Factors influencing the performance on crisis management tasks in dealing with escalating social media events

(1)

Institute of Security and Global Affairs

Faculty of Governance and Global Affairs – Leiden University

FACTORS INFLUENCING THE PERFORMANCE ON CRISIS

MANAGEMENT TASKS IN DEALING WITH ESCALATING

SOCIAL MEDIA EVENTS

Michel Soomers

A thesis submitted to the Faculty of Governance and Global affairs at Leiden University

in partial fulfillment of the requirements for the degree of Master of Science in the

Institute of Security and Global Affairs (Crisis and Security Management).

Student number:

s1233580

Thesis course coordinator: Dr. J. Matthys

Thesis supervisor:

Dr. S.L. Kuipers

(2)

(3)

Abstract

This thesis is a study into factors that influence the performance on crisis management tasks by crisis response leaders in the escalation of social media events into crises. To answer this question, a theoretical framework of factors influencing the performance on crisis management tasks is constructed and then applied to a number of ‘Project X’-events in a comparative case study. For the purpose of this study, crisis management is divided into three tasks relevant to the escalation of social media events: sense making, meaning making, and decision making.

The main findings are as follows. Factors influencing sense making are: whether an organization is looking for trouble, the sharing of information, and the perception of the threat. Factors influencing decision making and coordination are: experience in working together, internal conflict, good relationships, leadership, mobilization, upscaling procedures, and effective communication. Factors influencing meaning making are: preparedness for meaning making, coordination of communication, and credibility. Three different meaning making strategies: framing, rituals, and masking, have all been used by the actors in these ‘Project X’-events.

(4)

Acknowledgements

I would like to express my gratitude to Dr. Kuipers for her immense knowledge, excellent supervision, motivation, inexhaustible patience, and supportive attitude. I would also like to thank Dr. Landucci, Dr. Matthys, Ms. Bondarouk MSc, Ms. Park, the respondents that I have interviewed, and mr. dr. Keurentjes for their contributions to this thesis.

I would like to thank my parents, my brother, Laurens, my grandmother, and my family and friends for their support during this long and challenging process. I could not have achieved this without you.

“without your past, you could never have arrived- so wondrously and brutally,

By design or some violent exquisite happenstance ...here.” (Taylor Swift, 2017)

(5)

Ch. 1: Introduction

1.1 Problem outline

When one reads about 34 arrests, 36 citizens wounded, 15 police men wounded and almost a million euro in damages (Van Eck, 2017), one would expect these numbers to be from a news report about riots in a major city following political turmoil or a big soccer match. However, these numbers are actually related to the cancelled celebration of a sixteen-year old’s birthday – and the riots that followed. On September 21, 2012 about 3000-5000 youngsters gathered in the small-town of Haren to ‘celebrate’ a party that never took place. The occasion? A mistake by the birthday girl while setting up an event on Facebook.

On September 7, Merthe, a girl residing in the town of Haren, sent out an invite to 78 people to celebrate her upcoming birthday with her. She published the invite as ‘public’ on Facebook, so her friends could invite their friends. Immediately people started sharing the invitation and within a day 3500 people were attending the Facebook event and another 16000 people were invited (Commissie 'Project X' Haren (1), 2013). The event was deleted, but it was already too late. People copied the event and discussed it on social media, traditional media and news outlets picked it up, and in the end the event seemed unstoppable as a large number of people came to Haren, resulting in riots. More ‘Project X’-events followed, to name a few: Schiedam (Algemeen Dagblad, 2012, September 23), Arnhem (Van Gelder, 2017, November 22), and Enschede (Meesterburrie, 2017, October 14). The events in Schiedam, and Arnhem did not lead to a crisis. But then, only a year ago, the event in Enschede led to vandalism at a gas station and a number of arrests (Kerstens, 2016, September 21). This observation – similar events, different outcomes - forms a departure point for this research.

The event in Haren was deemed unstoppable. But was it really? Why did the events in other cities not lead to crises? How does a social media event lead to a crisis? Who is responsible for de-escalation? The mayor is responsible for maintaining public order and safety in their municipality (Nederlands Genootschap van Burgemeesters, 2018). However, the mayor is not the sole actor performing this task. The mayor leads other actors from the municipality and the police who are also involved. Therefore, it would be interesting to see what factors influence the performance on crisis management tasks by the responsible authorities, led by the mayor, in the escalation of social media events into crises. It could help us identify important signals, tasks, and actions in the lead-up to potential crises.

