EDMG330 Assignment

1 Crisis management in political systems: five leadership challenges

1.1 Crisis management and public leadership

Crises come in many shapes and forms. Conflicts, man-made accidents, and natural disasters chronically shatter the peace and order of societies. The new century has brought an upsurge of international terrorism, but also a creeping awareness of new types of contingencies - breakdowns in information and communication systems, emerging natural threats, and bio-nuclear terrorism - that lurk beyond the horizon.' At the same time, age-old threats (floods, earthquakes, and tsunamis) continue to expose the vulnerabilities of modern society.

In times of crisis, citizens look at their leaders: presidents and mayors, local politicians and elected administrators, public managers and top civil servants. We expect these policy makers to avert the threat or at least minimize the damage of the crisis at hand. They should lead us out of the crisis; they must explain what went wrong and convince us that it will not happen again.

This is an important set of tasks. Crisis management bears directly upon the lives of citizens and the wellbeing of societies. When emerging vulnerabilities and threats are adequately assessed and addressed, some potentially devastating contingencies simply d o not happen. Mispercep- tion and negligence, however, allow crises to occur. When policy makers respond well to a crisis, the damage is limited; when they fail, the crisis impact increases. In extreme cases, crisis management makes the difference between life and death.

These are no easy tasks either. T h e management of a crisis is often a big, complex, and drawn-out operation, which involves many organiza- tions, both public and private. T h e mass media continuously scrutinize and assess leaders and their leadership. It is in this context that policy makers must supervise operational aspects of the crisis management operation, communicate with stakeholders, discover what went wrong, account for their actions, initiate ways of improvement, and (re)establish I sense of normalcy. T h e notion "crisis management" as used in this

2 The Politics of Crisis Management Crisis management in political systems 3

book is therefore shorthand for a set of interrelated and exmaordinary vnvemance challenges. It provides an ultimate test for the resilience of u- - - political systems and their elites.

This is a book on public leadership in crisis management. It examines how public leaders deal with this essential and increasingly salient task of contemporary governance. It maps the manifold challenges they face in a crisis and identifies the pitfalls public leaders and public institutions encounter in their efforts t o manage crises. T o do SO, we must "unpack" the notions of crisis and crisis management. In this introductory chapter, we begin this task by outlining our perspective on crisis management. First, we explain what we mean by the term "crisis." Then we argue that crises are ubiquitous phenomena that cannot be predicted with any kind of precision. Next, we outline our perspective on crisis leadership. Finally, we present five key leadership tasks in crisis management, which form the backbone of this book.

1.2 The nature of crisis

The term "crisis" frequently features in book titles, newspaper headlines, political discourse, and social conversation. It refers to an undesirable and unexpected situation: when we talk about crisis, we usually mean that something bad is to befall a person, group, organization, culture, society, or, when we think really big, the world at large. Something must be done, urgently, to make sure that this threat will not materialize.

In academic discourse, a crisis marks a phase of disorder in the seemingly normal development of a system.' An economic crisis, for instance, refers to an interval of decline in a long period of steady growth and development. A personal crisis denotes a period of turmoil, pre- ceded and followed by mental stability. A revolution pertains to the abyss between dictatorial order and democratic order. Crises are transi- tional phases, during which the normal ways of operating no longer work.3

Most people experience such transitions as an urgent threat, which policy makers must a d d r e ~ s . ~ Our definition of crisis reflects its subject- ive nature as a construed threat: we speak of a crisis when policy makers experience "a serious threat to the basic structures or the hndamental values and norms of a system, which under time pressure and highly uncertain circumstances necessitates making vital decision^."^

Let us consider the three key components - threat, uncertainty, ur- gency - of this crisis definition in somewhat more detail. Crises occur when core values or life-sustaining systems of a community come under threat. Think of widely shared values such as safety and security, welfare

and health, integrity and fairness, which become shaky or even mean- ingless as a result of (looming) violence, destruction, damage, or other forms of adversity. T h e more lives are governed by the value(s) under threat, the deeper the crisis goes. That explains why a looming natural disaster (flood, earthquake, hurricane, extreme heat or cold) never fails to evoke a deep sense of crisis: the threat of death, damage, destruction, or bodily mutilation clearly violates the deeply embedded values of safety and security for oneself and one's loved ones.

