tailed) N Pearson Corr. Talk Sig. (2tailed) N Pearson Corr. Negotiate Sig. (2tailed) N Pearson Corr. Manipulate Sig. (2tailed) N

-.228**

-.123*

.242**

.102

.009 315

.029 315

.005 315

.071 315

-.158**

.157**

-.031

.110

.005 315

.005 315

.578 315

.052 315

.161**

.027

-.025

-.050

.004 315

.629 315

.661 315

.379 315

.033

.243**

-.180**

.055

.565 315

.005 315

.001 315

.326 315

.043

.087

-.117*

-.030

.452 315

.125 315

.039 315

.600 315

Note: *** = Correlation is significant at the .000 level (2-tailed); ** = Correlation is significant at the .01 level (2-tailed); * = Correlation is significant at the .05 level (2-tailed).

It is interesting to note that success-based helping as an overall cooperation strategy is not correlated with any of the cooperation behaviors identified at the decision-occasion level. There are two principle explanations for this result. One, success-based helping may be connected to a form of altruistic attitude and behavior that is not well represented by the cooperative behavior components. In order to capture the behavior that is associated with success-based helping we may need to include indicators that better represent measures of altruistic attitudes and behavior. Looking at the indicators that make up the component, we find that the two indicators have significant correlations with indicators at the decisionoccasion level, but these correlations are negative. If these correlations had been positive, they could have signaled a sixth behavioral component (not identified in

Linking Behavior and Strategies

109

this study)21 that would display more altruistic cooperation behavior at the decisionoccasion level. The negative correlation, however, indicates that if we observe the case level indicators then we are unlikely to see the correlated behavior. Two, it may be that success-based helping is a type of cooperation strategy that is not connected to any particular behavior in decision-situations. The organization pursuing this strategy may not be involved in the decision-making, but rather operating at the outskirts of the management effort, taking credit for what it is only marginally responsible for. This occurs while simultaneously trying to establish a footing, by actively helping facilitate the attainment of others’ goals, with the organization actually authorized to make decisions. Behavior as Predictors of Cooperation Strategies Because correlation analysis is non-directional it is possible to view behavior in specific decision-situations as predictors of cooperation strategies. In other words, if the organization displays a particular type of cooperation behavior in a decisionmaking setting then it is likely to pursue a particular cooperation strategy across the case: Fight If the organization fights in a decision-situation, it is likely to pursue a strategy of signaling trustworthiness. Furthermore, it is unlikely to choose bureaucratic politics or concurrence seeking as cooperation strategies across the case. Agree If the organization agrees in a decision-situation, it is likely to follow concurrence seeking as the cooperation strategy across the case, and unlikely to engage in bureaucratic politics as the cooperation strategy. Talk If the organization simply talks in a decision-situation, it is likely to engage in bureaucratic politics as an overall cooperation strategy. Negotiate If the organization displays negotiation behavior in a decision-situation, concurrence seeking is likely to be its overall cooperation strategy. Negotiating in a decision-situation also makes it unlikely that the organization will signal trustworthiness as its case wide cooperation strategy. 21 ��������������������������������������������������������������������������� If the CATPCA solutions chosen had been one with six components, the sixth component may have formed by these variables that do not fit in the existing five component solution. To maximize the fit and significance of the CATPCA solution a five component solution was chosen in this study, but it would be possible to explore a six component solution provided a poorer fit to the data is acceptable.

Organizational Cooperation in Crises

110

Manipulate If the organization manipulates in a decision-situation, it is unlikely that signaling trustworthiness will be its overall cooperation strategy.

Organization Managing a Crisis

Cooperation Strategies across Crisis Cases

Cooperation Behavior in Decision Situations

Fight Success-based helping

Negot.

Signal trustworthiness

Talk

Concurrence seeking

Org.

Bureaucratic politics

Agree

Manip.

Figure 6.1  Model of correlations between cooperation behavior at the decision-occasion level of analysis and cooperation strategies at the case level of analysis

Note: The dotted lines represent significant negative correlation and the full lines represent significant positive correlation between a cooperative behavior and a cooperative strategy.

The Relations between Cooperative Strategies and Cooperative Behaviors The interaction between organizations in crises may not be as clear cut as this schematic way of understanding and modeling cooperation implies. It seems likely that behavior in decision-situations in some cases generates a particular strategy or a shift in strategy. I also seems reasonable that some choices of strategies make particular behaviors (in individual decision-situations) much more or much less likely. The discussion of the bi-directional properties of the variable-to-variable correlations bears witness to this possibility. Sometimes it appears more reasonable to assume that case level variables would make a correlated decision-occasion indicator a likely behavior, and sometimes the decision level behavior make a case

Linking Behavior and Strategies

111

level strategy indicator more likely. This is the case, for instance, with consult as a decision-situation behavior indicator and the overall case indicators of claiming honesty and attributing fault to others. In this situation it is more likely that the behavior of consulting with others would precede strategies that involved claiming that you have been forthcoming while others have not. This alternative way of thinking about behavior and strategies culminates in two possible upshots: one temporal and one world-view, or attitude, related. The order in which the particular cooperative behavior occurs within the time span of a case may influence the strategy. Alternatively, if an organization’s cooperation strategy is predetermined in the interaction (as a general world view or attitude toward cooperating with others) or becomes firmly set at the very beginning of the interaction (due to the perceived nature of the event or the actors involved) then the cooperation strategy may shape all subsequent behavior in decision-situations. These possible mechanisms are described and examined by Axelrod (1984, 1988) in his work on the evolution of cooperation, i.e., the ways in which strategies and actions become linked in repeated or n-turn games. In the prisoners’ dilemma game, the actor’s cooperation strategy may be conditional as it is with tit-for-tat; the most successful of Axelrod’s identified cooperation strategies. An actor pursuing this strategy only makes two pre-interaction decisions; to start out cooperatively, if the other party does, and then to replicate whatever behavior the other actor displays (cooperating or defecting). In the real life organizational interaction setting we are examining here, the organization’s first decision (what first action to take with regard to others) may corresponds to its general attitude toward cooperation (more or less altruistically inclined towards others) or its perception of the specific crisis situation facing it (perceived severity of the situation, own coping capacity, and whether or not there is an obvious culprit). Following the logic of the tit-for-tat strategy, the organization’s behavior from that first assessment will then depend on the actions of others. In this interaction situation the interaction may start out with some consultations or with some expression of support and approval of other actors, the interaction may then take a turn for the better (showing more cooperative behavior in individual decision-situations) or the worse (showing more conflictual behavior in decision-situations).

This page has been left blank intentionally

Chapter 7

Cooperating in a Crisis Context The last empirical puzzle explored in this study is how the cooperative behaviors and strategies identified are related to the specific interaction setting of crises. Establishing whether any clear links exist, and what the nature of those links may be, is a step towards mapping the nature of crises and determining when we can expect a particular type of organizational behavior or strategy in a particular crisis. Establishing the relationship between cooperative behaviors and cooperative strategies on the one hand, and the relationship between specific crisis characteristics on the other, may help us think about whether there are particular crises, or characteristics of a crisis, that will determine the course of action and strategy of organizations early on. The discussion concerning the possibility of a temporal or world-view set of determinants for behavior and strategies is linked to this empirical test. If certain crisis characteristics are strongly correlated with particular behaviors or strategies, we come closer to setting up testable hypotheses about what triggers particular cooperation responses and what conditions influence organizations’ choice of cooperation strategies in crises. For the empirical analysis I used the distinguishing key characteristics of crises to see if any significant relations can be identified: the perceived threat to core values, urgency, and uncertainty. The TCM variables included in the model of crisis characteristics are surprise,I degree of threat,II origin of threat,III urgency,IV and uncertaintyV (see Key characteristics). I use a Pearson’s correlations coefficient and a one-tailed significance test (see Table 7.1).

Key characteristics I

Variable: Surprise

Description: How unexpected was the crisis for the decision makers? Values: 0 Low – Event was anticipated; indeed the possibility of it occurring had been discussed 1 Medium – Policymakers knew something like this was possible but it had not been discussed recently 2 High – Event was a complete surprise

Organizational Cooperation in Crises

114

Key characteristics II

Variable: Threat

Description: How powerful a threat do the decision makers perceive in the crisis? Values: 0 Low – Routine threat with little danger of long-lasting negative consequences 1 Medium – Decision makers’ values/policies may be weakened and/or undermined, but not beyond repair. In the case of threat to life or human health, the threat is considerable, but not unusual for the type of threat 2 High – Decision makers’ values are seriously threatened, and/or there is a risk of an unacceptable number of deaths and/or injuries

This and the following three variables are dummy variables for the values that thrtorig, the original variable can take: III

Variable: Thrtorig Description: Was the threat perceived by decision makers to be coming from the outside (external to the organizations/groups that set out to manage the crisis) or coming from the inside (problem generated in part or fully by one or more actors involved in managing the crisis)? Values: 0 There was no clear perception of where threat originated 1 Outside 2 Inside 3 Both outside and inside IV

Variable: Urgency

Description: How urgent do decision makers perceive the crisis to be? Values: 0 Not urgent – Decisions makers can go about their other activities while considering what action to take. The window of opportunity is considered to be open 1 Urgent – Decision makers must act quickly, but have time to consider multiple options. The window of opportunity is perceived to be brief but manageable 2 Highly Urgent – All attention is focused on the decision at hand. A short and narrow window of opportunity is perceived to be open

Cooperating in a Crisis Context

115

Key characteristics V

Variable: Uncrtnty

Description: How much uncertainty do the decision makers perceive in defining the crisis? Values: 0 1 2

Low – Crisis is easily defined with confidence Medium – Crisis can be defined, but without total certainty High – Lack of reference points, first time occurrence, and/or little or no access to relevant information

Variable: Unctyres Description: How much uncertainty do the decision makers perceive regarding how to respond to the crisis? Values: 0 1 2

Low – Decision makers have alternatives with which to approach the crisis and feel confident that the alternatives will work but are uncertain about which alternative to pick Medium – Decision makers have a system for managing the crisis, but it takes time to figure that out and they are unsure of the effectiveness of the strategy High – Decision makers have no way of approaching the crisis; there is a certain sense of chaos and high stress

Examining the Connection between Cooperative Behavior and Strategies, and the Characteristics of Crises Three cooperative behaviors are significantly correlated with crisis characteristics: fight, agree, and negotiate. In addition, all four cooperative strategies have significant correlations with certain crisis characteristics. Fighting, as a cooperative behavior in decision-situations, is positively correlated with it being perceived unclear where the threat in the crisis is coming from and perceived urgency. This means that if the organization facing the crisis cannot determine whether the threat the crisis presents originates within or outside of the group managing the crisis, and they perceive the situation as urgent, then the organization is likely to fight in decision-situations. Fighting behavior is also negatively correlated with threats that are perceived as coming from inside the group set to manage the crisis. In other words, if the threat in the situation is

 ����������������������������������������������������������������������������������������� The reverse may also be true; i.e., that the organization is more likely to perceive the threat in the situation as unclear and perceiving the situation as urgent when they fight in decision-situations. I deem this interpretation as the less likely one in terms of direction.

116

Organizational Cooperation in Crises

viewed as stemming, in part or fully, from the actors themselves, then it is unlikely that they will fight in decision-situations. With regards to agree, the correlations show that if the situation is perceived as containing a powerful threat, or if the threat in the situation is perceived as coming from outside of the group, then the organization is unlikely to agree in decisionsituations. If, on the other hand, the organization is uncertain about how to define the crisis, then it is likely to agree in decision-situations. Negotiating, as a cooperation behavior, is negatively correlated with a perceived outside threat, meaning that if the organization perceives the threat in the situation as generated from outside the group set to manage the situation will not negotiate. This means that if the organization perceives the threat as coming from outside of the group, it is unlikely to negotiate in individual decision-situations. Looking at the cooperation strategies at the case level of analysis, we find that all four strategies are significantly correlated with individual crisis characteristics. Bureaucratic politics is positively correlated with the threat of the situation being perceived as generated from within the group. This means that if the organization perceives the threat in the crisis to be caused or generated by the very group set to manage the situation, then it is likely to pursue a cooperation strategy of bureaucratic politics. Concurrence seeking as a cooperation strategy is positively correlated with both of the variables accounting for uncertainty as a crisis characteristic. In these situations, if the organization perceives uncertainty in how to define the crisis or how to respond to the crisis, then it is likely to pursue concurrence seeking across the crisis case. Furthermore, signaling trustworthiness as a cooperation strategy is negatively correlated with the threat of the situation being viewed as coming from inside the group. This means that if the organization views the threat in the crisis as being partially or fully generated from within the group set to manage the situation, then the organization is unlikely to signal trustworthiness as its overall cooperation strategy. Finally, with regard to success-based helping as an overarching strategy, there is a positive correlation with the threat being perceived as unclear and a negative correlation with threat being perceived as coming from outside the group. Simply stated, this shows that if it is unclear to the organization where the threat in the situation is coming from (who or what is generating it), then the organization is likely to pursue success-based helping as the overarching cooperation strategy. If, however, the organization perceives the threat as coming from outside the group set to manage the situation, then the organization will not have success-based helping as its case cooperation strategy.

 ��������������������������������������������������������������������������������������� The reverse is also true in this situation although theoretically less coherent, i.e., that if the actors fight in decision-situations they are unlikely to perceive the threat in the crisis as coming from within or being caused by themselves.

Table 7.1  Correlations between cooperation behavior and strategies, and crisis characteristics Pearson Correlation Surprise Power of threat Threat origin – unclear Threat origin – outside Threat origin – inside Threat origin – inside and outside Urgency Uncertainty – defining sit. Uncertainty – responding to sit. N

Fight

Agree

Negotiate

Bureaucratic politics

Concurrence seeking

0.044 -0.069 0.112* -0.009 -0.140** 0.055 0.149** -0.094 -0.049 300

0.062 -0.139** 0.073 -0.136** 0.048 0.085 0.042 0.135** 0.127* 300

0.107 0.052 0.01 -0.106* 0.064 0.074 0.004 0.016 0.048 300

0.037 0.134 -0.097 -0.138 0.282* 0.005 -0.161 0.038 0.078 63

0.057 0.152 -0.007 -0.179 0.128 0.124 0.104 0.350** 0.372*** 63

Signal Success-based trustworthiness helping 0.058 -0.091 0.177 0.042 -0.230* 0.026 0.011 -0.144 0.134 63

0.18 -0.153 0.229* -0.228* 0.032 0.126 0.158 0.098 0.066 63

Note: The correlation was conducted with Pearson’s correlation coefficients and 1-tailed significance. *** = Correlation is significant at the .000 level; ** = Correlation is significant at the .01 level; * = Correlation is significant at the .05 level.