1.2 Research question

The problem outlined above calls for a number of questions: What was the role of the mayor? What coordination mechanisms are in place? Which actions have been taken and were they effective? These questions are interesting in itself, but above all raises the question how it is possible that similar situations had dramatically different outcomes? This leads to the following research question, that will be answered in this thesis:

(8)

What factors influence the performance on crisis management tasks by crisis response leaders in the escalation of social media events into crises?

1.3 Academic relevance

Even though a lot has been written about the events in Haren, this will be the first systematic analysis of the behavior of mayors and other crisis leaders during ‘Project X’-situations in the Netherlands using the analytical framework of crisis management tasks by Boin, ‘t Hart, Stern and Sundelius (2005). The framework is much applied and often referenced to but needs further research to test and extend its utility (Jong, Dückers, & Van der Velden, 2016). This research aims to find to what extent this theory fits the performance of mayors and other crisis leaders on crisis management tasks and whether it is able to explain the different outcomes of the similar potential crisis situations that have been selected as cases. Additionally, the subject of this thesis is closely related to the mobilizing power of social media in broader terms. Not much research has been done yet into this relatively new phenomenon.

1.4 Societal relevance

The riots in Haren led to arrests, wounded citizens and police men, and almost a million euro in damages. Additionally, the sense of security of many residents was affected in a negative way. It is safe to say that ‘Project X Haren’ had a major impact on Dutch society and therefore research into this event, but also into the broader phenomenon of these ‘Project X’-situations is highly relevant to society. Especially since it is a contemporary issue that seems to be persisting over time, with many cases between 2012 and now. More insight in how these crises or near-crises should be managed is highly valuable to crisis leaders, policy makers and other people in the executive branch of government.

1.5 Reading guide

In this first chapter, I have introduced the subject and stated the research question that will be answered in this thesis. I have argued why it is relevant to conduct this research, both for academia and society. In the second chapter, I will describe the theory of crisis management tasks by Boin, ‘t Hart, Stern and Sundelius (2005). Sense making, decision making and meaning making are important concepts from this theory that will be explained in this chapter. I will make use of these concepts to analyze the various ‘Project X’-situations in the Netherlands. In the third chapter, the research method will be explained and justified. A brief overview of the cases is presented in the same chapter. In the fourth chapter, I will present the data and analyze it by applying the theoretical framework from the second chapter. In the fifth and final chapter I will draw conclusions from this analysis and answer the research question. I will also reflect on the research and discuss possibilities for further research into this subject.

(9)

Ch. 2: Concepts and theoretical framework

In this chapter, the concepts and theoretical framework that are used in this thesis will be discussed. The analytical framework of crisis management tasks by Boin, ‘t Hart, Stern and Sundelius (2005) will be described and combined with other relevant literature to gain a deep understanding of the performance on crisis management tasks by crisis response leaders.

2.1 Relation of the theory to the existing body of knowledge

The framework by Boin, ‘t Hart, Stern and Sundelius (2005) has been much applied and referenced elsewhere in the field of crisis management studies (Jong, Dückers, & Van der Velden, 2016, p. 47). According to a literature review on crisis management tasks by Jong, Dückers and Van der Velden (2016, p. 55) the key crisis management tasks have only been examined briefly in the literature up until now. The authors specifically call for “more studies that enable a comparison of mayors and governors in similar crisis situations” (Jong, Dückers, & Van der Velden, 2016, p. 56). Research that tests and may extend the utility of the framework by Boin, ‘t Hart, Stern and Sundelius (2005) when it comes to mayors and their counterparts is needed (Jong, Dückers, & Van der Velden, 2016, p. 56).

2.2 Crisis management tasks in crisis studies

Boin, ‘t Hart, Stern and Sundelius describe the political challenges and realities of leadership in times of crises. They do so by formulating five core tasks of crisis leadership: sense making, decision making, meaning making, terminating, and learning (Boin, 't Hart, Stern, & Sundelius, 2005, p. ix). The authors argue that policy makers can limit the damage by responding well to a crisis, but they can also increase the impact of a crisis when they fail (Boin, 't Hart, Stern, & Sundelius, 2005, p. 1). In their study, the authors are interested in how the performance on crisis management tasks relates to the crisis outcome (Boin, 't Hart, Stern, & Sundelius, 2005, p. 10). Existing research in this field has identified a number of important crisis management tasks, but not yet led to a theory that can fully explain the outcome of a crisis as a result of actions taken by crisis managers (Boin, Kuipers, & Overdijk, 2013, p. 80). The outcome of a crisis is the combined result of many factors and the behavior of many other actors, but research suggests that crisis management can make an impact on the outcome of a crisis (Boin, Kuipers, & Overdijk, 2013, p. 81).