T h e threat of mass destruction is, of course, but one path to c r i ~ i s . ~ A financial scandal in a large corporation may touch off a crisis in a society if it threatens the job security of many and undermines the trust in the economic system. In public organizations, a routine incident can trigger a crisis when media and elected leaders frame the incident as an indica- tion of inherent flaws and threaten to withdraw their support for the organization. T h e anthrax scare and the Washington Beltway snipers caused the deaths of relatively few people, but these crises caused wide- spread fear among the public, which - in the context of the 911 1 events - was enough to virtually paralyze parts of the United States for weeks in a

In other words, a crisis does not automatically entail victims or damages.'

Crises typically and understandably induce a sense of urgency. Serious threats that d o not pose immediate problems - think of climate change or h t u r e pension deficits - do not induce a widespread sense of crisis.9 Some experts may be worried (and rightly so), but most policy makers do not lose sleep over problems with a horizon that exceeds their polit- ical life expectancy. Time compression is a defining element of crisis: the threat is here, it is real, and it must be dealt with as soon as possible (at least that's the way it is perceived).

Time compression is especially relevant for understanding leadership at the operational level, where decisions on matters of life and death must sometimes be made within a few hours, minutes, or even a split second. Think of the commander of the US cruiser Vincennes who had only a few minutes to decide whether the incoming aircraft was an enemy (Iranian) fighter or a non-responsive passenger plane - it tragic- ally turned out to be the latter.I0 Leaders at the strategic level rarely experience this sense of extreme urgency, but their time horizon does become much shorter during crises.

In a crisis, the perception of threat is accompanied by a high degree of "ncenainzy. This uncertainty pertains both to the nature and the poten- tial consequences of the threat: what is happening and how did it happen? What's next? How bad will it be? More importantly, uncertainty clouds the search for solutions: what can we do? What happens if we

4 The Politics of Crisis Management

select this option? Uncertainty typically applies to other factors in the crisis process as well, such as people's initial and emergent responses to the crisis.

This definition of crisis enables us to study a wide variety of adversity: hurricanes and floods; earthquakes and tsunamis; financial meltdowns and surprise attacks; terrorist attacks and hostage takings; environmental threats and exploding factories; infrastructural dramas and organiza- tional decline - there are many unimaginable threats that can turn leaders into crisis managers. What all these dramatic events have in common is the impossible conditions they create for leaders: managing the response operation and making urgent decisions while essential information about causes and consequences remains unavailable.

This is, of course, an academic shortcut on the way toward under- standing crisis management. We know that in real life it is not always clear when exactly policy makers (who are they anyway?) experience a situation in terms of crisis. Some situations seem crystal clear, some are surely debatable. This fits our notion of crisis development: the defin- ition of a situation in terms of crisis is the outcome of a political process. Certain situations "become" crises; they travel the continuum from the "no problem" pole to the "deep crisis" end (and back). In our choice of literature and examples, we have tried to err on the safe side: we have selected crisis cases that most informed readers would probably categor- ize (if they were asked to) as situations of combined societal threat, urgency, and uncertainty.

We are also aware that the management of crisis may depend on the type of threat. A traditional distinction is the one between natural and man-made disasters. Managing the impact of a tsunami (killing tens of thousands) or the explosion of a fireworks factory (killing ten) involves different activities as most of us can undoubtedly imagine. However, we claim that the strategic - as opposed to the tactical and operational - challenges for leaders in dealing with these threats are essentially the same: trying to prevent or at least minimize the impact of adversity, deal with the social and political consequences, and restore public faith in the future. In fact, we take our argument one step further: leaders can prepare for crises of the future - always different from past events - only if they learn from the variety of experiences they themselves and other leaders have had in other typcs of crisis.

1.3 The ubiquity of crisis

Disruptions of societal and political order are as old as life itself." The Bible can be read as an introductory expose of the frightening crises that

Crisis management in political systems 5

have beset mankind. Western societies may have rooted out many of these adverse events, but most of the world still confronts these "old" crises on a daily basis. The costs of natural and man-made disasters continue to grow, while scenarios of future crises promise more mayhem. I "

Crises will continue to challenge leaders for a simple reason: the disruprions that cause crises in our systems cannot be prevented. This bold assertion arises from recent thinking about the causes of crises. It is now clear to most people that crises are not due to bad luck or God's punishment.13 Linear thinking ("big events must have big causes") has given Lvay to a more subtle perspective that emphasizes the unintended consequences of increased c o m p l e x i t y . ' ~ r i s e s , then, are the result of

causes, which interact over time to produce a threat with devastating potential.