118

Organizational Cooperation in Crises

It turns out the crisis characteristics are also correlated, indicating that the relationship between cooperation strategies and behaviors in crises is more complex than the correlations between behaviors and strategies show. The correlations between particular crisis characteristics tell us we can distinguish certain types of crises that take on correlated characteristics. Because the crisis characteristics are scaled as low, medium, high, there is likely variation between crises that score low on both correlated variables and those that score high. This, in turn, means that different types of crises may be connected with particular patterns of cooperative behavior and strategies. For example, it is possible that crises with a high degree of perceived threat also entail a high degree of surplice and have a particularly strong connection to the strategy of Concurrence seeking. Two out of three significant correlations seem particularly interesting. Urgency is positively correlated with both the perceived degree of threat (threat) and surprise. What these relationships mean is that if an organization perceive the situation as urgent they are also likely to perceive the situation as posing a serious threat and to be surprised by the situation. The reverse is, of course, equally possible, i.e., that if the organization perceives the situation as very threatening or if it is surprised by the situation it is then likely to perceive the situation as quite urgent. The last significant correlation is the one between the two measures of uncertainty, uncertainty of how to define the crisis and uncertainty with regard to how to respond to the crisis. This correlation is less surprising and therefore less interesting. Further Testing of the Robustness of the Results In some of these cases the correlations could be supported by overall significance of the model in an ANOVA (see Table 7.2) and by significant regression (OLS) results (see Table 7.3). For two of the decision level cooperation behaviors, Fight and Agree, the regression model as a whole was significant at the .001 and .01 level respectively.

Cooperating in a Crisis Context

119

Table 7.2  Significant ANOVA results for regression models of crisis characteristics and their effect on cooperative behaviors and strategies ANOVA Model 1

Agree

Fight

Regression Residual Total Regression Residual Total

Sum of Squares

df

Mean Square

F

Sig.

24.463 331.015 355.478 59.440 523.687 583.127

8 291 299 8 291 299

3.058 1.138   7.430 1.800  

2.688     4.129    

.007*     .000*    

Note: * Predictors: (Constant), Unctyres, Threat unclear, urgency, Threat inside and outside, Threat inside, Surprise, Threat, Uncrtnty.

Table 7.3  Significant effects of crisis characteristics on specific cooperative behaviors and strategies Fight (Constant) Surprise Threat severity Threat – unclear Threat – outside Threat – inside Urgency Uncertainty – how to def. crisis Uncertainty – how to respond Agree (Constant) Surprise Threat severity Threat – unclear Threat – outside Threat – inside Urgency Uncertainty – how to def. crisis Uncertainty – how to respond

B

Std. Error

Beta

Sig.

0.052 -0.153 -0.482 0.95 -0.48 0.329 0.607 -0.483 0.223

0.284 0.13 0.147 0.347 0.274 0.218 0.152 0.204 0.192

  -0.082 -0.219 0.166 -0.108 0.089 0.266 -0.177 0.087

.855 .241 .001 .007 .081 .132 .000 .019 .245

B

Std. Error

Beta

Sig.

-0.09 -0.051 -0.351 0.412 0.08 0.309 0.252 0.043 0.271

0.226 0.104 0.117 0.276 0.218 0.173 0.121 0.162 0.152

  -0.035 -0.204 0.092 0.023 0.107 0.142 0.02 0.135

.690 .621 .003 .136 .715 .076 .038 .793 .076

120

Organizational Cooperation in Crises

Table 7.3  continued Manipulate (Constant) Surprise Threat severity Threat – unclear Threat – inside Threat – inside and outside Urgency Uncertainty – how to def. crisis Uncertainty – how to respond Signal trustworthiness (Constant) Surprise Threat severity Threat – unclear Threat – outside Threat – inside Urgency Uncertainty – how to def. crisis Uncertainty – how to respond

B

Std. Error

Beta

Sig.

-0.417 0.085 0.348 -0.512 0.094 0.073 -0.13 0.381 -0.287

0.302 0.139 0.156 0.369 0.292 0.232 0.162 0.217 0.204

  0.044 0.155 -0.087 0.021 0.019 -0.056 0.137 -0.109

.169 .541 .026 .167 .747 .752 .423 .080 .160

B

Std. Error

Beta

Sig.

0.446 -0.119 -0.263 0.771 -0.196 -0.843 0.116 -0.711 0.689

0.495 0.2 0.239 0.648 0.339 0.504 0.246 0.298 0.284

  -0.087 -0.152 0.164 -0.091 -0.247 0.068 -0.376 0.385

.372 .555 .275 .239 .566 .100 .639 .020 .019

The individual variable effects from the correlation analysis of Fight were also confirmed in the test of the linear regression model (see Table 7.4). In addition, the regression model confirmed the negative effect of threat with a significant result. The negative and significant effect of threat on Agree was also confirmed in the regression analysis. However, slight changes can be noted in the regression analysis. Threat outside was the excluded variable in the model, which indicates that this was not a well-fitting variable in the model. Furthermore, urgency rather than the two uncertainty variables (uncrtnty and unctyres) had a positive and significant effect on Agree in the regression results. In the regression analysis, For Manipulate as a decision-situation behavior, with only near significant results in the correlation analysis, Threat showed a significant positive effect in the linear regression. Among the cooperative strategies only Signaling trustworthiness generated significant results in the regression analysis. For Signal trustworthiness, uncertainty regarding how to define the crisis (uncrtnty) showed a negative and significant effect. The significant correlation effect of threat being perceived as coming from inside the group managing the crisis (threat inside) on Signaling trustworthiness as an overall cooperation strategy did not repeat in the linear regression model.

Table 7.4  Correlations between the crisis characteristics  

 

Threat severity Corr. Coeff. Sig. N Threat – unclear Corr. Coeff. Sig. N Threat – outside Corr. Coeff. Sig. N Threat – inside Corr. Coeff. Sig. N Threat – inside Corr. Coeff. Sig. and outside N Surprise Corr. Coeff. Sig. N Urgency Corr. Coeff. Sig. N Uncertainty – Corr. Coeff. how to define Sig. N crisis Uncertainty – Corr. Coeff. how to respond Sig. N to crisis

Threat Threat Threat severity – unclear – outside

Threat – inside

1.000 . 66 -0.027 0.822 66 0.06 0.617 66 -0.001 0.991 66 -0.075 0.536 66 -0.077 0.501 66 .322** 0.006 66 0.051 0.675 65 0.219 0.063 66

-.001 .991 66 -.075 .545 66 -.471** .004 66 1.000 . 66 -0.154 0.214 66 0.16 0.175 66 -0.045 0.707 66 0.094 0.443 65 0.013 0.912 66

-.027 .822 66 1.000 . 66 -.298* 0.016 66 -0.075 0.545 66 -0.098 0.431 66 0.131 0.267 66 -0.056 0.641 66 0.137 0.266 65 0.092 0.446 66

.060 .617 66 -.298* .016 66 1.000 . 66 -.471** 0 66 -.611** 0 66 -0.049 0.674 66 0.12 0.319 66 -0.234 0.057 65 0.024 0.843 66

Threat – inside Surprise and outside -.075 .536 66 -.098 .431 66 -.611** .006 66 -.154 .214 66 1.000 . 66 -0.128 0.279 66 -0.125 0.297 66 0.095 0.442 65 -0.053 0.662 66

-.077 .501 66 .131 .267 66 -.049 .674 66 .160 .175 66 -.128 .279 66 1.000 . 66 .344** 0.003 66 -0.115 0.327 65 0.174 0.128 66

Urgency .322** .006 66 -.056 .641 66 .120 .319 66 -.045 .707 66 -.125 .297 66 .344** .003 66 1.000 . 66 0.02 0.869 65 0.147 0.207 66

Uncertainty – how Uncertainty – how to define crisis to respond to crisis .051 .675 65 .137 .266 65 -.234 .057 65 .094 .443 65 .095 .442 65 -.115 .327 65 .020 .869 65 1.000 . 65 .511** 0 65

.219 .063 66 .092 .446 66 .024 .843 66 .013 .912 66 -.053 .662 66 .174 .128 66 .147 .207 66 .511(**) .006 65 1.000 . 66

Note: The correlation was conducted with Kendall’s tau_b correlation coefficients and 2-tailed significance. *** = Correlation is significant at the .000 level; ** = Correlation is significant at the .01 level; * = Correlation is significant at the .05 level.

122

Organizational Cooperation in Crises

The correlation analysis demonstrates that there are a number of connections between, on the one hand, the behaviors that organizations display and the cooperative strategies that they pursue, and variation on specific crisis characteristics on the other hand. The most influential of the crisis characteristics are the perception of the threat in the situation and the type of uncertainty experienced. Each of these characteristics had significant relations to three cooperative behaviors and strategies. It is also worth noting that surprise, which several scholars have considered a key component of what constitutes a crisis (Brecher and James 1988, Holsti 1972), is not significantly related to or connected to cooperative behavior or strategies that organizations employ in crises. This does not, however, imply that surprise is not an important part of crises, merely that it is not key to how organizations do or do not get along in crises. The same goes for the rather narrower criteria of the threat in the situation being perceived as coming both from inside and outside of the group set to manage the crisis. It is not clear why this characteristic of the situation did not generate a significant effect on the interaction between organizations, but the results show that when the threat is perceived this way there is no significant relation to any particular cooperative behavior or cooperative strategy in the case.

Chapter 8

Organizing for Crisis Cooperation: Conclusions and Implications This study has taken as a point of departure the need for organizations to cooperate in crises and the many challenges that this type of cooperation may entail. Interorganizational network arrangements are the most common way for states to arrange their response capacity to crises today. The challenges of achieving joint, collaborative, and accommodative policy action among organizations in crisis are paramount. The need to look more closely at how these organizations actually behave and the strategies they employ in crisis interactions has long been a neglected part of crisis research. This study has presented a first systematic investigation of organizational interactions in crisis situations, drawing on the Transboundary Crisis Management (TCM) dataset and quantifying cooperative and conflictual indicators at both the occasion for decision level of analysis and across crises as a whole. By merging theoretical assertions and empirical findings from the cooperation and conflict research across disciplines, we have introduced two new sets of variables. These variables effectively operationalize cooperation at both levels of analysis. When tested against data on a large number of crisis cases, types of crises, decision-making dilemmas, and types of organizational mixes, the analysis revealed a number of patterns in the organizational crisis interactions at both levels of analysis. This part of the book places the findings of this study in the context of a larger research agenda. It does so in two ways. First, we will re-examine the research questions posed at the outset of the study in light of empirical findings. Then in the final chapter we will outline three independent variable sets that will help identify what influences the cooperation behaviors and strategies we have identified. The research agenda presented in Chapter 9 aims to further build on the results of  ������������������������������������������������������������������������������������� The variables at the decision-situation level were to yield, make/come to agreement, request/propose, decide to cooperate, express approval, consult/discuss, comment on, make demand, express disapproval, reject, threaten, reduce relations, and use structural violence. For a more detailed overview of the variables, see Table 2.1. The variables at the case level of analysis were to engage in voluntary concessions, facilitate others’ goals, facilitate the ingroup’s goals, coordinate, rally around the flag, groupthink, new group syndrome, display honesty, claim honesty, need for contact person, credit-seeking, stall implementation, content slippage, Game of Old Maid, blame-game, attribute fault to others, covert behavior, and break agreement. For a more detailed overview of these variables, see Table 2.2.

124

Organizational Cooperation in Crises

this study, allowing us to outline rigorously examined policy recommendations. The kinds of recommendations this further research should address are: how to structure organizational cooperation before a crisis hits; how to formulate good processes for interaction during the acute crisis; and how to organize cooperation during longer processes of recovery from a major societal crises. The policy implications here are a first attempt to place these findings in a planning context. The implications of this study suggest to practitioners ways to facilitate cooperation in decision-situations and promote cooperative strategies across whole crises. The discussion of the implications also highlights ways to prevent or minimize conflict in discrete situations and across whole crises. The Characteristics of Organizational Cooperation in Crises One aspect of this study was how theoretical work in three prominent research traditions on cooperation could facilitate the operationalization of organizational crisis cooperation. Some of the drawbacks of prior research, particularly in international relations and inter-organizational relations, are that it has tended to focus on the causes of cooperation rather than trying to define cooperation in and of itself. I contend that this tendency to treat cooperation as a uni-dimensional concept has been one of the main reasons prior research has been largely unsuccessful in explaining cooperation. Thus, the author was left with the question of defining the object of this study, even after consulting relevant literature extensively. Cooperation worked as an umbrella term for a range of organizational behaviors extending from cooperative to conflictual in this study. The conceptualization of cooperative behavior was developed using Gerner and Schrodt’s action and statement variables for actor interactions in international disputes. In addition, drawing indicators from all three fields (international relations, public administration, psychology), a conceptualization of cooperation strategies at the case level was developed and placed under the umbrella term “cooperation” at the case level of analysis. This conceptualization demonstrated how prior work by scholars that treats cooperation as uni-dimensional and as separate from conflict (e.g., Axelrod 1984; Axelrod and Dion 1988), does not fully capture the empirical reality of crises. In an effort to shed light on how organizations interact in crises, this study has also looked at both the cooperative and competitive behaviors and strategies organizations display in crises. At the outset of this book, the overarching research  ������������������������������������������������������������������������������������ Cooperative organizational behavior has in this study been defined as behavior that aims toward a common goal or objective. This activity may be joint or it may be initiated and performed largely by one party. Cooperative behavior, furthermore, is more or less facilitative and altruistic. Conflictual organizational behavior is behavior that aims toward divergent goals or objectives. This activity may, like cooperation, be joint or more unilateral. Conflictual organizational behavior can be more or less confrontational and coercive.

Organizing for Crisis Cooperation

125

interest in cooperation in crises was broken down into a number of specific questions. These research questions focused on empirical relations, such as how the identified behaviors were related to larger strategies, and how cooperation at the two levels of analysis were related to different characteristics of crises. Based on these empirical results, we can draw a number of general conclusions of the study. First, the results of this study confirm that cooperation and conflict in organizational interactions are intrinsically linked in crises. In crisis decisionsituations, the data shows organizations displaying behaviors ranging from highly cooperative to highly conflictual. In fact, the two strongest patterns of behavior were the two extremes; agreeing and fighting. The organizations also showed more neutral (talking) and more manipulative behaviors (negotiating or manipulating) in decision-situations. The latter two behaviors contained indicators of both cooperative and conflictive actions and statements. In short, the range of behavior that organizations choose to express in crisis decision-situations spans the spectrum from cooperative to conflictive; organizations engage all of these options. At the case level we also find cooperation intrinsically linked to conflict. The data revealed four distinct types of organizational interaction strategies across crises. These strategies represent much more complex concepts than the decisionsituation behaviors, and they do not readily lend themselves to classifications such as ‘more’ or ‘less’ cooperative. Each strategy contains variables that are cooperative in nature and variables that are conflictual in nature, making the overall case strategies theoretically challenging phenomena. What is clear from our results is that regardless of the strategy that organizations adopt, they are going to use a mix of cooperative and competitive actions to achieve their desired objectives. What also is clear from viewing the four types of strategies is that conflictual strategic actions are most common. Second, this study demonstrates that identifying different types of cooperation among organizations in crises is only part of the empirical story. A majority of the interactions between organizations in crises, both in decision-making situations and over the course of a crisis, are conflictual. That is, organizations interacting in crises tend to disagree, fight, and engage in other competitive behavior to a greater extent than they agree, are honest toward each other, or help each other. Looking at the components extracted at the two levels of analysis, we can conclude that the terms ‘cooperative behavior’ and ‘cooperation strategies’ are rather misleading. The empirical reality that these terms are used to describe is more competitive and conflictual than was expected at the outset of this study. In fact, a majority of the behavior types displayed in decision-making situations can be characterized as competitive or conflictual.