First, it is important to define the concepts of ‘crisis’ and ‘crisis management’ for our understanding of the theory and for further use in this thesis. Boin, ‘t Hart, Stern and Sundelius (2005) define a crisis as a situation wherein policy makers “experience a serious threat to the basic structures or the fundamental values and norms of a system, which under time pressure and highly uncertain circumstances necessitates making vital decisions” (Boin, 't Hart, Stern, & Sundelius, 2005, p. 2). To analyze crisis management, the authors look at what leadership in crises entails (Boin, 't Hart, Stern, & Sundelius, 2005, p. 9). They look at the strategic tasks that circumscribe all activities associated with the different stages of crisis management. The authors do not state that crisis management tasks are exclusively reserved for leaders, but they do argue that the formal leaders carry a “special responsibility for making

(10)

sure that these tasks (...) are properly addressed and executed” (Boin, 't Hart, Stern, & Sundelius, 2005, p. 9).

According to Boin, ‘t Hart, Stern and Sundelius (2005) crisis leadership involves five critical tasks: sense making, decision making, meaning making, terminating, and learning. In this thesis, social media events that developed into a crisis and social media events that did not develop into a crisis will be studied. This focus means that the terminating and learning tasks will not be a part of the analysis. Therefore, this research focuses on the three tasks important in the lead-up to and early stages of a crisis: sense making, decision making, and meaning making. These tasks will be described in the next paragraphs.

2.2.1 Sense making

The first critical task identified by Boin, ‘t Hart, Stern and Sundelius is sense making (2005, p. 10). Sense making is the process of recognizing that something extraordinary is developing (Boin, 't Hart, Stern, & Sundelius, 2005, pp. 10-11). Leaders have to do a threat assessment and consider their operational and strategic options. They also have to assess how they think the situation will develop over time. In this phase, leaders receive many different signals from many different sources. The central question is: How can they make sense of all this information? According to Boin, Kuipers and Overdijk effective sense making requires “a well-rehearsed method to process information, share it with the right people and consider their feedback, create a dynamic picture that everybody understands, analyze possible “futures” and potential consequences, and formulate specific information needs (which somehow must be met)” (2013, p. 83). Academics in the field of crisis management identify sense making as an important crisis management task (Boin, Kuipers, & Overdijk, 2013, p. 83; Jong, Dückers, & Van der Velden, 2016, p. 54).

Boin, ‘t Hart, Stern and Sundelius (2005, p. 19) make one core claim that consists of two major points. First, they argue that it is more or less impossible to predict a crisis. Second, the authors argue that it is possible to grasp the dynamics of a crisis once it occurs, but that it is also easy to make mistakes. They develop this argument by describing different barriers to crisis recognition. They do this on an organizational and on an individual level. Both individuals and organizations have to acquire and process enormous amounts of information once the intelligence and expert resources of a government have been appointed to a crisis (Boin, 't Hart, Stern, & Sundelius, 2005, p. 30).

2.2.1.1 Organizational barriers to crisis recognition

The authors identify a number of organizational barriers to crisis recognition (Boin, 't Hart, Stern, & Sundelius, 2005, p. 19). Their research findings show that many organizations lack expertise to detect and cope with safety and security threats. The system in which people operate often does not allow people to think “out of the box”, yet this is needed to detect a crisis that is about to happen. There are four main reasons why organizations often fail to recognize impending crises:

(11)

1. Many organizations are not designed to look for trouble;

2. Variable disjunction of information and the politics of organization; 3. Cognitive blinders and the perversity of intelligent design;

4. Organizational barriers to reality-testing (Boin, ‘t Hart, Stern & Sundelius, 2005, pp. 19-35). The first reason can be explained by the fact that many organizations were never designed to detect potential future crises (Boin, 't Hart, Stern, & Sundelius, 2005, p. 20). These organizations have a tendency to focus on achievements, instead of avoidance. This affects their capability to detect crises. To detect crises, these organizations would need to collect data on events that may develop into a crises, but most organizations simply do not do this (Boin, 't Hart, Stern, & Sundelius, 2005, p. 21). There is no room for intuition or “gut feeling”.

The second reason can be explained by the fact that policy makers are not able to put together the information that can warn them for an impending crisis, even when this information is available somewhere within the organization (Boin, 't Hart, Stern, & Sundelius, 2005, p. 21). Research shows that these organizations often fail to turn available data into useful information, because the interpretation of data often is part of a political process and because mechanisms that would facilitate swift sense making are not present (Boin, 't Hart, Stern, & Sundelius, 2005, pp. 22-23). Because this reason is closely linked to the way in which organizations look for trouble, and the characteristics of threat perception, this reason will not be separately examined as a factor.