This perspective is somewhat counterintuitive, as it defies the trad- itional logic of "triggers" and underlying causes. A common belief is that some set offactors "causes" a crisis. We then make a distinction between "external" and "internal" triggers. While this certainly facilitates conver- sation (both colloquial and academic), it would be more precise to speak of escalatory processes that undermine a social system's capacity to cope with disturbances. The agents of disturbance may come from anywhere - ranging from earthquakes to human errors - b u t the cause of the crisis lies in the inability of a system to deal with the disturbance.

An oft-debated question is whether modem systems have become increasingly vulnerable to breakdown. Contemporary systems typically experience fewer breakdowns, one might argue, as they have become much better equipped to deal with routine failures. Several "modem" features of society hospitals, computers and telephones, fire trucks and univenities, regulation and funds - have made some W e s of crisis that Once were rather ubiquitous relatively rare. Others argue that the resili- ence of modem society has deteriorated: when a threat does materialize (say an electrical Power outage), modem societies suffer disproportion- a l ' ~ . The point is often made by students of natural disasters: modem Society increases its vulnerability to disaster by building in places where

warns not to build. - 1 - k causes of crises thus seem to reside within the system: the causes

E P i c a l l ~ remain unnoticed, or key policy makers fail to attend to them.15 In the Process leading up to a crisis, seemingly innocent factors combine and transform into disruptive forces that come to represent an undeni- able threat to the system. These factors are sometimes referred to as pathogens, as they are typically present long before the crisis becomes manifest. I ('

6 T h e Politics of Crisis Management Crisis management in political systems 7

T h e notion that crises are an unwanted by-product of complex systems has been popularized by Charles Perrow's (1984) analysis of the nuclear power incident at Three Mile Island and other disasters in technological systems." Perrow describes how a relatively minor glitch in the plant was misunderstood in the control room. T h e plant operators initially thought they understood the problem and applied the required technical response. As they had misinterpreted the warning signal, the response worsened the problem. T h e increased threat baffled the oper- ators (they could not understand why the problem persisted) and invited an urgent response. By again applying the "right" response to the wrong problem, the operators continued to exacerbate the problem. Only after a freshly arrived operator suggested the correct source of the problem did the crisis team manage - just barely - to stave off a disaster.

T h e very qualities of complex systems that drive progress lie at the heart of most if not all technological crises. As socio-technical systems become more complex and increasingly connected (tightly coupled) to other (sub)systems, their vulnerability for disturbances increases expo- nentially.18 T h e more complex a system becomes, the harder it is for anyone to understand it in its entirety. Tight coupling between a system's component parts and with those of other systems allows for the rapid proliferation of interactions (and errors) throughout the system.

Complexity and lengthy chains of accident causation do not remain confined to the world of high-risk technology. Consider the world of global finance and the financial crises that have rattled it in recent

19 years. Globalization and I C T have tightly connected most world markets and financial systems. As a result, a minor problem in a seem- ingly isolated market can trigger a financial rncltdoivn in markets on the other side of the globe. Structural vulnerabilities in relatively weak economies such as Russia, Argentina, or Turkey may suddenly "explode" on Wall Street and cause worldwide economic decline.

T h e same characteristics can be found in crises that beset low-tech environments such as prisons or sports stadiums. Urban riots, prison disturbances, and sports crowd disasters seem to start off with relatively minor incidents." Upon closer inspection, however, it becomes clear that it is a similar mix of interrelated causes that produces major out- bursts of this kind. In the case of prison disturbances, the interaction between guards and inmates is of particular relevance. Consider the 1990 riot that all but destroyed the Strangeways prison in Manchester ( U K ) . ~ ~ In the incubation period leading up to the riot, prison guards had to adapt their way ofworking in the face of budgetary pressure. This change in staff behavior was negatively interpreted by inmates, who began to challenge staff authority, which, in turn, generated anxiety

and stress among staff. As staff began to act in an increasingly defensive and inconsistent manner, prisoners became more frustrated with staff behavior. A reiterative, self-reinforcing pattern of changing behavior and staff-prisoner conflict set the stage for a riot. A small incident started the riot, which in turn touched off a string of disturbances in other prisons.22 Many civil disturbances between protestors and police unfold according to the same pattern.23

Non-linear dynamics and complexity make a crisis hard to detect. As complex systems cannot be simply understood, it is hard to qualify the manifold activities and processes that take place in these systems.'.' Growing vulnerabilities go unrecognized and ineffective attempts to deal with seemingly minor disturbances continue. T h e system thus "fuels" the lurking crisis." Only a minor "trigger" is needed to initiate a de- structive cycle of escalation, which may then rapidly spread throughout the system. Crises may have their roots far away (in a geographical sense) but rapidly snowball through the global networks, jumping from one system to another, gathering destructive potential along the way.