 ��������������������������������������������������������������������������������� We continue to use the terms ‘cooperative/cooperation behavior’ and ‘cooperative/ cooperation strategies’ to refer to the components at the occasion for decision level and case

Organizational Cooperation in Crises

126

Third, the overarching crisis cooperation strategies demonstrates the complexity of organizational interactions in crises. All the strategies contain some mix of cooperative, less cooperative, and competitive behaviors. As such, the strategies are complex phenomena, something which calls into question the usefulness of asking questions such as “when do organizations cooperate” or “what makes them cooperate?” As a result, questions regarding crises should focus more on “what interaction strategies do the organizations employ” or “what makes organizations pursue a more cooperative versus a more conflictual interaction strategy.” In terms of common denominators between the strategies, what is perhaps most salient, and hence worth further examination, is the prominence of both political and psychological variables, as well as the opposing functions of organizational processes, i.e., both as a restraint on decision makers and as a tool for furthering decision makers goals and objectives. Fourth, the relationships between cooperation at the two levels of analysis, decision-situation behavior, and overall case strategies has not yet been fully analyzed. We have only begun to formulate ideas about how these different types of interactions are connected. However, based on the correlation analysis presented in Chapter 7, we can posit some reasonable propositions about these relationships. These propositions express the expected relationship between (a) individual indicators of cooperation at the decision-making level of analysis and individual cooperation indicators at the case level, (b) types of cooperative behaviors and cooperative strategies, and (c) cooperative behaviors and strategies, on the one hand, and crisis characteristics, on the other hand. A.

The Links between Cooperation Indicators at the Decision-making and the Case Levels

If the organization, … yields in a decision-situation, it is likely to display voluntary concessions, break agreements, and display content slippage in implementation across a crisis case. … makes a request or makes a proposal in a decision-situation, then it will claim honesty across the crisis case. … facilitates the attainment of the in-group’s goals across the crisis, it will not express approval in decision-situations.

level, respectively, in this chapter but the range of cooperative, competitive and conflictual aspects of these components should remain at the forefront of our minds when we do this.  ����������������������������������������������������������������������������������� This refers to the intentional use of bureaucracy to obstruct implementation or to break agreements.

Organizing for Crisis Cooperation

127

… breaks agreements across the crisis case, it will express approval in decisionsituations … consults with another organization in a specific decision-situation, it will claim honesty and attribute faults to others across the crisis case. … displays groupthink or facilitates the attainment of others’ goals across the crisis, it will not comment on the crisis in decision-situations. … either claims honesty, works to coordinate efforts, or if there is a need for contact persons across the crisis case, it will comment on the crisis in decisionsituations. … demands something from others in decision-situations, it will not display groupthink but will display content slippage in the implementation of decisions across the crisis case. … facilitates the attainment of others’ goals, it will not express disapproval of others in decision-situations. … makes voluntary concessions, needs contact persons, breaks agreements, displays content slippage in implementation, and attributes fault to others across the crisis case, it will express disapproval of others in decision-situations. … rejects others in decision-situations, it will not facilitate the attainment of others’ goals across the crisis case. It will break agreements and display content slippage in the implementation of decisions across the crisis case. … threatens others in decision-situations, it will need contact persons, break agreements, display content slippage in the implementation of decisions, and attribute faults to others across the crisis case. … reduces normal relations with others in decision-situations, it will make voluntary concessions, need contact persons, break agreements, and display content slippage in the implementation of decisions across the crisis case. … uses violence or coercion in decision-situations, it will make voluntary concessions. B1. Cooperation Strategies and Specific Decision-setting Behaviors The correlation analyses lead us to expect the following relationships between strategies and decision-situation behaviors:

128

Organizational Cooperation in Crises

If the organization, Bureaucratic politics … pursues a strategy of bureaucratic politics, the organization is likely to just talk in individual decision-situations. It is unlikely to either fight or agree with others in decision-situations. Signal trustworthiness … pursues a strategy of signaling trustworthiness, it is likely to fight in individual decision-situations. It will not, however, negotiate or manipulate in these situations. Concurrence seeking … pursues concurrence seeking as a cooperation strategy across the case, it is likely to agree and negotiate in individual decision-situations but not fight. B2. Behaviors as Predictors of Cooperation Strategies Based on the earlier correlation analyses we can also expect that if the organizations display the following cooperation behavior, they will pursue the following organizational cooperation strategy across the case: If the organization, Fight … fights in a decision-situation, it is likely to pursue a strategy of signaling trustworthiness. Furthermore, it is unlikely to choose bureaucratic politics or concurrence seeking as cooperation strategies across the case. Agree … agrees in a decision-situation, it is likely to follow concurrence seeking as the cooperation strategy across the case and unlikely to engage in bureaucratic politics as the cooperation strategy. Talk … simply talks in a decision-situation, it is likely to engage in bureaucratic politics as an overall cooperation strategy. Negotiate … displays negotiation behavior in a decision-situation, concurrence seeking is likely to be its overall cooperation strategy. Negotiating in decision-situation also makes it unlikely that the organization will signal trustworthiness as its case-wide cooperation strategy.

Organizing for Crisis Cooperation

129

Manipulate … manipulates in a decision-situation, it is unlikely that signaling trustworthiness will be its overall cooperation strategy. The model of organizational cooperation in crises that is suggested by the data is presented in Figure 8.1.

Fight

Success-based helping

Signal trustworthiness

Negot.

Concurrence seeking

Talk

Bureaucratic politics

Agree

Manip.

Figure 8.1  Model of relations between organizational cooperation behavior (decision-occasion level) and strategies (case level) in crises

Note: The dotted lines represent significant negative correlation and the full lines represent significant positive correlation between a cooperative behavior and a cooperative strategy.

C.

Examining the Connection between Cooperative Behaviors and Strategies and the Characteristics of Crisis

A number of the cooperative behaviors and strategies turned out to have significant correlations with particular crisis characteristics, relations that lead us to expect the following:

130

Organizational Cooperation in Crises

Fight If the organization facing the crisis cannot determine if the threat originates from within or without the group managing the crisis, and they perceive the situation as urgent, then the organization is likely to fight in decision-situations. If the threat in the situation is viewed as stemming, in part or fully, from the actors themselves, it is unlikely that they will fight in decision-situations. Agree If the situation is perceived as containing a severe threat, or if the threat in the situation is perceived as coming from outside the group, then the organization is unlikely to agree in decision-situations. If the organization is uncertain about how to define the crisis, then it is likely to agree in decision-situations. Manipulate If the organization perceives the threat as coming from outside the group, then it is unlikely to negotiate in individual decision-situations. Bureaucratic politics If the organization perceives the threat in the crisis to be caused or generated by the very group set to manage the situation, then it is likely to pursue a cooperation strategy involving bureaucratic politics. Concurrence seeking If the organization perceives uncertainty in how to define or respond to the crisis, then the organization is likely to pursue concurrence seeking across the crisis case. Signal trustworthiness If the organization views the threat in the crisis as being partially or fully generated from within the group set to manage the situation, then the organization is unlikely to signal trustworthiness as its overall cooperation strategy. Success-based helping If the organization perceives that the threat in the crisis is unclear, it is likely to pursue success-based helping as its overall strategy. If, however, the organizations perceive the threat in the crisis as coming from outside the group, it is unlikely to engage in success-based helping.

Organizing for Crisis Cooperation

Threat severity

131

Fight

Threat unclear

Urgency

Negot.

Success-based helping

Talk

Signal trustworthiness

Threat outside

Concurrence seeking

Org.

Bureaucratic politics

Agree Threat inside

Uncert. def. Uncert. respond

Manip.

Figure 8.2  Relations between crisis characteristics and organizational cooperation behaviors and strategies

Note: The dotted lines represent significant negative correlation and the full lines represent significant positive correlation.

It is worth noting in Figure 8.2 that two of the behaviors displayed in decisionsituations do not have any significant correlations to the crisis characteristics. Talk and Manipulate as cooperative behaviors appear to have no direct relation to the crisis characteristics or variations in crisis as an interaction setting. This primary map of crisis cooperation and the propositions found in this study suggest important challenges that need to be developed in the next phase of this research effort (see Chapter 9). Implications for Practitioners in the Field This section is devoted to the practical implications of the study and specifically addresses practitioners in the field. By practitioners in the field, we mean persons who set up and participate in collaborative forums, task forces, or interagency groups to deal with the acute management of crises or their aftermath. They may be part of heavily operational organizations such as rescue services, military, affected businesses, NGOs, or volunteer organizations. They may be representatives of more strategic or politically oriented organizations like governments at different

132

Organizational Cooperation in Crises

levels, agencies, bureaus, departments, and special interest groups. This study has a number of practical implications for these parties, and suggests ways to facilitate organizational cooperation across crises and eliminate factors that promote conflict in specific decision-situations. In What Kind of Situation Can You Expect Problems? The first thing that a practitioner should be aware of is that a majority of the interactions between organizations in crises are conflictual, whether in decisionsituations or and over the course of a crisis. Organizations interacting in crises tend to disagree, fight, or engage in other competitive behavior to a greater extent than they agree, are honest toward each other, or help each other. To expect and prepare appropriately for conflict, both mentally and organizationally, minimizes the negative effect of this type of behavior and interaction strategy in a crisis. What these appropriate preparations might look like will be discussed a little later in our conclusion, but simply knowing that conflict is the predominant interaction feature across crises, enables practitioners to act in a manner that does not preclude them from promoting cooperation. There are a number of situational characteristics that may tip practitioners off that cooperation in the crisis is likely going to be more challenging and that conflict is likely. There are a number of crisis conditions that makes conflict or cooperation the likely behavior in decision-situations and prominent strategies across crises. The more severe the threat is in the situation, the less likely organizations are to agree when they meet to make decisions. This obviously presents a serious problem. Unfortunately, it is also true that the more urgent the situation seems to the organizations, the more likely they are to fight in decision-situations. If it is unclear where the threat in the situation is coming from, the more likely organizations are to openly fight in decision-situations. However, an unclear threat that persists across a crisis will make organizations more likely to pursue a strategy of success-based helping. Such scenarios can provide opportunities to turn the fighting in a crisis around if particular strategies are employed (see the section Fostering Cooperative Strategies among Organizations Interacting in Crises). If the threat in the situation is perceived by those involved as coming from, or being caused in part by, the very group set to manage the crisis, one can expect bureaucratic politics, but not open conflict in decision-situations. The conflict will be kept under the surface and any action taken by organizations to undermine the parties they perceive as responsible will take place in the implementation phase. Likewise, there will be no efforts made to signal trustworthiness across the crisis. A perceived inside threat makes it less likely that organizations will pursue this strategy over the course of a crisis, which eliminates some possibilities for cooperation explained below. If the threat is seen as coming from outside the group set to manage the crisis, this group is likely to ‘harden’ and be less likely to agree or negotiate in decision-

Organizing for Crisis Cooperation

133

situations. Under these circumstances, however, they are not afraid to openly argue and fight in decision-situations. The more uncertain the organizations are about how to define the crisis, or about how to respond to the crisis, the more likely they are to cooperate. They are more likely to agree in decision-situations and tend, as an overarching strategy, to pursue concurrence seeking across the crisis. In these situations practitioners need to be mindful of excessive consensus building that could potentially undermine critical examination of the situation and the options available. In addition to situational characteristics, behavior in decision-situations can be used by practitioners to predict what strategy an organization will pursue in a crisis. If an organization fights in decision-situations it is likely to use a strategy of signaling trustworthiness, but is unlikely to pursue strategies of bureaucratic politics or concurrence seeking across the crisis. If an organization agrees in decision-situations it is likely to pursue concurrence seeking as a strategy across the crisis, but is not likely to engage in bureaucratic politics. If, however, an organization talks in a decision-situation it is more likely to pursue bureaucratic politics as its overall strategy. If the organization negotiates in decision-situations, concurrence seeking is likely as an overall strategy, but signaling trustworthiness is not. Finally, if an organization manipulates in a decision-situation it is unlikely to use signaling trustworthiness as its overall strategy. Keeping track of how different organizations behave in individual decision-situations can signal to the practitioner, an organization’s overall attitude toward pursuing its goals. Tools for Managing Crisis Cooperation As a practitioner facing interactions with other organizations in decision-situations, how can you use your understanding of potentially problematic situations to promote cooperation and decrease conflict with others? There are three different sets of tools available to you for organizing for cooperation. Drawing on theory and case-based research findings in international relations, public administration, social psychology, and crisis research may suggest a number of relevant measures. While the present disadvantage is that we do not know the specific effect of each of these measures on the different types of behavior and strategies organizations display in crises, we can present suggestions based on the best research in the field up to this point. Clearly there needs to be more research specifying which tools and measures will be most effective for practitioners. An outline of future research is presented in Chapter 9. The first tool is related the structuring of cooperation. Simply put, adapting the design of the interaction forum or the inter-organizational structure to meet the characteristics and needs of the situation. Some examples of this are: hierarchies vs. networks, centralized vs. decentralized command, adjusting the number of cooperating organizations, opting for a formal or an informal cooperative arrangement, creating a one-time temporary (ad hoc) work group or an institutionalized permanent crisis management forum, change the reward