The third reason can be explained by the fact that some organizations suffer from their rationality and efficiency (Boin, 't Hart, Stern, & Sundelius, 2005, pp. 23-24). This allows mistakes to travel “fast” through an organization. Even though these organizations often have warning mechanisms for crises in place, they often are not able to detect these because of two factors. First, this situation can lead to a “normalization of risk” (Boin, 't Hart, Stern, & Sundelius, 2005, p. 24). Risks that are deemed small become neglected. Second, by focusing on detecting particular crises, unforeseen and unimagined crises will not be noticed (Boin, 't Hart, Stern, & Sundelius, 2005, p. 25).

The fourth reason includes a number of organizational constraints to reality-testing. For people within an organization, sharing information with other people or organizations is not necessarily their first concern. They may perceive the world differently than others, have different viewpoints, interests, priorities, or look at information as a valuable resource that can be used as a leverage or to gain power (Boin, 't Hart, Stern, & Sundelius, 2005, p. 34). This behavior is not always a sign of a dysfunctional organization. Sometimes, it is the result of intentional (security) policies that prevent people from sharing information.

2.2.1.2 Individual barriers that hinder effective sense making

Boin, ‘t Hart, Stern and Sundelius (2005) identify a number of barriers to crisis recognition on the individual level. The authors distinguish three main barriers:

(12)

1. Because of the characteristics of threat perception, a potential threat may fail to get recognized as such;

2. Stress can seriously influence the performance of decision makers;

3. There are individual constraints that hinder effective sense making (Boin, ‘t Hart, Stern & Sundelius, 2005, pp. 25-34).

In order for a crisis to be recognized as such, a considerable number of players must agree that a threat exists and that it is an urgent matter (Boin, 't Hart, Stern, & Sundelius, 2005, p. 25). Governmental leaders have to define a situation as a crisis in order for action to be taken. Due to time constraints, short-term considerations can keep long-short-term threats from the policy agenda. Sometimes, ‘claim makers’ manage to influence this political process. Because of these characteristics of threat perception, sometimes a potential threat may fail to get recognized as such in the sense making process. Boin, ‘t Hart, Stern and Sundelius (2005, p. 28) argue that “warnings of impending crises stand a chance only if they are understood as direct threats to the way of living that is protected and serviced by the political-administrative elites”.

Stress can seriously influence the performance of decision makers (Boin, 't Hart, Stern, & Sundelius, 2005, p. 29). The amount of stress seems to be related to the amount of experience a crisis leader has. Experienced actors are usually more effective at maintaining performance under pressure than newcomers (Boin, 't Hart, Stern, & Sundelius, 2005, p. 29). Under heavy stress, individuals are thought to: focus on the short term, fall back on and cling to old patterns, narrow and deepen their span of attention, be more likely to rely on stereotypes and be more easily irritable (Boin, 't Hart, Stern, & Sundelius, 2005, p. 30). These effects of stress on an individual can hinder effective sense making. There are many individual constraints that hinder effective sense making. First, humans make sense of information by combining what they process with past experience and their expectations. This can lead policy makers in the right direction, but it can also lead to misinterpretation of information (Boin, 't Hart, Stern, & Sundelius, 2005, p. 31). Second, there are simply fundamental limitations to the amount of information a human can acquire and process (Boin, 't Hart, Stern, & Sundelius, 2005, p. 31). Third, humans are biased in their assessment of situations (Boin, 't Hart, Stern, & Sundelius, 2005, p. 32). Fourth, under high levels of stress humans resort to cognitive, affiliative, and egocentric rules of thumb (Boin, 't Hart, Stern, & Sundelius, 2005, p. 33). Cognitive rules of thumb allow humans to make quick choices but are often dangerously simplistic. Affiliative rules amount to sense making processes where policy makers seek solutions that will not have a negative impact on their relationships with other people within the organization. Egocentric rules include rules that are self-serving and aim to satisfy personal motives and rules that are emotive and aim to satisfy emotional needs. During urgent threats, affiliative and egocentric rules of thumb become highly relevant to sense making (Boin, 't Hart, Stern, & Sundelius, 2005, p. 33).

(13)

2.2.2 Decision making and coordination

The second critical task is decision making. In times of crises governments and leaders have to address pressing issues. They have to make decisions on how to act and what resources to prioritize. This all happens under exceptional circumstances under high pressure and with highly restricted time (Boin, 't Hart, Stern, & Sundelius, 2005, pp. 11-12). Leaders have to make hard calls, involving difficult tradeoffs and potentially huge political risks. Studies have found that effective decision making is hindered by the complexity of networks, underlining the importance of taking crisis coordination politics into account when assessing the performance of government actors on the decision-making task (Jong, Dückers, & Van der Velden, 2016, p. 54). Therefore, crisis coordination is also included in this crisis management task. Boin, ‘t Hart, Stern and Sundelius (2005) look not only at the leadership decisions, but also at the institutional context. In ‘Executive Politics in Times of Crisis’ Boin and ‘t Hart (2012) provide an additional framework for analyzing crisis coordination that will be elaborated on later in this chapter. Boin, ‘t Hart, Stern and Sundelius state that many studies of crisis management find an “upward” shift in decision making. Decisionmakers adjust their response to a crisis to the scale of the crisis. The responsibility for coordinating government responses will shift to a higher level of government. This may happen in the spur of the moment, but often it is prescribed in contingency plans (2005, p. 42).