Is it really impossible to predict crises? Generally speaking, yes. There is no clear "moment X" and "factor Y" that can be pinpointed as the root of the problem. Quite sophisticated early-warning systems exist in cer- tain areas, such as humcane and flood prediction, and some pioneering efforts are under way to develop early-warning models for ethnic and international ~ o n f l i c t . ' ~ These systems may constitute the best available shot at crisis prediction, but they are far from flawless. They cannot predict exactly when and where a humcane or flash flood will emerge. In fact, the systems in place can be dangerously wrong.

All this explains why some of the most notorious crises of our times were completely missed by those in charge. As the crisis process begins to unfold, policy makers often do not see anything out of the ordinary. Everything is still in place, even though hidden interactions eat away at the pillars of the system. It is only when the crisis is in full swing and becomes manifest that policy makers can recognize it for what it is. There are many reasons for this apparent lack of foresight, which we will discuss in Chapter 2.

1.4 Crisis management: leadership perspectives

Crises that beset the public domain - this may happen at the local, regional, national, or transnational level - are occasions for public lead- ership. Citizens whose lives are affected by critical contingencies expect governments and public agencies to do their utmost to keep them out of h a m ' s way. They expect the people in charge to make critical decisions

8 The Politics of Crisis Management Crisis management in political systems 9

and provide direction even in the most di6cult circumstances. SO do the journalists who produce the stories that help to shape the crisis in the minds of the public. And so do members of parliament, public interest groups, institutional watchdogs, and other voices on the political stage that monitor and influence the behavior of leaders.

However misplaced, unfair, or illusory these expectations may be, it hardly matters. These expectations are real in their political conse- quences. When events or episodes are widely experienced as a crisis, leadership is expected. If incumbent elites fail to step forward, others might well seize the opportunity to fill the gap.

In this book, we confine ourselves to crisis management in democratic settings. T h e embedded norms and institutional characteristics of liberal democracies markedly constrain the range of responses that public leaders can consider and implement. Many crises could be terminated relatively quickly when governments can simply "write o f f certain people, groups, or territories, or when they can deal with threats regard- less of the human costs or moral implications of their actions. In coun- tries with a free press, a rule of law, political opposition, and a solid accountability structure this is not possible.

In a liberal democracy, public leaders must manage a crisis in the context of a delicate political, legal, and moral order that forces them to trade off considerations of effectiveness and efficiency against other embedded values - something leaders of non-democracies do not have to worry about as much.27

If crisis management was hard, it is only getting harder. T h e demo- cratic context has changed over the past decades. Analysts agree, for instance, that citizens and politicians alike have become at once more f e a h l and less tolerant of major hazards to public health, safety, and prosperity. The modern Western citizen has little patience for imper- fections; he has come to fear glitches and has learned to see more of what he fears. In this culture of fear - sometimes referred to as the "risk society" - the role of the modem mass media is c r ~ c i a l . ' ~

A crisis sets in motion extensive follow-up reporting, investigations by political forums, as well as civil and criminal juridical proceedings. It is not uncommon for public officials and agencies to be singled out as the responsible actors for prevention, preparedness, and response in the crisis at hand. The crisis aftermath then turns into a morality play. Leaders must defend themselves against seemingly incontrovertible evidence of their incompetence, ignorance, or insensitivity-. When their strategies fail, they come under severe pressure to atone for past sins. If they refuse to bow, the crisis will not end (at least not any time soon).

This study aims to capture what leadership in crises entails. We are interested to learn how public leaders seek to protect their society from

how they prepare for and cope with crises. T o organize our inquiry, we define leadership as a set of strategic tasks that encompasses all associated with the stages of crisis management.29

This perspective does not presume that these tasks are exclusively for leaders only. On the contrary: these tasks are often per-

formed throughout the crisis response network. In fact, during a crisis one may find situational leadership, which diverges from regular, formal leadership arrangements. We do believe, however, that the formal leaders carry a special responsibility for making sure that these tasks - which we specify in the following section - are properly addressed and executed (if not by the leaders then by others).

We do not wish to suggest that the performance of a set of tasks will ~ r o v i d e fool-proof relief from crises (of whatever kind). This would be both a presumptuous claim and one-sidedly instrumental. It would deny the pivotal, yet highly volatile and complex political dimension of crises and crisis management.'' In all fairness, one could criticize the field of crisis management studies for its overtly instrumental orientation. There is a large and fast-growing pile of self-help, how-to books that promise to make organizations crisis free.