134

Organizational Cooperation in Crises

structure from a zero sum allocation of resources and control to one where several organizations benefit more if they cooperate, and increase the way issues are connected so that organizations have more incentives to stick with the process and agree on issues even when they are not immediately gaining from it. The second tool available is process adaptation. Adapting the processes involved in organizational interactions involves picking appropriate rules to govern how the group makes decisions. Some decision-making rules, such as unanimity, make cooperation in decision-making more challenging. You may also change the order in which organizations contribute to the common resources of a network, thereby influencing the voluntary contributions. You may also put mechanisms in place to stave off some of the negative group dynamics (concurrence seeking, groupthink, new group syndrome) that influence problem framing, option generation, and risk assessment. You may also consider changing the communication style of the group from sender-oriented to one that allows for more feedback, making it easier to tell when cooperation with others is getting off track. The third tool available to you as a practitioner is agency. This entails working with individual initiative and active leadership. For example, assume that the organizations are going to be cooperative and make your first action a cooperative one. Your actions have the power to influence the choice of behavior and strategy that other organizations will make. Furthermore, frame issues in an inclusive way and make them seem like an external threat to the group set to cooperate. These measures induce a shared sense of responsibility and common destiny. Avoid divisive politicization of the crisis by framing the origin of the crisis as the symptom of a systemic or network failure, leaving the responsibility of fixing the problem diffuse (i.e., what Brandstrom and Kuipers (2003, 302) term ‘network failure’). Manage communications, both factual and symbolic, and, by extension, manage perceptions of the crisis. Important components of managing communications and the shaping of perceptions is active sense-making and meaning-making. In the face of compromise and decisions work to identify the range of options and preferences that overlap between organizations (win-sets). Punish un-cooperative behaviors (altruistic punishment) so that defecting organizations are induced to switch strategies while cooperating organizations feel that hurtful behavior is justly corrected. Facilitating Cooperation in Crisis Decision-situations Even though talking often precedes agreements, this study shows that the only truly cooperative behavior in decision-situations is agreeing. This behavior involves expressing support for, or agreement with, the actions of another organization, as well as the actual act of entering into agreements. Talking, that is meeting to discuss issues or talking about the situation in public, as a decision-situation behavior should not be interpreted as good news. This is a neutral, non-committal behavior that organizations are likely to display instead of agreeing. If you take into consideration that the other behaviors organizations may engage in decision-

Organizing for Crisis Cooperation

135

situations is fighting, manipulating, and negotiating, then a neutral display of behavior is not necessarily a positive response. Talking, as opposed to agreeing, may indicate that an organization will resort to one of the three conflictive behaviors in the next decision-situation rather than the cooperative behavior. Consequently, in cases where organizations display anything but agreeing behavior there is reason for those in charge to anticipate the need to minimize the likelihood of conflictive behavior. As a practitioner setting up cooperation forums or heading up an interorganizational task force, how can you work with this information? If the members of the group (organizational representatives) are very committed to getting through the crisis together as a group, then the most important factor shaping whether or not an effective cooperative process will take place will be how the group deals with conflicts that arise. The three ways that you can deal with conflict in this kind of group is to ignore it, paper over it, or to recognize it and accept that there are differing opinions without doing anything more about it. Groups that ignore conflict and simply try to come to a consensus tend to be successful in making decisions, but not always very good decisions. This is because the process is flawed by a lack of critical thinking and discussion. Groups that try to compromise, or paper over conflicts, also tend to get through the decision-making process, but they often need to revisit decisions as parties dissatisfied with a compromised outcome will try to raise the issue again for a chance to gain a new outcome. Research on coalition decision-making tells us that whether or not there are rules in place for making decisions and what the rules in place are, matters greatly in shaping the effectiveness of the decision process and the quality of the decision output. It should be pointed out that coalitions are groups of organizational representatives that do not primarily identify with the group pulled together to make decision, but that depend on others in the group to accomplish the decision objective. Their primary loyalty is with the organization they represent, either in this group or another outside group. This makes them less vested in seeing the joint decision-group succeed and make it through the crisis intact. In order to minimize the possibility that more powerful organizations might take advantage of a crisis situation and bully others in decision-situations, it is helpful to have established decision-rules even before the organizations get together to make decisions. Game-theoretic research tells us that enforcement mechanisms that will allow participants to punish defectors are important factors in getting parties to choose to cooperate and to not ‘cheat’ when interacting. Furthermore, decision-rules that propose a majority rule rather than everyone having to agree in order to make a decision make it more likely that you will get decisions made in the group. With a majority voting rule it is likely that organizations in the group will rally others around their position in order to get a majority vote in favor of their proposal or position. Manipulation and negotiation in these situations are common. Ultimately, however, there will be a relatively ordered process for making decisions. There will be a certain amount of accountability and productive channeling of conflict due to the clarity of the majority vote rule. If the

136

Organizational Cooperation in Crises

decision-rule is that the group has to be unanimous in order to make a decision, then the group is likely to become deadlocked and no decision will get through. Unfortunately, this does not prevent the participating organizations from going off and doing whatever they deem necessary to get through when the group cannot come to a decision. In a deadlocked situation you are likely to end up with fragmented, decentralized actions that are often more symbolic than effective joint management efforts. Fostering Cooperative Strategies among Organizations Interacting in Crises The overarching crisis cooperation strategies point to the complexity of organizational interactions in crises. All the strategies contain some mix of cooperative, less cooperative, and competitive behaviors. Consequently, even when organizations adopt a generally cooperative strategy for interacting with other organizations we should expect that there will be competitive elements to their interactions. It is also the case that when organizations adopt more cooperative strategies in crises they will not do so unconditionally. There seems no reason to believe that organizations will act altruistically. They will adopt the most cooperative strategies when the management of the crisis is going well, or is perceived at least by outside stakeholders as being managed well. In other words, organizations will be most cooperative, when they see an opportunity to join a winning team or a popular cause. How can you work with this? Image management is important. If you can manage perceptions of the management effort and develop good relations with both the public and the media, you are more likely to get other organizations to cooperate with you. If you are perceived as managing a bad situation well, other organizations will be more inclined to help you. Ironically, organizations who perceive themselves to be doing well, that they have most things in place and that the management of the situation is progressing steadily are naturally less likely to ask for help. As a result, the increased willingness by other organizations to cooperate (success-based helping) is wasted on situations where it is least likely to be needed. Looking to the more conflict-heavy strategies, how can you as a practitioner work to reverse these courses or mitigate the negative consequences? In general, research tells us that the most successful way of achieving cooperation with others that you repeatedly interact with is to start out cooperatively and then to reciprocate the actions of others. This would be a standard tit-for-tat strategy. If others cooperate, then you cooperate as well. If they fail to cooperate or pursue competitive actions, you do the same, hoping that they will change their behavior. Furthermore, if you can tie issues together – make package deals, link measures of one organization to measures taken by another, or somehow address several organizations’ concerns in a joint output – you are more likely to see organizations willing to cooperate across the board, not just in instances where they are gaining

Organizing for Crisis Cooperation

137

most. By fostering an open process that facilitates accountability and minimizes organizations’ ability to cheat the more comfortable other organizations will be that they are not going to be cheated. Likewise, this type of process makes it easier to intervene and punish organizations that are not doing their part. This brings up a final measure, which is to punish organizations that fail to cooperate or do their part in a deal (either by criticizing or exposing shortcomings to the media or interest groups, or by cutting them out of the information loop). This serves both as a deterrent from cheating, and as a reinforcement of already cooperating organization’s support of the groups’ goals and norms. Increased accountability and enforcement of norms may be the only way to lessen the impact of bureaucratic politics. We do not really know any good measures for reducing or managing bureaucratic politics. Identifying what factors increase and reduce bureaucratic politics is a vital part of future research. Understanding what characterizes concurrence seeking, you are best served as a practitioner to avoid this strategy. The negative consequences of excessive concurrence seeking involve uncritical analysis and decision-making about key components of the crisis such as problem identification, problem framing, option generation and risk assessment. In order to avoid these negative aspects of concurrence seeking an alternative would be to promote a more nuanced view of the threat or problem presented in the situation. Do not let black and white thinking and stereotypes dominate the view of ‘the enemy’ in crises where there seems to be an adversary. It is important to create a group perception of the threat in the crisis as something capable of changing and moving, and to identify aspects of the threat or adversary as something or someone that the group can identify with and understand. Actual mechanisms for achieving this include assigning someone the role of criticizing and questioning the group’s assumptions, assessments and analyses, and the role of devil’s advocate. In the case of signaling trustworthiness, act ‘as if’. Acting on the signals of trustworthiness that an organization shows across a case may open up possibilities to act cooperatively, even in the face of more hawkish stakeholder groups and other competing organizational facets. A successful example of this strategy can be found in John F. Kennedy’s (JFK’s) response to two very different letters from Khrushchev during the Cuban Missile Crisis. While the first letter appealed to JFK’s administration to not escalate the situation but to cooperate by making a deal, the second communication challenged the administration not to make any further hostile moves and threatened swift military action in response to any further actions. In this situation, the administration answered the first letter, acting as if the second one had never existed. This opened up a channel of action for the Khrushchev administration to broker a deal, despite hawkish opposition at home. In addition, face-saving measures can become important tools when organizations adopt this strategy. The signaling of trustworthiness suggests that there is a gap between the organizations trying to cooperate. This gap may be characterized by mistrust and power balances. If this is the case, providing face-saving measures or mechanisms may help turn a stalemate in the relationship into a cooperative

138

Organizational Cooperation in Crises

situation. As shown in the Cuban Missile Crisis, Khrushchev realized that he had gotten himself and the Soviet Union into quite a jam and a part of him was wondering how he could turn the situation around without having to look bad to his stakeholders (The Fog of War 2003).

Chapter 9

An Agenda for Continued Research The findings in this study pave the way for important advances in our understanding of organizational cooperation and conflict in crises. Having established the five behaviors in decision-situations and four overarching strategies that organizations use when the interact in crises, we can now proceed to establish what makes organizations more or less likely to display a particular behavior or pursue a particular strategy. The implications of the study for the world of practice presented in Chapter 8 highlights the need for more specific information on how to prevent, mitigate and minimize conflict in crises and how to facilitate, support, and maximize cooperation between organizations in these situations. Following the outline of tools for managing crisis cooperation in Chapter 8, this research should work to identify the specific impact factors related to structure, process, and agency. With the two-fold objective of furthering our understanding of cooperation in crises and formulating more precise and practically applicable recommendations on how to structure and manage cooperation in crises, the continuation of this research ought to take place along the following lines: drawing on the three traditions of cooperation research combined in this study we need to formulate research questions addressing the specific effect factors identified by each perspective as influencing cooperation in crises. The effect of these competing sets of variables should then be systematically tested against the organizational behaviors and strategies identified in this study. Our Next Research Questions The following research questions aim to further establish the nature of organizational cooperation in crises and determine what factors facilitate and constrain organizational interactions. These questions aim to pin-point the contribution of specific variables that shape the relationship between organizations in crisis settings: How do crisis situational factors influence organizational cooperation behaviors and strategies? How do structural factors and organizational relations facilitate cooperation behaviors and strategies among organizations in crises?

140

Organizational Cooperation in Crises

How do perceptual and psychological factors influence organizational cooperation behaviors and strategies in crises? What set of factors seem to matter most for facilitating organizational cooperation behaviors and strategies in crises? Do the three proposed sets of factors adequately capture what seems to be influencing organizational cooperation, or are there intervening variables that are as yet unaccounted for? How do Crisis Situational Factors Influence Organizational Cooperation Behaviors and Strategies? This research question aims to establish factors other than the crisis characteristics that have been examined here that shape the interaction setting, crises, in important ways. Examples include power relations between the organizations involved, the types of values threatened in the crisis, the timeline of the crisis, the number of actors engaged in the sector hit, and the availability of resources and leadership. The experience and training of the leaders involved, as well as the nature of the organizations they manage, are other situational variables that may affect perceptions of the situation at hand. These factors may also affect the organizations’ ability to take coordinated actions and communicate effectively with others in crisis situations. Organizational crisis cooperation needs neither be voluntary nor planned. The need to interact, and even cooperate, with poor prior coordination can cause problems for any type of organization. It may also be that the particular nature of the event, be it a natural disaster or an armed attack by an aggressor, generates more or less cooperative behavior or particular strategies such as concurrence seeking. Other types of crises, such as corruption scandals or financial crises, may be particularly sensitive for organizations, making cooperation with others particularly important and difficult. How do Structural Factors and Organizational Relations Facilitate Cooperation Behaviors and Strategies among Organizations in Crises? This second question approaches a set of variables that are closely linked to the unit of analysis: organizations. The focus of the question is on forms of organizational relations and the structure, importance, and role of these relations for organizational interactions in crisis situations. Examples of structural factors include: the level of formality of cooperation arrangements, centralized and decentralized response structures, network relations versus hierarchies, decision-making rules, standard operating procedures, degree of accountability to external stakeholders, and the nature of the channels for communication. Some organizational factors increase control and accountability, but hamper flexibility and decrease interpersonal

An Agenda for Continued Research

141

trust-building. Establishing the effect of factors such as these on organizational cooperation will be a key part in addressing the long-standing dilemma that many leaders of organizations face; the importance of planning for crises while simultaneously retaining a degree of flexibility in order to meet inevitable unforeseen conditions. How do Perceptual and Psychological Factors Influence Organizational Cooperation Behaviors and Strategies in Crises? This question narrows in on a set of factors that pertain to group psychology and the importance of perception in crisis situations. The salience of a situation on the organizational agenda and how an organization chooses to respond depends greatly on the perceptions of key decision makers and street-level bureaucrats. Heuristics, problem framing, stress, group dynamics, trust, and leadership styles are just some of the psychological factors that are likely to have an effect on organizational cooperation. While many of these factors are perceived as making cooperation more difficult, others facilitate cohesion and cooperation to an extent that undermines high quality decision-making. In this study, crises have been defined as situations where decision makers perceive a number of potential stressors. The research question above aims to explore how these perceptual and psychological factors influence different types of cooperation. What Set of Factors Seem to Matter Most for Facilitating Organizational Cooperation Behaviors and Strategies in Crises? The fourth question sets out to distinguish those influencing factors that seem particularly supportive of cooperation. Here we are looking for those variables that facilitate the more cooperative behaviors and strategies we identified in this study. Moreover, we are also looking for those variables that have a negative effect on more conflictual behaviors and strategies. An important aim of this question is to make a practical contribution to the structuring and management of organizational interactions that need to take place under challenging circumstances. The potential benefits of cooperation, as well as and the high individual and societal costs that failures to cooperate in crises have brought about, make it pertinent that future research efforts work toward furthering cooperation by drawing practical implications from empirical findings. Do the Three Proposed Sets of Factors Adequately Capture What Seems to be Influencing Organizational Cooperation or are there Intervening Variables that are as of yet Unaccounted for? The fifth and final question is geared toward developing and refining an integrated model of organizational interaction in crises and that demonstrates outcomes of a given interaction. The initial model presented earlier in this chapter sets out a map

142

Organizational Cooperation in Crises

where the test of independent variable sets will further specify mechanisms and relations in the model. By bringing together prior research in international relations, public administration, and social psychology, this integrated model will propel our understanding of how agency, process, and structure govern organizational relations under dire circumstances. Since little research has focused on organizational cooperation in crises, there is much that we do not know, but the potential for expanding our knowledge in this area is substantial. Testing the Effect of Independent Variables With regard to cooperation in crises, there are two alternative ways of formulating the tests of influencing variables. The first is theoretically valid and deductively derived. The second is theoretically informed but inductively organized. The first alternative involves gathering the variables identified in the fields of international relations, public administration, and psychology that appear to have a modifying effect on cooperation and organize them in theoretically appropriate independent variable models. The primary way of selecting variables for statistical models is based on theoretical arguments about cause and effect with regard to the unit of analysis (organizations, groups) and the object of study (cooperation behaviors and strategies). This is particularly the case if we want to use statistical tests such as multiple regression analysis where the results are only ever as valid as the theoretical arguments underlining them. The second alternative involves using theory to select influencing variables, but sorting the variables and simplifying models using factor analysis. This approach would allow us to empirically test whether some independent variables go together more often than others, i.e., if they cluster into a discernible component or independent variables sets. This quantitative technique was used in this study to identify the cooperation behaviors and strategies across crises. This approach would arguably combine the strength of prior research that has identified factors that influence cooperation and conflict, while mitigating shortcomings in theoretical conceptualizations or applicability of theories to influence a construction of a theory of organizational cooperation in crises. Furthermore, with regard to exploring data where the number of potentially important variables is great and the amount of theoretical work is limited, factor analysis is a particularly helpful quantitative technique to guide the placement of variables into specific models. In future empirical work I propose setting up competing models of the three sets of independent variables (centered around situational variables, structural variables, and perceptual variables) and compare the results against sets of independent variables extracted through factor analysis (empirical clusters of independent variables). Comparing the effects of these models and the effects of individual variables on the types of cooperative behaviors and cooperative