2.2.2.1 Decision making

Public leaders are not the only ones who make decisions during a crisis. Other actors include government officials and people outside government. Additionally, crisis decision-making is not the only factor that determines if a crisis response will become successful. The implementation of those decisions is also highly important. The implementation often involves many different organizations and groups (2005, p. 43). The authors argue that executive-level crisis decision making is an important answer to the question why some crisis response operations are successful, and others fail, but it is not the only answer. They claim that “crisis management depends not so much on critical decision making but on the facilitation of crisis implementation and coordination throughout the response network” (2005, p. 43).

During a crisis, leaders have to make choices that are highly consequential, contain genuine dilemmas, present leaders with major uncertainties – all under great time pressure. This combination of characteristics makes decision-making in crisis situations difficult. Boin, ‘t Hart, Stern and Sundelius state that for a successful decision leaders have to “discount the uncertainties, overcome any anxieties that they may feel, control their impulses, and commit the government’s resources to a course of action that they can only hope is both effective and appropriate in the political context they are in” (2005, p. 44). According to the authors poor decision-making can often be explained by stress. They argue that this is cause for special concern (2005, p. 45).

Crisis decision-making usually takes place in an institutionally sanctioned small-group setting, guided by crisis contingency plans that include requirements for crisis teams to gather and start coordinating the crisis response. These crisis teams become critical nodes in highly complex crisis response networks

(14)

(Boin, 't Hart, Stern, & Sundelius, 2005, p. 45). This has both advantages and disadvantages. A potential advantage is the increased intellectual and cognitive capacity of a group, but group dynamics can also be a disadvantage. According to Boin, ‘t Hart, Stern and Sundelius (2005, p. 45) individuals in groups often do not share and use information effectively, because of a number of factors. These factors include two forms of group behavior: conflict and conformity. In addition to that, groups can fall apart due to pressure. Even loyalty can become a problem when members of a group deem good relations and unity within the group more important than the actual decisions. These problems tend to be enhanced by the high degree of informality that often occurs within crisis response groups. This leads to the disappearance of procedural rules and safeguards that usually stabilize policy making (Boin, 't Hart, Stern, & Sundelius, 2005, p. 46).

Leaders are not always in control of the crisis response (Boin, 't Hart, Stern, & Sundelius, 2005, p. 51). Often, many decisions are not taken by leaders but emerge during the process of crisis decision making and coordination. Boin, ‘t Hart, Stern and Sundelius identify three situations where this happens. The first is non-decision making. Decisions that are not taken determine the course events just as much as intentional choices (Boin, 't Hart, Stern, & Sundelius, 2005, p. 51). The authors distinguish decisions that are not taken, decisions not to decide, decisions not to act, and strategic evasion of choice opportunities. Non-decision making can have both good and bad consequences. According to Boin, ‘t Hart, Stern and Sundelius it “may provide a recipe for chaos, but it may equally well mean a refreshing absence of the kind of political interference in operational affairs so often detested by agency leaders” (2005, p. 52). The second situation where leaders are not in control of the crisis response is decentralization (Boin, 't Hart, Stern, & Sundelius, 2005, p. 53). Centralization can be a potential liability in times of crisis, because the actual crisis is different from the one that is foreseen in the formal crisis contingency plans (Boin, 't Hart, Stern, & Sundelius, 2005, p. 54). Centralization may produce policies that do not reflect what is happening “on the ground”. Boin, ‘t Hart, Stern and Sundelius argue that crisis experience tends to favor decentralization of crisis response authority (2005, p. 54). The third situation where leaders are not in control of the crisis response is when improvisation occurs (Boin, 't Hart, Stern, & Sundelius, 2005, p. 55). Due to the nature of crisis, plans and structures often become irrelevant. The situation may require improvisation to adjust the response to what is actually happening at that moment. Improvisation can have both good and bad effects on the crisis performance. In a best-case scenario, structure provides the crisis responders with something to fall back on, while allowing the exploitation of improvisation (Boin, 't Hart, Stern, & Sundelius, 2005, p. 55).