Our book is an attempt to redress this imbalance. We view crisis management not just in terms of the coping capacity of governmental institutions and public policies but first and foremost as a deeply contro- versial and intensely political activity. We want to find out what crises "do" to established political and organizational orders; we seek to under- stand how crisis leadership contributes to defending, destroving, or renovating these orders. The distinctive contribution we seek to make is to highlight the political dimensions of crisis leadership: issues of conflict, power, and legitimacy.31

We thus use a more task-related than person-related perspective on crisis leadership. In general discourse, leaders are often seen as the personification of leadership. This is the myth of the "great" leader, which pervades so many efforts to understand both great accomplish- ments and massive failures. In this book we talk loosely of policy makers and leaders, but we concentrate on the efforts of all those holding high ofices and strategic positions from which public leadership functions can be performed. Hence our "sample" of leaders includes presidents, prime ministers, cabinet ministers, senior civil servants, and public managers. We agree that charismatic bonds between leaders and follow- ers, and personal idiosyncracies of policy makers may be important to explain how certain leadership tasks are fulfilled, but we are more

10 The Politics of Crisis Management Crisis management in political systems 11

interested to see how the performance of these tasks relates to the crisis

T h e adjective "strategic" is important here: we study the overall direction of crisis responses and the political process surrounding these responses. This book is not about operational commanders and their leadership predicaments, however important these have proven to be in resolving various types of crisis. Moreover, we only touch upon the more technical activities of the comprehensive crisis management con- tinuum (such as risk assessment or the use of tort law).33 Let us now turn to the key challenges of crisis leadership.

1.5 Leadership in crisis: five critical tasks

T h e normative assumption underlying our approach is that public leaders have a special responsibility to help safeguard society from the adverse consequences of crisis. Leaders who take this responsibility seriously would have to concern themselves with all crisis phases: the incubation stage, the onset, and the aftermath. In practice, policy makers have defined the activities of crisis management in accordance with these stages - they talk about prevention, mitigation, critical decision making, and a return to normalcy. We stick closely to this phase model of crisis management, but we have slightly adapted it to account for the political perspective used in this book.

Crisis leadership then involves five critical tasks: sense making, deci- sion making, meaning making, terminating, and learning. We devote one chapter to each of these tasks. We present our reading of the relevant literature, including some of our own research, on each of these areas of crisis management. Each chapter is organized to illustrate a central claim that we hope to defend persuasively, sometimes defying conventional wisdom and common practice.

Sense vzaking

T h e acute crisis phase seems to pose a straightforward challenge: once a crisis becomes manifest, public leadership must take measures to deal with the consequences. Reality is much more complex, however. Most crises do not materialize with a big bang; they are the product of escalation. Policy makers must recognize from vague, ambivalent, and contradictory signals that something out of the ordinary is developing. T h e critical nature of these developments is not self-evident; policy makers have to "make sense" of them.34

Leaders must appraise the threat and decide what the crisis is about. However penetrating the events that trigger a crisis - jet planes hitting ,bscrapers, thousands of people found dead in mass graves - a uniform picture of the events rarely emerges: do they constitute a tragedy, an outrage, perhaps a punishment, or, inconceivably, a blessing in disguise? Leaders will have to determine how threatening the events are, to what or whom, what their operational and strategic parameters are, and how the situation will develop in the period to come. Signals come from all kinds of sources: some loud, some soft, some accurate, some widely off the mark. But how to tell which is which? How to distill cogent signals from the noise of crisis?

In Chapter 2 we describe and analyze the sense making process in crises. We explain that crises are hard to detect in their early phases. Once they have become manifest, however, it is possible for policy makers and their organizations to construct reliable representations of crisis realities.

Decision rlzaking

Crises leave governments and public agencies with pressing issues to be addressed. These can be of many kinds. T h e needs and problems triggered by the onset of crisis may be so great that the scarce resources available will have to be prioritized. This is much like politics as usual except that in crisis circumstances the disparities between demand and supply of public resources are much bigger, the situation remains un- clear and volatile, and the time to think, consult, and gain acceptance for decisions is highly restricted. Crises force governments and leaders to confront issues they do not face on a daily basis, for example concerning the deployment of the military, the use of lethal force, or the radical restriction of civil liberties.