An Agenda for Continued Research

143

strategies that this study has identified could bring us closer to a model specifying conditions that both facilitate and constrain organizational cooperation in crises. Furthermore, I suggest revisiting the case studies that the quantitative data in the TCM dataset rests on. While statistical tests can move us forward by specifying conditions that generate behaviors or strategies, quantitative techniques cannot tell us how this happens, i.e., what the causal mechanisms are. As mentioned earlier, our main taproot for identifying causal mechanisms is theory. But in the wake of stronger theoretical frameworks with regard to crisis cooperation, we can look to specific cases that may help us identify potential mechanisms. Looking at the qualitative case studies can provide the kinds of rich context that allows us to identify causal mechanisms specific to crisis interaction settings. These findings may indicate differences and similarities to those mechanisms originally discussed by scholars of cooperation in international relations, public administration, and psychology that this study has drawn upon. Disparate research efforts in three fields have lent a fruitful beginning to our understanding of what cooperation among organizations in crises looks like. The empirical results of the study presented here paves way for the further systematic exploration of cooperative and conflictive organizational behaviors and strategies in situations when it matters most. The specification of when and in what ways structures of cooperative arrangements matter, how the processes that take place in the interactions shape the outcomes, and the role agency in crisis cooperation is what lies ahead. There is room to be optimistic about cooperation in crises in the future, but more than anything this study has confirmed the nature of the challenges and the role that organizing and planning for cooperation plays in making this type of organizational behavior and strategy come about.

This page has been left blank intentionally

Bibliography 9/11 Commission (National Commission on Terrorist Attacks Upon the United States) (2004), ‘Staff Statement 14’ [Online] available at: , accessed 23 August 2005. Alkan, M. (2006), ‘Things I Learned from Working with Victims of Hurricane Katrina’, In the European Crisis Management Academy, The Future Challenges for Crisis Management in Europe [conference], Stockholm, Sweden, May 2006. Allison, G. (1971), Essence of Decision: Explaining the Cuban Missile Crisis (Boston: Little, Brown). Arana, J. (2004), U.S. Presidential Decision-Making Process During the Gulf War of 1991 (M.A. Syracuse, NY: Syracuse University). Ashman, D. (1999), The Organization of Joint Action by Nongovernmental Organization Networks (Interorganizational Networks) [PhD] (Boston, MA: Boston University). Austin, W.G. and Worchel, S. (eds) (1979), The Social Psychology of Intergroup Relations (Monterey: Brooks/Cole). Axelrod, R. (1984), The Evolution of Cooperation (New York: Basic Books). Axelrod, R. and Dion, D. (1988), ‘The Further Evolution of Cooperation’, Science 242:4884, 1385–90. Axelrod, R. and Keohane, R. (1985), ‘Achieving Cooperation Under Anarchy: Strategies and Institutions’, World Politics 38:1, 226–54. Baker, W. and Oneal, J. (2001), ‘Patriotism or Opinion Leadership?: The Nature and Origins of the “Rally Round the Flag” Effect’, Journal of Conflict Resolution 45:5, 661–87. Barnard, C. (1938), The Functions of the Executive (Cambridge, MA: Harvard University Press). Baumgartner, F. and Jones, B. (1993), Agendas and Instability in American Politics (Chicago: University of Chicago Press). Boin, A. (2004), ‘The Forum: Lessons from Crisis Research’, International Studies Review 6:1, 165–94. Boin, A., ‘t Hart, P., Stern, E. and Sundelius, B. (2005), The Politics of Crisis Management: Public Leadership Under Pressure (Cambridge: Cambridge University Press). Bolton, G., Katok, E. and Ockenfels, A. (2004), ‘Trust Among Internet Traders: A Behavioral Economics Approach’, Analyse & Kritik 26:1, 185–202. Bovens, M. and ‘t Hart, P. (1996), Understanding Policy Fiascoes (New Brunswick: Transaction).

146

Organizational Cooperation in Crises

Brändström, A. (2000), Minkar Eller Ubåtar I Den Svenska Ubåtsjakten (Stockholm: Ateljé/Faktor AB). Brändström, A. (2001), Coping with Credibility Crisis: The Stockholm JAS Fighter Plane Crash (Stockholm: Ateljé/Faktor AB). Brändstrom, A. and Kuipers, S. (2003), ‘From “Normal Incidents” to Political Crises: Understanding the Selective Politicization of Policy Failures’, Government and Opposition 38:3, 279–305. Brecher, M. and James, P. (1988), ‘Patterns of Crisis Management’, Journal of Conflict Resolution 32:3, 426–56. Brecher, M. and Wilkenfeld, J. (1989), Crisis, Conflict and Instability (Oxford: Pergamon Press). Brecher, M. and Wilkenfeld, J. (1997), A Study of Crisis (Ann Arbor: University of Michigan Press). Brecher, M., Wilkenfeld, J. and Moser, S. (1988), Crises in the Twentieth Century: Vol. I. Handbook of International Crises (Oxford: Pergamon Press). Brehmer, B. (1991), ‘Dynamiskt och Fördelat Beslutsfattande’, in n.e. (eds). Brehmer, B. (2000), ‘Dynamic Decision Making in Command and Control’, in C. McCann and R. Pigeau (eds). Buus, S. and Olsson, E.-K. (2006), ‘The SARS Crisis: Was Anybody Responsible?’ Journal of Contingencies and Crisis Management 14:2, 71–81. Bynander, F. (2003), The Rise and Fall of the Submarine Threat: Threat Politics and Submarine Intrusions in Sweden 1980–2002 [PhD] Uppsala, SWE: Uppsala University. Bynander, F. and Stern, E. (eds) (1998), Crisis and Internationalization: Eight Case Studies from a Cognitive-Institutional Perspective (Stockholm: Swedish Agency for Civil Emergency Preparedness). Carruth, R.A. (1985), Compliance Mechanisms in Intergovernmental Organizations: The Case of Steel in the Organization for Economic Cooperation and Development and the European Coal and Steel Community (Policy Coordination, Metal Industry, Sanctions) [PhD] Minneapolis, MN: University of Minnesota. Characky, D.J. (2002), Patterns of Cooperation and Conflict within the United Nations Security Council: The Iraq-Kuwait Crisis [MA] San Diego, CA: Alliant International University. Child, J. and Faulkner, D. (1998), Strategies of Cooperation: Managing Alliances, Networks and Joint Ventures (Oxford: Oxford University Press). Choi, J. (2001), Economic Crisis, Elite Cooperation, and Democratic Stability: Asia in the Late 1990s (Indonesia, Korea, Thailand, Malaysia, The Philippines) [PhD] The University of Texas at Austin. Comfort, L. (2006), ‘Summary Report, Workshop 3: Escalation from Disaster to Catastrophe: Plans, Policies, and Uncertainties in Complex, Metropolitan Regions’, in the European Crisis Management Academy, The Future Challenges for Crisis Management in Europe [Conference] Stockholm, Sweden, May 2006.

Bibliography

147

Como, N. (2003), The 1997 Albanian Pyramid Schemes: A Case of Social and Political Crisis in Leadership [MA] Syracuse, NY: Syracuse University. Copeland, B.R. (1987), Conflict and Cooperation in International Trade: Two Essays on the Theory of Protectionism [PhD] Palo Alto, CA: Stanford University. Cotter, C. (eds) (1975), Political Science Annual, Vol. 6 (Indianapolis: BobbsMerrill). Cyert, R. and March, J. (1963), A Behavioral Theory of the Firm: Attempts to Develop a Theory of the Firm which is based on Empirical Studies of Decisionmaking within the Firm (Englewood Cliffs: Prentice Hall). D’Agostino, A. (2003), Nuclear Accident at Tokai-mura, Japan 30 September 1999 [MA] Syracuse, NY: Syracuse University. Das, S. and Teng, B. (1998), ‘Resource and Risk Management in the Strategic Alliance Making Process’, Journal of Management 24:1, 21–42. Dasgupta, N. (2004), ‘Implicit Ingroup Favoritism, Outgroup Favoritism, and Their Behavioral Manifestations’, Social Justice Research 17:2, 143–69. Deng, P. (1998), Relative Gains Problem and Case Studies of Economic Cooperation in East Asia (International Trade, Taiwan, China, Japan) [PhD] Norfolk, VA: Old Dominion University. Department of Homeland Security (2004a), ‘Department of Homeland Security Tom Ridge Approves National Incident Management System (NIMS)’ [Online] Press release (Published 1 March 2005) , accessed 30 August 2005. Department of Homeland Security (2004b), National Incident Management System (Washington DC: Department of Homeland Security). Deutsch, M., Epstein, Y., Canavan, D. and Gumpert, P. (1967), ‘Strategies of Inducing Cooperation: An Experimental Study’, The Journal of Conflict Resolution 11:3, 345–60. Diekmann, A. (2004), ‘The Power of Reciprocity: Fairness, Reciprocity, and Stakes in Variants of the Dictator Game’, Journal of Conflict Resolution 48:4, 487–505. Drescher, S., Burlingame, G. and Fuhriman, A. (1985), ‘Cohesion: An Odyssey in Empirical Understanding’, Small Group Behavior 16:1, 3–30. Dynes, R. and Aguirre, B. (1979), ‘Organizational Adaptation to Crises: Mechanisms of Coordination and Structural Change’, Disasters 3:1, 71–4. Eckhardt, R. (1986), ‘Punctuated Equilibrium: From the Other Side’, Science 232:4753, 920. Einolf, C. (2006), ‘A Comprehensive Theory of Altruistic Behavior’, in The Southern Sociological Society, Annual Conference of the Southern Sociological Society, New Orleans, LA, 1–4 April 2006. Emergency Preparedness; Guidance on Part 1 of the Civil Contingencies Act 2004 its Associated Regulations and Non-Statutory Arrangements, London: HMSO. Available at: , accessed 28 August 2005.

148

Organizational Cooperation in Crises

Evangelista, M. (1990), ‘Cooperation Theory and Disarmament Negotiations in the 1950s’, World Politics 42:4, 502–28. Feridhanusetyawan, T. (2001), ‘APEC Labour Markets: Structural Change and the Asian Financial Crisis’, Journal of Contemporary Asia 31:4, 491–532. Field, A. (2005), Discovering Statistics Using SPSS (and Sex, Drugs and Rock ‘n’ Roll), 2nd Edition (London; Thousand Oaks; New Delhi: Sage Publications). Försvarsberedningen (2007), Säkerhet I Samverkan: Försvarsberedningens Omvärldsanalys (Stockholm: Regeringskansliet). Gaertner, S. and Dovidio, J. (2005), ‘Understanding and Addressing Contemporary Racism: From Aversive Racism to the Common Ingroup Identity Model’, Journal of Social Issues 61:3, 615–39. Gajbhiye, K. (2003), Communal Riots in Malegaon, Maharashtra, India on October 26, 2001, Public Administration, Syracuse, Syracuse University. Galinsky, A., Ku, G. and Wang, C. (2005), ‘Perspective-taking and Self-other Overlap: Fostering Social Bonds and Facilitating Social Coordination’, Group Processes & Intergroup Relations 8:2, 109–24. George, A. (ed.) (1991), Avoiding War: Problems of Crisis Management (Boulder: Westview Press). Gerner, D. and Schrodt, P. (2003) ‘CAMEO: Conflict and Mediation Event Observation Codebook’, Version 0.7b3 (updated November 2003), available at: , accessed 15 March 2004. Gerner, D., Schrodt, P., Abu-Jabr, R. and Yilmaz, Ö. (2002), ‘Conflict and Mediation Event Observations (CAMEO): A New Event Data Framework for the Analysis of Foreign Policy Interactions’, in The International Studies Association, Annual Meeting of the International Studies Association, New Orleans, March 2002, available at: , accessed 15 March 2004. Gerodimos, R. (2004), ‘The UK BSE Crisis as a Failure of Government’, Public Administration 82:4, 911–29. Glaeser, L., Laibson, D. and Sacerdote, B. (2002), ‘An Economic Approach to Social Capital’, Economic Journal 112:433, F437–58. Goldberg, J., Markoczy, L. and Zahn, G.L. (2005), ‘Symmetry and the Illusion of Control as Bases for Cooperative Behavior’, Rationality and Society 17:2, 243–70. Goldstein, J. and Pevehouse, J. (1997), ‘Reciprocity, Bullying and International Cooperation: A Time Series Analysis of the Bosnia Conflict’, American Political Science Review 91:3, 515–30. Gowa, J. (1995), ‘Democratic States and International Disputes’, International Organization 49:3, 511–22. Gowa, J. and Mansfield, E. (1993), ‘Power Politics and International Trade’, American Political Science Review 87:2, 408–20. Grzelak, J. and Derlega, V. (eds) (1982), Cooperation and Helping Behavior: Theories and Research (Academic Press: New York).