2.2.2.1.1 Vulnerabilities

Boin, ‘t Hart, Stern and Sundelius (2005, p. 46) mention a number of vulnerabilities of crisis teams. The first vulnerability of a group is newness and conformity. The people that gather during a crisis will not always be familiar to one another or have experience in working together as a group. They are likely to conform to the leaders of the group and to avoid conflict resulting in ill-considered decision-making. Leaders can counteract this by intervening actively and encouraging participation (Boin, 't Hart, Stern,

(15)

& Sundelius, 2005, p. 47). The second vulnerability is excessive cordiality and conformity. Crisis teams can become safe places for the individuals within a crisis group, guarding them from the pressures and dilemmas of the crisis. This can reduce stress, but also lead to a reduce in the capacity for reality-testing (Boin, 't Hart, Stern, & Sundelius, 2005, p. 47). People in a crisis group can become isolated from the actors in the outside world. Whether this happens has to do with the relations prior to the crisis. Boin, ‘t Hart, Stern and Sundelius argue that the accountability structure is an important factor. If group members feel that they are not accountable for the situation or will be held accountable as a group, they are more likely to engage in conformity behavior (2005, p. 48). The third vulnerability is centrifugality and politicking (Boin, 't Hart, Stern, & Sundelius, 2005, p. 48). As opposed to conformity, crisis may also produce internal conflicts within a group. When internal conflicts occur, group members can use their information and expertise for personal gains rather than to help the group as a whole. This happens in the form of rumors, leaks, silences, and misrepresentations or as attempts to form or break up sub-groups within the group (Boin, 't Hart, Stern, & Sundelius, 2005, p. 48).

2.2.2.1.2 Success factors

Crisis teams do not only have vulnerabilities. They can also work well when a number of conditions are met. Boin, ‘t Hart, Stern and Sundelius (2005, p. 49) state a number of success factors for effective crisis decision making. The first success factor is experience in working together (Boin, 't Hart, Stern, & Sundelius, 2005, p. 49). Crisis teams are more likely to perform effectively in areas or organizations where crises happen often and are reoccurring. The second success factor is good relations and mutual trust prior to the crisis between the people in a crisis group (Boin, 't Hart, Stern, & Sundelius, 2005, p. 49). The people within a group need to have a shared understanding of the purposes of the group and their roles in it. It must be clear who leads and who advises. Procedures can help a group to create conditions for optimal information sharing, collaboration, and frank discussion (Boin, 't Hart, Stern, & Sundelius, 2005, p. 50). Multiple advocacy is a well-known procedure that can be employed to stimulate this. Different actors within the group get a chance to argue their case before a leader. The process is guided by a neutral manager. This procedure is very appropriate for the coordination of crisis units (Boin, 't Hart, Stern, & Sundelius, 2005, p. 50). However, the authors note that due to the nature of crisis situations this procedure is often difficult to adopt in times of crisis.

2.2.2.2 Crisis coordination

Leaders depend on organizations to execute their decisions (Boin, 't Hart, Stern, & Sundelius, 2005, p. 56). Most crises involve a multitude or organizations, so leaders have to coordinate the crisis response between these organizations. This is both an important and difficult crisis management task (Boin, 't Hart, Stern, & Sundelius, 2005, p. 56). Boin and ‘t Hart (2012, p. 181) discuss crisis coordination as a separate crisis management task, but the problems of crisis coordination and the conditions for effective crisis coordination described by the authors in this book are very similar to the ones described by Boin, ‘t Hart, Stern and Sundelius (2005, pp. 57-63). This supports the findings by Boin, ‘t Hart, Stern and Sundelius (2005, pp. 57-63) which are in turn used in this study.

(16)