The classic example of crisis decision making was the Cuba Missile Crisis (1963), during which United States President John F. Kennedy was presented with pictures of Soviet missile installations under con- struction in Cuba. T h e photos conveyed a geostrategic reality in the making that Kennedy considered unacceptable, and it was up to him to decide what to do about it. Whatever his choice from the options presented to him by his advisers - an air strike, an invasion of Cuba, a naval blockade - and however hard it was to predict the exact conse- quences, one thing seemed certain: the final decision would have a momentous impact on Soviet-American relations and possibly on world peace. Crisis decision making is making hard calls, which involve tough value tradeoffs and major political risks.35

12 The Politics of Crisis Management Crisis management in political systems 13

An effective response also requires interagency and intergovernmental coordination. After all, each decision must be implemented by a set of organizations; only when these organizations work together is there a chance that effective implementation will happen. Getting public bur- eaucracies to adapt to crisis circumstances is a daunting - some say impossible - task in itself. Most public organizations have been designed to conduct routine business that answers to values such as fairness, lawfulness, and efficiency. T h e management of crisis, however, requires flexibility, improvisation, redundancy, and the breaking of rules.

Effective crisis responses also require coordination of the many differ- ent groups or agencies involved in the implementation of crisis decisions; these organizations are all under pressure to adapt rapidly and effect- ively. Coordination is pivotal to prevent miscommunication, unneces- sary overlap, and conflicts between agencies and actors involved in crisis operations. Coordination is not a self-evident feature of crisis manage- ment operations. T h e question of who is in charge typically arouses great passions. In disaster studies, the "battle of the Samaritans" is a well- documented phenomenon: agencies representing different technologies of crisis coping find it difficult to align their actions. Moreover, a crisis does not make the public suddenly "forget" the sensitivities and conflicts that governed the daily relations between authorities and others in fairly recent times.

In Chapter 3 we argue that time and again crisis leaders experience how difficult it is to retain control over the course of events. We show that the crisis response is not determined only by crucial leadership decisions but, to a considerable extent, also by the institutional context in which crisis decision making and implementation take place.

Meaning making

A crisis generates a strong demand from citizens to know what is going on and to ascertain what they can do to protect their interests. Author- ities often cannot provide correct information right away. They struggle with the mountains of raw data (reports, rumors, pictures) that quickly amass when something extraordinary happens. Turning them into a coherent picture of the situation is a major challenge by itself. Getting it out to the public in the form of accurate, clear, and actionable infor- mation requires a major communication effort. This effort is often hindered by the aroused state of the audience: people whose lives are deeply affected are anxious if not stressed. Moreover, they do not neces- sarily see the government as their ally. And pre-existing distrust of government does not evaporate in times of crisis.

In a crisis, leaders are expected to reduce uncertainty and provide an account of what is going on, why it is happening, and what

needs to be done. When they have made sense of the events and have arrived at some sort of situational appraisal and made strategic policy

leaders must get others to accept their definition of the situ- ation. They must impute "meaning" to the unfolding crisis in such a way that their efforts to manage it are enhanced. If they don't, or if they do not succeed at it, their decisions will not be understood or respected. If

actors in the crisis succeed in dominating the meaning-making process, the ability of incumbent leaders to decide and maneuver is severely constrained.

T o this end, leaders are challenged to present a compelling story that describes what the crisis is about: what is at stake, what are its causes, what can be done. Whatever one might think about his subsequent policies, there is no disputing that President George W. Bush was effective in framing the meaning of the September 11 attacks to the American public and to the world. This appears all the more true when we compare Bush with his Spanish counterpart Josi Maria Aznar who, after the March 2004 attack in Madrid, hastily tried to pin it down as yet another ETA atrocity. H e failed miserably - a few days after the attack an outraged electorate voted his party out of power.

Leaders are not the only ones trying to frame the crisis. News organizations use many different sources and angles in their frenetic attempts at fact-finding and interpretation. Among this cacophony of voices and sentiments, leaders seek to achieve and maintain some degree of control over the images of the crisis that circulate in the public domain. Their messages coincide and compete with those of other parties, who hold other positions and interests, who are likely to es- pouse various alternative definitions of the situation and advocate differ- ent courses of action. Censoring them is hardly a viable option in a democracy.

In Chapter 4 we examine the meaning-making process. We argue that leadership credibility enhances the quality of the crisis response and increases the chances of political survival in the post-crisis phase. But leaders cannot depend on their credibility. They must excel in crisis communication if they want to reduce the public and political uncertainty that crises cause.