Bibliography

149

Grzymata-Kaztowska, A. (2005), ‘From Ethnic Cooperation to In-group Competition: Undocumented Polish Workers in Brussels’, Journal of Ethnic and Migration Studies 31:4, 675–97. Hage, J., Aiken, M. and Marrett, C. (1971), ‘Organization Structure and Communications’, American Sociological Review 36:5, 860–71. Haggard, S. and Simmons, B. (1987), ‘Theories of International Regimes’, International Organization 41:3, 491–517. Hagström, A.-Z. and Sundelius, B. (2001), Krishantering På Göteborska: En Studie Av Brandkatastrofen Den 29–30 Oktober 1998 (Stockholm: Ateljé/ Faktor AB). Haley, K. and Fessler, D. (2005), ‘Nobody’s Watching? Subtle Cues Affect Generosity in an Anonymous Economic Game’, Evolution and Human Behavior 26:3, 245–56. Halperin, M. (1974), Bureaucratic Politics and Foreign Policy (Washington DC: Brookings Institute). Haney, P.J. (1997), Organizing for Foreign Policy Crises: Presidents, Advisers, and the Management of Decision Making (Ann Arbor: University of Michigan Press). Hansén, D. and Stern, E. (eds), Crisis Management in Estonia (Stockholm, Swedish National Defence College). Harris, S. (2000), ‘Asian Multilateral Institutions and Their Response to the Asian Economic Crisis: The Regional and Global Implications’, The Pacific Review 13:3, 495–516. ‘t Hart, P. (1991), ‘Irving L. Janis’ Victims of Groupthink’, Political Psychology 12:2, 247–78. ‘t Hart, P. (1993), ‘Symbols, Rituals and Power: The Lost Dimensions of Crisis Management’, Journal of Contingencies and Crisis Management 1:2, 36–50. ‘t Hart, P. and Pijnenburg, M. (1989), ‘The Heizel Stadium Tragedy’, in U. Rosenthal, M. Charles and P. ‘t Hart (eds). ‘t Hart, P., Stern, E. and Sundelius, B. (eds) (1997), Beyond Groupthink: Political Group Dynamics and Foreign Policymaking (Ann Arbor: University of Michigan Press). Haspers, A. (1998), ‘The Estonia Ferry Disaster: The Making of a Second Order Disaster’, in E. Stern and F. Bynander (eds). Hatzenbichler, G. (2001), ‘EYEWITNESS I – Civil-Military Cooperation in UN Peace Operations Designed by SHIRBRIG’, International Peacekeeping 8:1, 117–21. Hawkins, R. (2003), The 1988 Yellowstone Forest Fires: A Management Crisis for the United States Government [MA] Syracuse, NY: Syracuse University. Hayashi, N., Ostrom, E., Walker, J. and Yamagishi, T. (1999), ‘Reciprocity, Trust and the Sense of Control’, Rationality and Society 11:1, 27–46. Helleiner, E. (1994), States and the Reemergence of Global Finance: From Bretton Woods to the 1990s (Ithica, NY: Cornell University Press).

150

Organizational Cooperation in Crises

Hermann, C. (1972), International Crises; Insights from Behavioral Research (New York: Free Press). Hermann, M. (1979), ‘Indicators of Stress in Policymakers During Foreign Policy Crises’, Political Psychology 1:1, 27–46. Hermann, M., Dayton, B. and Svedin, L. (2003), A Guide to Doing Comparative Case Studies Maxwell Style (Syracuse: Syracuse University). Hermann, M. and Hermann, C. (1990), ‘Hostage Taking, the Presidency, and Stress’, in W. Reich (ed.). Hetherington, M. and Nelson, M. (2003), ‘Anatomy of a Rally Effect: George W. Bush and the War on Terrorism’, PS: Political Science & Politics 36:1, 37–42. Hinde, R. and Groebel, J. (eds) (1991), Cooperation and Prosocial Behaviour (Cambridge: Cambridge University Press). Hoeschele, L. (2003), ‘The Small Town YMCA’, MPA Capstone Project, Public Administration, Syracuse, Syracuse University. Holsti, O. and George, A. (1975), ‘The Effects of Stress on the Performance of Foreign Policy Makers’, in C.P. Cotter (ed.). Holsti, O. (1972), Crisis, Escalation, War (Montreal: McGill-Queen’s University Press). Homans, G. (1961), Social Behavior (New York: Harcourt, Brace, Jovanovich). Honig, J. (2001), ‘Avoiding War, Inviting Defeat: The Srebrenica Crisis, July 1995’, Journal of Contingencies and Crisis Management 9:4, 189–99. Hood, C. (2002), ‘The Risk Game and the Blame Game’, Government and Opposition 37:1, 15–37. Hopkins, A. (2000), Lessons From Longford: The Esso Gas Plant Explosion (Sydney: CCH Australia Limited). Hopkins, A. (2005), Safety, Culture and Risk: The Organisational Causes of Disasters (Sydney: CCH Australia Limited). Horrell, D. (2003), ‘The Love Canal Crisis of 1978’, MPA Capstone Project, Department of Public Administration, Syracuse, Syracuse University. Huddy, L. (2004), ‘Contrasting Theoretical Approaches to Intergroup Relations’, Political Psychology 25:6, 947–67. Hugemark, B. (1991), ‘Delegering Eller Central Styrning: Ett Tidlöst Problem’, in n.e. (eds). Husted, B. (1990), Contracts and Covenants: A Study of Cooperation in Interorganizational Relationships [PhD] University of California at Berkeley. Huxtable, P. (1997), Uncertainty and Foreign Policy-Making: Conflict and Cooperation in West Africa [PhD] Lawrence, KS: University of Kansas. Hyatt, A. (2003), The 1981 Air Traffic Controller’s Strike: Ronald Reagan v. The PATCO Challenge [MA] Syracuse, NY: Syracuse University. Isett, K. (2001), Collective Action in Interorganizational Networks [PhD] Tucson, AZ: The University of Arizona. Janis, I. (1972), Victims of Groupthink (Boston: Houghton Mifflin). Janis, I. (1982), Groupthink, 2nd Edition (Boston: Houghton Mifflin).

Bibliography

151

Janis, I. (1989), Crucial Decisions: Leadership in Policymaking and Crisis Management (New York; London: Free Press; Collier Macmillan). Jennings, E. (1998), ‘Interorganizational Coordination Administrative Consolidation, and Policy Performance’, Public Administration Review 58:5, 417–28. Jervis, R. (1978), ‘Cooperation Under the Security Dilemma’, World Politics 30:2, 167–214. Jervis, R. (1985), ‘From Balance to Concert: A Study of International Security Cooperation’, World Politics 38:1, 58–79. Johann, S. (1983), Cooperation and Coordination between Correctional Education and Other Programs in the Prison System in New York State [Dissertation, DAI, 44, No. 11A] Harvard University. Johnson, L., Zorn, D., Yung, K. Tam, B., Lamontagne, M., and Johnson, S. (2003), ‘Stakeholders Views of Factors That Impact Successful Interagency Collaboration’, Exceptional Children 69:2, 195–209. Jost, J., Banaji, M. and Nosek, B. (2004), ‘A Decade of System Justification Theory: Accumulated Evidence of Conscious and Unconscious Bolstering of the Status Quo’, Political Psychology 25:6, 881–919. Joyce, J. (2000), ‘The IMF and Global Financial Crises’, Challenge 43:4, 88–107. Kanner, M. (2004), ‘Framing and the Role of the Second Actor: An Application of Prospect Theory to Bargaining’, Political Psychology 25:2, 213–39. Katastrofkommissionen (2005), ‘Sverige och tsunamin: gransking och förslag (SOU 2005: 105)’ (Stockholm, Statens Offentliga Utredningar). Keohane, R. (1984), After Hegemony (Princeton: Princeton University Press). Kettl, D. (2007), System under Stress: Homeland Security and American Politics, 2nd Edition (Washington DC: CQ Press). Kickert, W. and Koppenjan, J. (1997), ‘Public Management and Network Management: An Overview’, in W. Kickert, E.-H. Klijn and J. Koppenjan (eds). Kickert, W., Klijn E.-H. and Koppenjan, J. (eds) (1997), Managing Complex Networks: Strategies for the Public Sector (London: Sage Publications). Kim, Y. (2004), ‘Daewoo Group’s Crisis Settlement’, MPA Capstone Project, Department of Public Administration, Syracuse, Syracuse University. Kingdon, J. (1995), Agendas, Alternatives, and Public Policies, 2nd Edition (New York: HarperCollins College Publishers). Koester, W. (2003), ‘Disaster in Waco: The 1993 Confrontation at the Mt. Carmel Complex of the Branch Davidians, Waco, Texas’, MPA Capstone Project, Department of Public Administration, Syracuse, Syracuse University. Krasner, S.D. (1983), International Regimes (Ithaca: Cornell University Press). Kydd, A. (2000), ‘Trust, Reassurance, and Cooperation’, International Organization 54:2, 325–57. Lamberte, R. (2000), ‘1997 Floods in Latvia’, in D. Hansén and E. Stern (eds). Langewiesche, W. (2005), ‘American Ground: Unbuilding the World Trade Center’, in R. Stillman (ed.).

152

Organizational Cooperation in Crises

Lindblom, C. (1965), The Intelligence of Democracy (New York: Free Press). Linting, M., Groenen, P. and Meulman, J. (2006a), ‘Stability of Nonlinear Principal Component Analysis: An Empirical Study’ [Unpublished manuscript] Leiden, NL. Linting, M., Meulman, J., Groenen, P. and van der Kooij, A. (2006b), ‘Nonlinear Principal Component Analysis: Introduction and Application’ [Unpublished manuscript] Leiden, NL. Lipsky, M. (1980), Street Level Bureaucracy: Dilemmas of the Individual in Public Service (New York: Russel Sage Foundation). Mandell, M. (1999), ‘The Impact of Collaborative Efforts: Changing the Face of Public Policy through Networks and Network Structures. Introduction to a Symposium’, Policy Studies Review 16:1, 4–17. Manisera, M., Dusseldorp, E. and van der Kooij, A.J. (2005), ‘Component Structure of Job Satisfaction Based on Herzberg’s Theory’ [Unpublished manuscript] Brescia, IT. March, S. (2003), ‘Accident at Three Mile Island – 1979’, MPA Capstone Project, Public Administration, Syracuse, Syracuse University. Maresca, J. (2003), ‘Post-conflict Reconstruction: Developing Public-Private Partnerships’, UN Chronicle 40:3, 14–15. Martin, L. (1992), Coercive Cooperation: Explaining Multilateral Economic Sanctions (Princeton: Princeton University Press). Martin, R. (1999), ‘The Joint Terrorism Task Force: A Concept That Works’, FBI Law Enforcement Bulletin 68:3, 23–7. Marwell, G. and Schmitt, D. (1975), Cooperation (New York: Academic Press). McCann C. and Pigeau R. (eds) (2000), The Human in Command: Exploring the Modern Military Experience (New York: Kluwer Academic/Plenum Publishers). McConnell, A. and Drennan, L. (2006), ‘Mission Impossible? Planning and Preparing for Crisis’, Journal of Contingencies and Crisis Management 14:2, 59–70. McKeown, T. (2000), ‘The Cuban Missile Crisis and Politics as Usual’, The Journal of Politics 62:1, 70–87. Meulman, J., van der Kooij, A. and Heiser, W. (n.d.), ‘Chapter 3: Principal Component Analysis with Nonlinear Optimal Scaling Transformations for Ordinal and Nominal Data’, Unpublished manuscript, Leiden, NL. Miller, A. (2001a), ‘The Los Angeles Riots: A Study in Crisis Paralysis’, Journal of Contingencies and Crisis Management 9:4, 189–99. Miller, E. (2001b), ‘The New World Order in Theory and Practice: The Bush Administration’s Worldview in Transition’, Presidential Studies Quarterly 31:1, 56–68. Milner, H. (1992), ‘International Theories of Cooperation Among Nations: Strengths and Weaknesses’, World Politics 44:3, 466–96. Moment of Crisis: Anatomy of a Riot, 1993 [DVD] Oak Forest, IL: MPI Home Video (narrated by Ted Koppel).

Bibliography

153

Morgenthau, H. and Thompson, K. (1985), Politics Among Nations: The Struggle for Power and Peace, 6th Edition (New York: Knopf). Mueller, J. (1970), ‘Presidential Popularity from Truman to Johnson’, The American Political Science Review 64:1, 18–34. Nakayama, M. (1999), Information Technology Impact on Interorganizational Relationships in Marketing Channels: Management of Cooperation, Conflict, and Information Sharing [PhD] University of California at Los Angeles. N.e. (eds), Ledning och Beslutsfattande: Informationsteknologi Till Samhällets Försvar, Stockholm, Försvarsmedia. Newlove, L., Stern, E. and Svedin, L. (2003), Auckland Unplugged: Coping with Critical Infrastructure Failure (Lanham: Lexington Books). Nye, J.S. (1984), ‘U.S.–Soviet Relations and Nuclear-risk Reduction’, Political Science Quarterly 99:3, 401–14. O’Brien, K. (1996), Interorganization Cooperation and the Effectiveness of Service Delivery to Homeless Children and Youth (School Districts, Human Service Agencies) [PhD] State University of New York at Albany. O’Gorman, R., Wilson, D. and Miller, R. (2005), ‘Altruistic Punishing and Helping Differ in Sensitivity to Relatedness, Friendship, and Future Interactions’, Evolution and Human Behavior 26:5, 375–87. O’Toole, L. (1995), ‘Rational Choice and Policy Implementation: Implications for Interorganizational Network Management’, American Review of Public Administration 25:1, 43–58. Oye, K. (1985), ‘Explaining Cooperation Under Anarchy: Hypotheses and Strategies’, World Politics 38:1, 1–24. Page, S. (2003), ‘Entrepreneurial Strategies for Managing Interagency Collaboration’, Journal of Public Administration Research and Theory 13:3, 311–39. Park, D. (2003), The Currency Crisis of South Korea in 1997 [MA] Syracuse, NY: Syracuse University. Parsons, T. (1951), The Social System (Glencoe: Free Press). Pathak, A. (2004), War Tension between Two Nuclear Powers: India and Pakistan [MA] Syracuse, NY: Syracuse University. Patterson, J. (1993), Cooperation for Competition: The Formation and Growth of ‘Strategic’ Networks in the International Motor Vehicle Industry, 1959–1988 (Strategic Networks) [PhD] University of Illinois at Urbana-Champaign. Perrow, C. (1984), Normal Accidents: Living with High-Risk Technologies (New York: Basic Books). Perry, J. (1996), ‘Measuring Public Service Motivation: An Assessment of Construct Reliability and Validity’, Journal of Public Administration Research and Theory 6:1, 5–22. Perry, J. (1997), ‘Antecedents of Public Service Motivation’, Journal of Public Administration Research and Theory 7:2, 181–97. Perry, J. and Wise, L. (1990), ‘The Motivational Basis of Public Service’, Public Administration Review 50:3, 367–73.