2.2.2.2.1 Problems of crisis coordination

Boin, ‘t Hart, Stern and Sundelius identify a number of problems in coordinating a crisis response. The first problem undermining the coordination of crisis management efforts is the “limited adequacy of the planned response” (Boin, 't Hart, Stern, & Sundelius, 2005, p. 57). This means that organizations that have to deal with a crisis, experience problems in doing what they are supposed to do because they face problems with the mobilization of the members (Boin, 't Hart, Stern, & Sundelius, 2005, p. 57). A second problem undermining the coordination of crisis management efforts is the “matching of place and function” (Boin, 't Hart, Stern, & Sundelius, 2005, p. 57). Crises challenge the organizational designs of involved organizations. Problems that arise during a crisis may require the organizations to perform tasks or come up with solutions that they were not or only partly designed to perform, sometimes outside of their jurisdiction or geographical area (Boin, 't Hart, Stern, & Sundelius, 2005, p. 57). The third problem undermining the coordination of crisis management efforts is the “upscaling dilemma” (Boin, 't Hart, Stern, & Sundelius, 2005, p. 58). It is not always clear when central authorities should step in or who decides when this needs to happen. Additionally, upscaling can increase the gap between decision-makers and the people on the ground. A fourth problem is found in the interorganizational relations within the crisis response network (Boin, 't Hart, Stern, & Sundelius, 2005, p. 58). Cooperation between parties can be hindered by communication problems or a different understanding of the meaning of coordination (Boin, 't Hart, Stern, & Sundelius, 2005, p. 59). When a crisis becomes more complex, this problem often increases. A fifth problem is the unwillingness of some organizations to cooperate with others (Boin, 't Hart, Stern, & Sundelius, 2005, p. 59). Pre-existing tensions often persist or even intensify during a crisis. Friction increases when different crisis responders that are not used to working together are forced by a crisis to cooperate. Because of these interorganizational tensions the central organizations coordinating the crisis have to reach compromises that are not always the most effective decisions.

2.2.2.2.2 Effective crisis coordination

Boin, ‘t Hart, Stern and Sundelius (2005, p. 60-61) do not only identify problems in crisis coordination. They also identify a number of factors that can foster effective crisis coordination. Successful crisis coordination emerges almost as a natural by-product of crisis, instead of being imposed by leaders (Boin, 't Hart, Stern, & Sundelius, 2005, pp. 60-61). Crisis coordination can be a by-product of the search for information and the adjustments that actors make in that search. People who possess information interact with people who seek information. In order to exchange this information successfully, the exchange is institutionalized. Crisis leaders can nurture the right conditions for this. An important condition for successful crisis coordination is the formalization of the coordination process (Boin, 't Hart, Stern, & Sundelius, 2005, p. 63). The coordination process requires rules that facilitate the interaction between the different crisis responders and structure the flows of information. Without these rules, the coordination structure may collapse.

(17)

2.2.3 Meaning making

The third critical task is meaning making. Meaning making is the process of reducing uncertainty and provide the public with answers to the questions of what is going on, why it is happening, and what needs to be done (Boin, 't Hart, Stern, & Sundelius, 2005, p. 13). Studies show that meaning making is considered a particularly important crisis management task for mayors and governors (Jong, Dückers, & Van der Velden, 2016, p. 54). This is an important crisis management task, because effective crisis communication can make the crucial difference between obtaining and losing the support that leaders need during a crisis to implement their decisions (Boin, 't Hart, Stern, & Sundelius, 2005, p. 70). Much has been written about crisis communication, but Boin, ‘t Hart, Stern and Sundelius (2005, p. 70) is particularly relevant to this study as they focus on crisis communication in a political setting.

Boin, ‘t Hart, Stern and Sundelius (2005, p. 13) explain that leaders must give meaning to the crisis in order to enhance their efforts to manage it. They can do this by telling a convincing story. In giving meaning to a crisis, leaders and other stakeholders attempt to reduce the uncertainty and confusion that is caused by crises. According to Boin, ‘t Hart, Stern and Sundelius (Boin, 't Hart, Stern, & Sundelius, 2005, p. 69) they do so by communicating a compelling narrative that “explains what happened, why it had to be that way, what its repercussions are, how it can be resolved, who can be relied upon to do so, and who is to blame” (Boin, 't Hart, Stern, & Sundelius, 2005, pp. 69-70). They have to do this while competing for attention with other stakeholders that may have other interests.

In the political context of crisis management, the capacity to capture public attention and a reputation for accuracy and trustworthiness are essential (Boin, 't Hart, Stern, & Sundelius, 2005, p. 71). Credibility is especially important, because without credibility “even the most basic tasks become difficult and subject to scrutiny by the media and other watchdogs” (Boin, 't Hart, Stern, & Sundelius, 2005, p. 71). Boin, ‘t Hart, Stern and Sundelius describe the context in which policy makers communicate as a triangular relationship between political actors, the mass media, and the citizenry (2005, p. 72). Each of these has their own behavior, role, and interests.

The mass media play an important role in discovering, conveying and (de)escalating crises (Boin, 't Hart, Stern, & Sundelius, 2005, p. 72). There is much debate about the distribution of power among the media and the government. Some emphasize the possibilities for crisis managers to use the media to take control of the agenda and to lead them in desirable directions. Others emphasize the power of the media in shaping public opinion. These are both extreme viewpoints, whereas in practice the tone of the mass media’s crisis reporting can vary widely. The character of crisis reporting is influenced by political ideology, editorial policy, organizational factors, and news organizations’ crisis response capacities (Boin, 't Hart, Stern, & Sundelius, 2005, pp. 74-75).