Governments - at least democratic ones - cannot afford to stay in crisis mode for ever. A sense of normalcy will have to return sooner or later. It

14 The Politics of Crisis Management Crisis management in political systems 15

is a critical leadership task to make sure that this happens in a timely and expedient fashion.

Crisis termination is two-fold. It is about shifting back from emer- gency to routine. This requires some form of downsizing of crisis oper- ations. At the strategic level, it also requires rendering account for what has happened and gaining acceptance for this account. These two aspects of crisis termination are distinct, but in practice often closely intertwined. The system of governance - its rules, its organizations, its power-holders - has to be (re)stabilized; it must regain the necessary legitimacy to perform its usual functions. Leaders cannot bring this about by unilateral decree, even if they may possess the formal mandate to initiate and terminate crises in a legal sense (by declaring a state of disaster or by evoking martial law). Formal termination gestures can follow but never lead the mood of a community. Premature closure may even backfire: allegations of underestimation and cover-up are quick to emerge in an opinion climate that is still on edge.

Political accountability is a key institutional practice in the crisis- termination game. T h e burden of proof in accountability discussions lies with leaders: they must establish beyond doubt that they cannot be held responsible for the occurrence or escalation of a crisis. These accountability debates can easily degenerate into "blame games" with a focus on identifying and punishing "culprits" rather than discursive reflection about the full range of causes and ~ o n s e ~ u e n c e s . ~ % e chal- lenge for leaders is to cope with the politics of crisis accountability without resorting to undignified and potentially self-defeating defensive tactics of blame avoidance that serve only to prolong the crisis by transforming it into a political confrontation at knife's edge.

In Chapter 5, we argue that crisis termination depends on the way leaders deal with these accountability processes. We also show that in these accountability processes leaders are at best only partially in control of their political fate, let alone over the evolution of the crisis as a whole.

Learning

A final strategic leadership task in crisis management is political and organizational lesson drawing. The crisis experience offers a reservoir of potential lessons for contingency planning and training for future crises. We would expect all those involved to study these lessons and feed them back into organizational practices, policies, and laws.

Again, reality is a bit messier. In fact, it turns out that lesson drawing is one of the most underdeveloped aspects of crisis management. In add- ition to cognitive and institutional barriers to learning, lesson drawing is

constrained by the role of these lessons in determining the impact that crises have on a society. Crises become part of collective memory, a source of historical analogies for h t u r e leaders. The political depiction of crisis as a product of prevention and foresight failures would force people to rethink the assumptions on which pre-existing policies and rule systems rested. Other stakeholders in the game of crisis-induced lesson-drawing might seize upon the lessons to advocate measures and policy reforms that incumbent leaders reject. Leaders thus have a big stake in steering the lesson-drawing process in the political and bureau- cratic arenas. The crucial challenge here is to achieve a dominant influ- ence on the feedback stream that crises generate into pre-existing policy networks and public organizations.

The documentation of these inhibiting complexities has done nothing to dispel the near-utopian belief in crisis opportunities that is found not only in academic literature but also in popular wisdom. A crisis is seen as a good time to clean up and start anew. Crises then represent discontinu- ities that must be seized upon - a true test of leadership, the experts claim. So most people are not surprised to see sweeping reforms in the wake of crisis: that will never happen again! They intuitively distrust leaders who claim bad luck and point out that their organizations and policies have a great track record.

In Chapter 6, we reject the thesis that prescribes structural reforms in the wake of crisis. In fact, we posit the claim that crisis response, lesson- drawing, and reform craft (the repertoire of skills and strategies that leaders use to make reform work) typically imply orientations and strategies that are fundamentally at odds with each other.

Toward policy advice

At the end of this study, in Chapter 7, we move from our primarily descriptive and interpretive aims and discourse into a more prescriptive mode. We present lessons for crisis leadership conveying the practical implications of our central claims. Together these lessons constitute an agenda for improving public leadership in crises that we hope will reach, inspire, and provoke those who govern us.

Notes

1 See, for instance, Rosenthal, Boin, and Comfort (2001); OECD (2003); Sundelius and Gronvall (2004).

2 See, for instance, Almond, Flanagan, and Mundt (1973); Linz and Stepan (1978); Tilly and Stinchcombe (1997).

16 The Politics of Crisis Management Crisis management in political systems 17

3 T h e core idea of the interdisciplinary subfield of crisis studies is that in a crisis the modus operandi of a political system or community differs markedly from the functioning in normal times. This assumption is, of course, more tenable in highly stable systems (Rosenthal, 1978).