154

Organizational Cooperation in Crises

Petee, J. (1986), The Development of Collective Action in Hierarchical and Nonhierarchical Settings (Collaboration, Intergroup Relations, Interorganizational, Cooperation, Coalition Formation, Social Action) [PhD] Cleveland, OH: Case Western Reserve University. Peterson, S. (1992), Deterrence, Cooperation, and Bargaining in International Crises (Crimean War, Berlin Crisis, Fashoda Crisis) [PhD] New York, NY: Columbia University. Pevehouse, J. and Goldstein, J. (1999), ‘Serbian Compliance or Defiance in Kosovo?’ Journal of Conflict Resolution 43:4, 538–46. Pigeau, R. and McCann, C. (2000), ‘Redefining Command and Control’, in C. McCann and R. Pigeau (eds). Piskalyuk, Y. (2004), Construction of the Dam by the Russian Federation towards the Ukranian Island of Tuzla (Sept. 18–Oct. 30, 2003) [MA] Syracuse, NY: Syracuse University. Poulsen, A. and Svendsen, G. (2005), ‘Social Capital and Endogenous Preferences’, Public Choice 123:1, 171–196. Powell, W., Koput, K. and Smith-Doerr, L. (1996), ‘Interorganizational Collaboration and the Locus of Innovation: Networks of Learning in Biotechnology’, Administrative Science Quarterly 41:1, 116–45. Preston, T. and ‘t Hart, P. (1999), ‘Understanding and Evaluating Bureaucratic Politics: The Nexus Between Political Leaders and Advisory Systems’, Political Psychology 20:1, 49–98. Provan, K. and Milward, H. (1991), ‘Institutional-Level Norms and Organizational Involvement in a Service-implementation Network’, Journal of Public Administration Research and Theory 1:4, 391–417. Puterbauch, A. (2002), The Federal Raid on Ruby Ridge [MA] Syracuse, NY: Syracuse University. Putnam, R. (2000), Bowling Alone: The Collapse and Revival of American Community (New York: Simon & Schuster). Putnam, R., Leonardi, R. and Nanetti, R. (1993), Making Democracy Work: Civic Traditions in Modern Italy (Princeton: Princeton University Press). Raub, W. (2004), ‘Hostage Posting as a Mechanism of Trust: Binding Compensation and Signaling’, Rationality and Society 16:3, 319–65. Reagan, R. (2003), The Iran Hostage Crisis: Four Decisions from 4 November 1979 to 25 April 1980 [MA] Syracuse, NY: Syracuse University. Reich, W. (ed.) (1990), Origins of Terrorism (Cambridge, MA: Cambridge University Press). Reicher, S. (2004), ‘The Context of Social Identity: Domination, Resistance, and Change’, Political Psychology 25:6, 921–45. Rogowski, R. (1989), Commerce and Coalitions: How Trade Affects Domestic Political Alignments (Princeton: Princeton University Press). Rosati, J. (1981), ‘Developing a Systematic Decision-Making Framework: Bureaucrataic Politics in Perspective’, World Politics 33:2, 234–52.

Bibliography

155

Rosenthal, U. and ‘t Hart, P. (eds) (1998), Flood Response and Crisis Management in Western Europe: A Comparative Analysis (Berlin: Springer). Rosenthal, U., Charles, M and ‘t Hart, P. (eds) (1989), Coping with Crises: The Management of Disasters, Riots, and Terrorism (Springfield: C.C. Thomas). Rosenthal, U., ‘t Hart, P. and Kouzmin, A. (1991), ‘The Bureau-Politics of Crisis Management’, Public Administration 69:2, 211–33. Rosenthal, U., ‘t Hart, P., Van Duin, M., Boin, A., Kron, M., Otten, M. and Overdijk, W. (1994), Complexity in Urban Crisis Management: Amsterdam’s Response to the Bijlmer Air Disaster (London: James and James). Roy, M. (1999), Change in World Politics and International Trade Openness (Cooperation, Balance of Power, Security) [PhD] Washington DC: Johns Hopkins University. Rubin, M. and Hewstone, M. (2004), ‘Social Identity, System Justification, and Social Dominance: Commentary on Reicher, Jost et al., and Sidanius et al.’, Political Psychology 25:6, 823–44. Ruggie, J. (1982), ‘International Regimes, Transactions and Change: Embedded Liberalism in the Post-War Economic Order’, International Organization 36:2, 379–415. Ryan, V., Agnitsch, K., Zhao, L. and Mullick, R. (2005), ‘Making Sense of Voluntary Participation: A Theoretical Synthesis’, Rural Sociology 70:3, 287– 313. Sabatier, P. (2007), ‘Fostering the Development of Policy Theory’, in P. Sabatier (ed.). Sabatier, P. (ed.) (2007), Theories of the Policy Process, 2nd Edition (Boulder: Westview Press). Sager, M.A. (1999), Cooperation without Borders: Federalism and International Trade (Minnesota, New York, Nafta, Gatt) [PhD] Fairfax, VA: George Mason University. Sampson, R. (1991), ‘Linking the Micro- and Macro-level Dimensions of Community Social Organization’, Social Forces 70:1, 43–64. Santander, C. and Matos, Z. (2003), Breakdown of Pastrana’s Peace Process [MA] Syracuse, NY: Syracuse University. Saroglou, V., Pichon, I., Trompette, L., Verschueren, M. and Dernelle, R. (2005), ‘Prosocial Behavior and Religion: New Evidence Based on Projective Measures and Peer Ratings’, Journal for the Scientific Study of Religion 44:3, 323–48. Scanlon, J. (1990), A Study of the Team Approach to Planned Change: Cooperation and Coordination Among Individuals Serving Special Needs Preschool Children [PhD] University Park, PA: The Pennsylvania State University. Schafer, M. (1999), ‘Cooperative and Conflictual Policy Preferences: The Effect of Identity, Security, and Image of the Other’, Political Psychology 20:4, 829– 44. Scott, W. (2001), ‘Foreword’, in J. Shafritz, and S. Ott (eds). Shafritz, J. and Ott, S. (2001), ‘Introduction’, in J. Shafritz and S. Ott (eds).

156

Organizational Cooperation in Crises

Shafritz, J. and Ott, S. (eds) (2001), Classics of Organization Theory, 5th Edition (Orlando: Harcourt). Shah, J., Brazy, P. and Higgins, E. (2004), ‘Promoting Us or Preventing Them: Regulatory Focus and Manifestations of Intergroup Bias’, Personality and Social Psychology Bulletin 30:4, 433–46. Sherman, R. and Eliasson, J. (2006), ‘Trade Disputes and Non-State Actors: New Institutional Arrangements and the Privatization of Commercial Diplomacy’, The World Economy 29:4, 473–89. Shiels, F. (1991), Preventable Disasters: Why Governments Fail (Savage: Rowan & Littlefield Publishers). Shin, Y.-S. (2003), ‘Y2K Crisis in Korea’, MPA Capstone Project, Department of Public Administration, Syracuse, Syracuse University. Shinada, M., Yamagishi, T. and Ohmura, Y. (2004), ‘False Friends are Worse than Bitter Enemies: “Altruistic” Punishment of In-group Members’, Evolution and Human Behavior 25:6, 379–93. Sidanius, J., Pratto, F., van Laar, C. and Levin, S. (2004), ‘Social Dominance Theory: Its Agenda and Method’, Political Psychology 25:6, 845–80. Smith, R. (2005), ‘How Many Have Died in Darfur?’ BBC News [internet] 16 February , accessed 16 August 2006. Staub, E. (2003), ‘Social and Prosocial Behavior: Personal and Situational Influences and Their Interactions’, in E. Staub (eds). Staub, E. (eds) (2003), The Psychology of Good and Evil: Why Children, Adults, and Groups Help and Harm Others (Cambridge: Cambridge University Press). Staw, B., Sandelands, L. and Dutton, J. (1981), ‘Threat Rigidity Effects in Organizational Behavior: A Multilevel Analysis’, Administrative Science Quarterly 26:4, 501–24. Statsoft, 1984–2008 Electronic Textbook [Online] available at: , accessed 26 July 2007. Stepanovic, D. (2003), NATO and the Kosovo Conflict February 23, 1998–March 24, 1999 [MA] Syracuse, NY: Syracuse University. Stern, E. (1997), ‘Probing the Plausibility of Newgroup Syndrome’, in P. ‘t Hart, E. Stern, and B. Sundelius (eds). Stern, E. (1999), Crisis Decisionmaking: A Cognitive Institutional Approach (Stockholm: The Swedish National Defense College). Stern, E. and Bynander, F. (eds) (1998), Crisis and Internationalization: Eight Crises Studies from a Cognitive-Institutional Perspective (Stockholm: ÖCB). Stern, E. and Sundelius, B. (1994), ‘Review: The Essence of Groupthink’, Mershon International Studies Review 38:1, 101–7. Stern, E. and Sundelius, B. (2002), ‘Crisis Management Europe: An Integrated Regional Research and Training Program’, International Studies Perspectives 1:3, 71–88. Stillman, R. (ed.) (2005), Public Administration: Concepts and Cases, 8th Edition (Boston: Houghton Mifflin).

Bibliography

157

Sundelius, B., Stern, E. and Bynander, F. (1997), Krishantering på Svenska: Teori och Praktik (Stockholm, Nerenius Santérus Förlag). Svedin, L. (1998a), Kriminella MC-gäng: Hotbildsanalys, Scenarier och Åtgärdsförslag (Stockholm, Swedish Defence Research Agency). Svedin, L. (1998b), ‘Mc-krig i Sverige och Skandinavien 1994–97 [Biker-wars in Sweden and the Nordic countries 1994–97]’, in F. Bynander and E. Stern (eds). Svedin, L. (1998c), ‘The Red River Flood of 1997: Crisis Mitigation and Resilience’, Unpublished manuscript, Stockholm: CRISMART. Svedin, L. (2001), Perception as a Stepping-Stone to Cooperation [MA] Stockholm, SWE: Stockholm University. Switky, B. (1995), International Cooperation and the Determinants of European Union External Trade Policy: The Case of Textiles and Automobiles [Dissertation, DAI, 56, No. 03A], The Claremont Graduate School. Synonym.com [Website] , 2001–2007 Hillclimb Media, accessed 13 April 2006. Tajfel, H. and Turner, J.C. (1979), ‘An Integrative Theory of Intergroup Conflict’, in W. Austin and S. Worchel (eds). Thee, M. (1977), ‘Arms Control: The Retreat from Disarmament the Record to Date and the Search for Alternatives’, Journal of Peace Research 14:2, 95–114. The Fog of War: Eleven Lessons from the Life of Robert S. McNamara, 2003 [DVD] Morris, E. United States: Sony Pictures (narrated by Robert S. McNamara). Thompson, J. (1997), Cooperation in the Commons: The Emergence and Persistence of Reciprocal Altruism and Collective Action in Farmer-Managed Irrigation Systems in Kenya [Dissertation, DAI, 58, No. 04A], Clark University. Ullberg, S. (2001), Environmental Crisis: The Boliden Dam Rapture (Stockholm: Ateljé/Faktor AB). Verbeek, B. (1994), ‘Do Individual and Group Beliefs Matter? British Decision Making During the 1956 Suez Crisis’, Cooperation and Confict 29:4, 307–32. Vertzberger, Y. (1990), The World in Their Minds: Information Processing, Cognition, and Perception in Foreign Policy Decisionmaking (Stanford: Stanford University Press). Walt, S. (1988), ‘Testing Theories of Alliance Formation: The Case of Southwest Asia’, International Organization 42:2, 275–316. Waltz, K. (1979), Theory of International Politics (Reading: Addison-Wesley). Watson, P. (2003), ‘Exxon Valdez Oil Spill’, MPA Capstone Project, Department of Public Administration, Syracuse, Syracuse University. Weick, K. (1998), ‘Foresights of Failure: An Appreciation of Barry Turner’, Journal of Contingencies and Crisis Management 6:2, 72–5. Williams, P. and Bellamy, A. (2005), ‘The Responsibility to Protect and the Crisis in Darfur’, Security Dialogue 36:1, 27–47. Wispé, L. (1972), ‘Positive Forms of Social Behavior: An Overview’, Journal of Social Issues 28:3, 1–19.

158

Organizational Cooperation in Crises

Wood, D. and Peake, J. (1998), ‘The Dynamics of Foreign Policy Agenda Setting’, American Political Science Review 92:1, 173–84. Wu, F. (1989), Information Systems Supported End-User Interaction: Impacts on Inter- and Intra-Organizational Connectivity and Cooperative Behavior (Interorganizational) [PhD] The University of Texas at Austin. Wydick, B. (1999), ‘Can Social Cohesion be Harnessed to Repair Market Failures? Evidence from Group Lending in Guatemala’, Economic Journal 109:457, 463–475. Yamagishi, T., Makimura, T., Foddy, M., Matsuda, M., Kiyonari, T. and Platow, N. (2005), ‘Comparisons of Australian and Japanese on Group-based Cooperation’, Asian Journal of Social Psychology 8:2, 173–190. Yarbrough, B. and Yarbrough, R. (2003), ‘The Contractual Role of Boundaries: Law and Economics Meets International Organization’, European Journal of International Relations 9:4, 543–590. Yoo, H.-S. (2003), Wido Crisis in Korea: Decision on Nuclear Waste Disposal Site and Residents’ Protest [MA] Syracuse, NY: Syracuse University. Young, O. (1989), International Cooperation: Building Regimes for Natural Resources and the Environment (Ithaca: Cornell University Press). Yum, B. (2003), The 1997 Hanbo Scandal as an Initial Trigger of the Korean Financial Crisis [MA] Syracuse, NY: Syracuse University.

Index actors, 1, 2, 4, 5, 7, 10, 15–9, 21–2, 24–5, 28–9, 31, 34–5, 37–9, 41, 45–55, 57, 66–9, 71, 73, 78, 80–81, 84–5, 87–8, 90–1, 96–7, 99–100, 102, 111, 114, 116, 130, 140 agreement, 7, 24, 26, 28–9, 31, 51, 58, 62, 64–9, 77, 80, 88, 94, 123, 134 break agreement, 31, 41, 74, 76–7, 94–6, 98, 102–105, 123, 126–7 make/come to agreement, 28–9, 58, 62, 64–7, 69, 94, 123 Aiken, M., 38 alliance, 22, 46 altruism, 15, 21–2, 32–3, 52 approve, 64, 70, 97–8 express approval, 28–9, 98, 123, 126–7 attributing fault to others, 31, 40–41, 48, 51, 90, 99, 111 authority, 8, 10, 54, 63 authority to make decision, 54 Axelrod, R., 25, 37, 111, 124 bankruptcy, 50, 86 bargaining, 23, 25, 35, 39 battle of the good Samaritans, 40 Baumgartner, F., 19 behavior, 10–12, 19, 25, 27, 31–2, 34–5, 37–8, 40, 43–4, 46, 49, 50, 52–62, 64–8, 71–81, 83, 85, 86, 90, 93, 95, 97, 99, 100, 101, 103, 105, 108, 110, 115–26, 132, 134, 139–40, 142, 145, 147, 152, 155–7, 159, 163–72, 175–6, 178–9, 180, 182, 184–7, 192–3, 197–202, 206, 209, 211, 217 competitive behavior, 32, 38, 87, 125, 132 conflictive behavior, 48, 50, 135 covert behavior, 31, 40, 74, 76–7, 80, 94–5, 123 neutral behavior, 125, 134–5

pro-social behavior, 3, 32, 52 blame-game, 31, 48, 49, 123 board of directors, 66, 82–4 Boin, A., 1, 9, 39, Brändström, A., 24, 41–2, 90, 134 Brecher, M., 13, 19, 23, 122 Bynander, F., 5, 19–20, 42, 90 Canada, 52, 53 Carter, 64 case example, 13, 45–6, 49–52, 54, 63–70, 79–82, 85–6, 88–9 case research, x, 15 categorical principal component analysis, 12, 57, 62, 77 CATPCA, 43, 58–9, 63, 73–7, 90, 109 cohesion, 9, 15, 21–3, 25, 35, 52, 80–81, 100–101, 141 collective action, 17, 21–2 collective goods, 2 Comfort, L., 5 comment, 28–9, 58, 62, 66, 70, 94, 99, 100, 123, 127 communication, 24, 28–31, 35, 38, 78–80, 85, 90, 104, 134, 137, 140 competition, 7, 11, 23, 40, 47, 53, 83, 88 component, 12, 14, 17, 43–4, 57, 60–70, 73, 75, 76–8, 80–81, 84–5, 88, 90, 106, 108–9, 122, 142 consult, 28–9, 47, 58, 62, 66, 70, 94, 98, 111, 123 content slippage, 31, 39, 49, 74, 76–8, 94–6, 101–3, 105, 123, 126–7 cooperate, 1–5, 7, 10, 12, 18, 23–4, 26, 28–9, 33–4, 38, 50, 53–4, 58, 62, 64, 66, 94, 105, 123, 126, 133–7, 140–1 decide to cooperate, 28–9, 58, 62, 64, 66, 94, 123 cooperation, 1–39, 41–55, 58, 60, 61, 66, 70–71, 73–4, 76–7, 80–81, 84–5,