4 A brief period of spectacular growth may be a transition, but it is usually considered a boon rather than a crisis.

5 Rosenthal, Charles, and t' Hart (1989: 10). See also Stem (2003). 6 A threat does not have to materialize before it becomes widely seen as one.

T h e often-cited Thomas Theorem teaches us that it is the perception that makes a threat real in its consequences (Thomas and Thomas, 1928).

7 See Kettl (2004) for an analysis of this period. 8 The use of the term "disaster" usually does presuppose damage, death, and

destruction (see Boin, 2005; Smith, 2005). 9 So-called creeping crises, notably long-term environmental crises such as

desertification and deforestation, soil salination and fertilizer use, global ,warming and the rise of the sea level, constitute a particularly interesting category of complex problems with a high crisis potential.

10 See Flin (1996) and Flin and Arbuthnot (2002) for informed treatises on operational leadership.

11 For an overview of natural disasters see Keys (1999). 12 Recent scenarios feature radical weather changes, biological terrorism, and

asteroid collisions. See Pentagon weather scenarios (Schwartz and Randall, 2003); OECD crisis scenarios (2003); for a clear overview of climate contin- gencies, see Bryson (2003).

13 See Bovens and 't Hart (1996); Quarantelli (1998); Rosenthal (1998); Steinberg (2000). This understanding has become widespread even outside the academic community. See, for instance, the introduction to the Report of the Columbia Accident Investigation Board (2003).

14 See Buchanan (2000). Not all academic fields show this development. It is interesting to note that the long-standing adherence to linear thinking in the international relations (IR) field correlates with a long history of failed early- warning systems (Jervis, 1997; Bernstein et al., 2000).

15 Turner and Pidgeon (1997). 16 Reason (1990). 17 Perrow (1984). 18 See Turner (1978) and Perrow (1 994). 19 For an excellent introduction, see Eichengreen (2002). 20 Useem and Kimball (1989). 21 This example is taken from Boin and Rattray (2004). For an application of

Perrow's theory to a stadium crowd disaster, see Jacobs and 't Hart (1992). 22 Similar dynamics of destructive escalation have marked the incubation

phases of corporate and organizational crises. Examples in 2004-5 included the troubles at Shell and the BBC.

23 For a classic statement, see Smelser (1962). More recent contributions include Waddington (1992) and Goldstone and Useem (1999).

24 The laws of complex systems are still largely unknown. And the more we learn about the behavior of complex systems, the less we seem to understand. Complexity theorists are busy uncovering the hidden patterns that they say

underlie this process, but practical insights (for our purposes at least) have yet to emerge. For an introduction see Buchanan (2000).

2 4 See f.i. Rijpma (1997) who argues that redundancy - an often-prescribed tool to help prevent incidents -may actually help cause them.

26 The internet is a great resource for tracing the many systems that are now in operation, partly run by non-government organizations (NGOs), partly by specialized academics, partly by commercial organizations catering to busi- ness investor audiences. See www.reliefweb.int~resources/ewam.html, which is a pivotal portal in this regard.

2 , One may, of course, ask whether some of the crisis responses in Western democracies - counterterrorism policies in the 1970s and in the wake of the 911 1 events come to mind - do not amount to what Juan Linz ( I 978) once called "an abdication of democratic authenticity" (e.g. an expansion of feasibility boundaries at the price of sacrificing values of democratic rule).

28 Beck (1992). 29 Pioneering works in this tradition include Barnard (1938), Selznick (1957),

and Wilson (1 989). 30 The dangers of a functional approach in studying society, politics, and their

crises have been discussed extensively. Binder et al. (1971) and Almond, Flanagan, Mundt (1973) are key examples of the tradition. For a recent revivalist interpretation of functionalism, see Wilson (2002).

31 Developed further in 't Hart (1993). See also Habermas (1975); Edelman (1977); Linz and Stepan (1978); Turner (1978).

32 There is a large field of leadership studies in which the relation between personal characteristics and task fulfillment receives ample attention. A classic account is MacGregor Bums (1978). A good introduction to the entire literature is Northouse (200 1).

33 For strategic leadership issues in risk assessment and crisis prevention, see among others Wildavsky (1988); Wildavsky (1995); and Meltsner (1990).

34 A classic point made by Edelman (1977). 35 Brecher (1980); Janis (1989). 36 Although much more pronounced today, the tendency to search for culprits

following the occurrence of disaster and crisis is age-old - see Drabek and Quarantelli (1967) as well as Douglas (1992).