160

Organizational Cooperation in Crises

88–91, 93, 96, 98, 100, 102, 104, 106–111, 113, 116–8, 120, 123–36, 139–43 cooperative behavior, 12, 22, 25–8, 32–3, 35, 38, 41, 43, 47, 50, 53, 55, 57, 61–4, 66–7, 69–71, 96, 108, 110–11, 115, 118, 122, 124–5, 129, 134–5, 140 coordinate, 2, 4, 17, 31, 38, 51, 54, 74, 76–8, 91, 94–5, 99, 100, 123, 127 coordination, 1–3, 5, 8–9, 17, 22–5, 35, 38, 49, 51–2, 54, 79, 90, 140 correlation, 44, 57, 60, 61, 90, 93, 95–6, 98–110, 116–8, 120–22, 126–9, 131 correlation analysis, 44, 93, 96, 107, 109, 120, 122, 126 credit-seeking, 31, 39, 51, 123 crisis, x, 1–20, 23–4, 26–9, 32–3, 35–6, 38–55, 63–7, 69, 71, 73–4, 78, 80–86, 88–91, 93, 96–106, 110–11, 113–141, 143 crisis case, 11, 14, 17, 26, 28, 32, 36, 42, 45, 52, 55, 63, 66, 69, 73–4, 80–81, 88, 96, 97–105, 110, 116, 123, 126–7, 130 crisis communication, see communication crisis coordination, see coordination crisis decision-making, see decisionmaking crisis definition, 10, 17, 19–20, 36, 42, cut-off point, 60–61, 75, cut-off criterion, see cut-off point cut-off level, see cut-off point Cyert, R., 2 Daewoo, 65, 88 Dasgupta, N., 9 Dayton, B., ix, 13, 20, 41 decision-making, ix, 2, 8–12, 15, 18, 20–21, 25–6, 35–6, 42, 53–4, 80, 97, 102, 106, 109, 123, 125–6, 135, 137, 140–41 definition of crisis, see crisis definition demand, 3, 28–30, 34, 40, 58, 62–3, 95, 101, 123 make a demand, 28–30

Deng, P., 22 Department of Homeland Security, 1–2 Derlega, V., 22, 32 disapprove, 58, 62–3, 69, 95, 101 express disapproval, 28, 30, 102, 123, 127 discuss, 28–9, 31, 45, 123, 134 Drescher, S., 22 Dutton, J., 35 Dynes, R., 38 enforce, 1, 54, 70 Evangelista, M., 37 extraction method, 58, 73 facilitate, 20, 34, 38, 51–2, 78, 98–9, 102–3, 109, 123–4, 127, 132, 139–41, 143 facilitate in-group’s goals, 51, 98, 123 facilitate others’ goals, 34, 52, 99, 102–3, 109, 123, 127 factor, see factor analysis factor analysis, 12, 43, 44, 106, 142 factor score, see factor analysis factors, 2, 5, 9,15, 17–9, 24, 42, 132, 135, 137, 139, 140–2 perceptual factors, 140, 141, 142 psychological factors, 16, 18, 21–2, 32, 49, 52, 55, 126, 140–41 structural factors, 5, 17, 55, 73, 77, 139–40, 142 FARC, 68, 70 FBI, 69, 79 Field, A., ix, 62, 73, 75 financial, x, 37, 39, 49, 50–1, 54–5, 65, 82–4, 86–7, 140 financial crisis, 49, 50–51, 54–5, 65, 82–3, 140 financial institutions, 49, 65 flood, 5, 52–4 formal arrangements, 29, 64, 78, 88, 133 formal relations, see formal arrangements free-rider problem, 52 gains, 7, 25 relative gains, 7

Index game, 4, 7, 11, 12, 20, 23–5, 31, 33, 37, 40, 48–9, 74, 76, 78, 80–81, 90, 94–5, 99, 111, 123, 135 game theory, 7, 11, 20, 23, 25, 33, 37, 135 n-turn, 12, 111 prisoners’ dilemma, 24, 111, Game of Old Maid, 31, 40, 41, 73–4, 123 George, A., 9, 20, 23, 41–2 Gerner, D., 27–8, 124 Goldstein, J., 27 Gowa, J., 21–22 group, ix, 3, 7, 9, 15, 16, 18–9, 21–3, 25, 27, 31, 35–7, 40, 42, 48, 50–52, 54, 65, 68, 73–4, 79–81, 84–6, 88, 98, 100–101, 115–6, 120, 122–3, 130, 132–7, 141 group decision-making, 15, 21, 25, 35 group dynamics, 9, 15, 21–2, 52, 80–81, 134, 141, groupthink, 25, 31, 35–7, 49, 51, 74, 76, 80–81, 94–5, 99–101, 123, 127, 134 Grzelak, J., 22, 32 Hage, J., 35, 38 Halperin, M., 7, 40 Hanbo Group, 86 Hansén, D., ix, 42 harmony, 32, 34–35 Hayashi, N., 22 Hermann, C., 9, 23 Hermann, M., ix, 6, 9, 13, 20, 41, 43, Holsti, O., 9, 19, 23, 122 honesty, 20, 31, 37–38, 49, 54, 74, 76, 84, 85, 94–95, 97–100, 111, 126–127 claim honesty, 31, 94–95, 97–99, 123, 126–127 display honesty, 31, 94–95, 123 Hood, C., 41, 43, 90 Huddy, L., 9 incentives, 5, 26, 33, 134 indicator, 29, 30, 33, 35, 47, 61–66, 69–71, 75, 77–78, 80–81, 85, 88, 91, 96, 97, 105, 106, 110–111 informal, 133 institutional, 8, 10–11, 45–6

161

institutional design, 8 institutionalization, 18 interaction, 4, 6–12, 15–16, 18–24, 26–7, 32–3, 37, 39, 45–6, 49, 52, 74, 76, 78–9, 84–5, 110–111, 113, 122, 124–6, 131–3, 140–41, 143 interaction setting, 10, 20–21, 45, 111, 113, 131, 140 interaction strategy, 126, 132 international organizations, 4, 46 Iran hostage crisis, 64 Janis, I., 35–6 Jervis, R., 5, 19 Jones, B., 19 Jost, J., 9, 33 Keohane, R., 25, 34–5, 37–9 Kettl, D., 5 Kickert, W., 8 Kingdon, J., 19, 90 Koppenjan, J., 8 Korea, 49–51, 85–6 Kosovo crisis, 46–8, 51 Krasner, S., 21 levels of analysis, 12–13, 16, 21, 26, 28, 31, 41–2, 57, 59–60, 64, 69, 73, 75, 90, 93, 96, 110, 116, 123–6 Linting, 43–4, 59–60, 106 Lipsky, M., 18 loading, 61–2, 64, 75, 77–8, 80, 85, 88 Mandell, M., 22 March, J., 2 Marrett, C., 38 Martin, L., 21, 24, 35, 39 Meulman, 43, 59, 106 Milner, H., 24–5 Milward, H., 34 minorities, 89 NATO, 46–7, 80, 88 need for contact person, 31, 38, 74, 76–8, 84–5, 90–91, 94–5, 99–100, 104–5, 123, 127 network, 2, 8, 78, 123, 134, 140

162

Organizational Cooperation in Crises

new group syndrome, 31, 35, 37, 73–4, 123, 134 Newlove, L., 41, 43, 78

psychological research, see factors; psychological Putnam, R., 22

O’Toole, L., 35 occasion for decision, 13, 42, 59–60, 63, 66–7, 69–70, 123, 125 oppression, 30, 69, 105 organization, 2, 7–8, 10–12, 15, 17–20, 23, 25–6, 28–30, 32–4, 40–1, 47–9, 52–4, 70, 79, 82–4, 96–107, 109–110, 115–6, 118, 126–8, 130, 133–7, 140–41 organizational, 1–2, 4–18, 20–8, 32, 34, 38–43, 45, 49, 53, 55, 57, 73, 93, 107, 111, 113, 123–6, 128–9, 131–7, 139–43 organizational psychology, 9 organizational relations, 8, 15, 21–2, 139–140, 142 organizing, 100, 123, 133–43 Oye, K., 37

qualitative case studies, 13, 28, 143 quantitative data, 15, 143

Pastrana’s Peace Process, 68, 70, 80 perception, 3, 6, 9–10, 15, 19–22, 39, 51–2, 81, 89, 97, 111, 114, 122, 137, 141 perceptual factors, 140–42 Perrow, C., 78 Perry, J., 3 Pevehouse, J., 27 politicization, 90, 134 Powell, W., 22 practitioner, 132–7 predictors, 109, 119, 128 principal component analysis, 12, 14, 43–4, 57, 61–2, 73, 75, 77, 106 problem, 3–4, 7, 9–10, 16, 31, 36, 40, 50, 65, 70, 79–81, 84, 89–90, 99, 114, 132, 134, 137, 141 problem frame, 3–4 problem framing, 80, 134, 137, 141 propose, 28–9, 58, 62, 69–70, 94, 97, 123, 135, 142 pro-social behavior, see behavior; pro-social Provan, K., 34 psychological variables, see factors; psychological

rally around the flag, 2, 31, 35, 74, 76, 80–81, 94–5, 123 rational, 7, 13, 18, 23 Red River, 52–5 reduce relations, 28, 30, 58, 62–3, 95, 104, 123 regimes, 21–22, 24, 46 regression, 44, 93, 118–120, 142 forward regression, 137, 139, 232 regression analysis, 120, 142 regression results, 120 Reicher, S., 9, 23 reject, 28, 30, 50, 58, 62, 67, 69, 95, 103–4, 123 repeated interaction, 12 request, 28, 29, 53, 58, 62–3, 69–71, 94, 97, 123, 126 research questions, 11–12, 123, 125, 139–41 resources, 1–2, 8, 11, 17, 34, 40, 134, 140 riots, 6, 42, 89 Rosenthal, U., 1, 5, 9, 19, 39–40 rotation, 61–2, 74, 76, 90 Ruby Ridge, 79 Ruggie, J., 37 rules, 7, 17–18, 21, 65, 70, 78–79, 134–5, 140 Sabatier, P., 11 Sandelands, L., 35 Schrodt, P., 27–8 self-interest, 4, 7, 33 selfishness, 88 unselfish, 91 Sidanius, J., 9 Simmons, B., 21 situational factors, 139–140 social capital, 18, 22–3, 33, 52 social contract, 2 social coordination, 9, 22–3, 52

Index social psychology, 6, 14–5, 21–2, 45, 52, 133, 142 stakeholders, 8, 17, 28–30, 66, 79, 84, 100, 136–8, 140 stalling implementation, 31, 39, 49, 74, 76–8, 94–5, 97, 123 Staub, E., 32 Staw, B., 35, 40 Stern, E., 17, 19–20, 23–5, 35–7, 41–3 strategy competitive strategy, 6, 11, 26, see also behavior; competitive conflictive strategy, 143 see also behavior; conflictive cooperation strategy, 12, 26, 32, 81, 84–5, 88, 98, 100, 104, 107–111, 116, 120, 128–130 cooperative strategy, 32, 80, 102, 110, 122, 129, 136 see also behavior; cooperative structural variables, see factors; structural structural violence, see violence; structural Sundelius, B, ix, 5, 17, 19–20, 25, 36–7, 39, 41–3 surprise, 89, 100, 113, 117–122 Svedin, L., x, 5, 10, 40, 42–3, 52–4 ‘t Hart, P., 1, 5, 24–5, 39–41, 43 Tajfel, H., 9, 23 threat, 1–2, 5–6, 9–10, 12, 14, 19–20, 28, 30–31, 35–36, 40, 46, 55, 63, 80–82, 84, 90, 99, 108, 113–122, 130–132, 134, 137 power of threat, 117 threat inside, 114–7, 119–122, 131–2 threat origin, 113, 117 threat outside, 36, 55, 80–81, 114–117, 119–122, 130–132, 135–6 threat perception, 10, 81, 114, 122, 137 threat severity, 119–121, 131 threat to values, 1, 10, 15, 19–20, 36, 113, 114, 140 threaten, 28, 30, 58, 62–4, 95, 103–4, 123 Three Mile Island, 67 trade, 15, 18, 21–2, 46, 49 training, 5, 8, 10–11, 140

163

Transboundary Crisis Management, ix, 13, 41–3, 45, 63, 66, 113, 123, 143 TCM dataset, xi, 13, 41, 43, 63, 66, 123, 143 Turner, J., 9, 23 uncertainty, 2, 5, 10, 12, 14–15, 19–20, 36–38, 82, 90, 113, 115–122, 130 uncertainty with regard to the definition of the crisis, 115–116, 118, 120–121, 130, 133 uncertainty with regarding how to respond to the crisis, 115–6, 118–121, 130–1, 133 urgency, 5, 9–10, 12, 14–15, 19–20, 36, 82, 90, 113–5, 117–121, 131 value, 2, 33, 36, 38, 42–3, 57, 59, 61, 82–3, 106 core values, 10, 15, 20, 36, 113 Verbeek, B., 23 Vertzberger, Y., 6 violence, 5, 28, 30, 48, 58, 62, 69, 82, 95, 105, 123 government-sponsored structural violence, 28, 30, 69, 123 violence against civilians/persons, 30, 69, 105 violence against property, 30, 69, 105 violence against rights, 30, 69, 105 voluntary concessions, 31–4, 48, 51, 74, 76–8, 94–6, 102, 104, 105, 123, 126–7 Walt, S., 19 Waltz, K., 19 Wilkenfeld, J., 13, 19, 23 Wise, L., 3 Wispé, L., 32 Yamagishi, T., 9, 22 Yarbrough, B., 2 yield, 28–9, 58, 62, 67, 69, 94, 96–7, 123 YMCA, 82–4 Young, O., 21, 86 zero sum, 134

This page has been left blank intentionally