All people react differently to crisis situations. During a crisis, citizens can play various roles. They can be spectators, victims, or witnesses. Because of this, public authorities should not base their crisis

(18)

responses and communication strategies on stereotypes and myths about the behavior of citizens in times of crisis (Boin, 't Hart, Stern, & Sundelius, 2005, p. 75). Boin, ‘t Hart, Stern and Sundelius argue that even though the average citizen may not be an expert on the crisis at hand, authorities and journalists cannot mislead the public for a long period of time and get away with it (2005, p. 75). During a crisis, the public is more interested in news than usually. As Halper (1971) explain in Boin, ‘t Hart, Stern and Sundelius (2005, p. 76) this can make citizens more willing to receive information from a legitimate government, but it can also make the public more alert to flaws in the story.

During a crisis, government actors seek to direct or influence the behavior and opinions of the public. To reach the public, the government has to work together with the mass media. This can lead to difficulties. Moreover, many government leaders and organizations are not prepared for crisis communication.

2.2.3.1 Important factors in determining the effectiveness of governmental crisis communications efforts

Boin, ‘t Hart, Stern and Sundelius identify three factors for assessing the effectiveness of governmental crisis communications efforts. The first important factor is the degree of preparedness (Boin, 't Hart, Stern, & Sundelius, 2005, p. 76). A lack of preparedness hinders effective crisis communication because it slows down the response in the first stages of a crisis. Instead of shaping the story, authorities are catching up on it. Due to a lack of organization, the officials who perform the crisis communication may not possess the skills to do this effectively. They may, for example, use very technocratic language that the public does not understand and can misinterpret (Boin, 't Hart, Stern, & Sundelius, 2005, pp. 76-77). An effective way to combat this is by proactively monitoring and engaging in rumor-control, but research shows that these efforts are often lacking, especially in ill-prepared organizations. The second factor for effective crisis communication is the degree of coordination of outgoing information (Boin, 't Hart, Stern, & Sundelius, 2005, p. 77). Internal division between all the involved organizations should be avoided and the message to the outside world should be coordinated. However, this does not mean that the government should engage in government communication that is intentionally misleading the public. The third factor for effective crisis communication is the degree of professionalization (Boin, 't Hart, Stern, & Sundelius, 2005, p. 77). A professional approach to dealing with the mass media is necessary, but at the same time the policy makers have an interest to keep the media at a distance or even to manipulate them. If the professional spokespersons do not get this balance right, they risk getting trapped between the policy makers that they serve and the mass media (Boin, 't Hart, Stern, & Sundelius, 2005, p. 77). Boin, ‘t Hart, Stern and Sundelius (2005, p. 78) argue that “PR officials often lack the authority that others derive from their badge or their technical expertise”.

2.2.3.1. Credibility

The degree of credibility that public actors possess is the most important factor in determining the effectiveness of government crisis communication (Boin, 't Hart, Stern, & Sundelius, 2005, p. 78). To influence public opinion and to shape the political meaning of a crisis, the public has to trust the crisis response leaders and deem them credible. Distrust will lead to disbelieve and may cause the public to

(19)

refuse to act in the desired way. Therefore, pre-existing trust issues or credibility deficits or those which emerge in the early stages of a crisis make managing the crisis much harder (Boin, 't Hart, Stern, & Sundelius, 2005, p. 78). Sometimes, leaders are tempted to play down the seriousness of the situation, but this can easily backfire.

2.2.3.2.1 Sources of credibility

Boin, ‘t Hart, Stern and Sundelius (2005, p. 80) name different factors that can influence the degree of credibility that governments possess. First, the history and reputation of organizations and their spokespersons affect the way the public interpret their credibility (Boin, 't Hart, Stern, & Sundelius, 2005, p. 80). Naturally, well-known examples of successful crisis management in similar situations increase the credibility of a person or organization. Second, the initial responses set the tone (Boin, 't Hart, Stern, & Sundelius, 2005, p. 80). The public tends to be more forgiving to overreaction than to underestimation of a crisis. Third, the timing of messages is a deciding factor in the way they will be perceived (Boin, 't Hart, Stern, & Sundelius, 2005, p. 81). Organizations can make use of news cycles and typical rhythms to select the best time for sharing information, depending on what type of information it is.

2.2.3.2.1 Credibility management

Boin, ‘t Hart, Stern and Sundelius (Boin, 't Hart, Stern, & Sundelius, 2005, p. 81) mention two classic errors that can lead to a self-inflicted decline in one’s reputation of trustworthiness. First, leaders can overemphasize hopeful scenarios. Their credibility then becomes dependent on how accurate their prognosis turns out to be. Second, leaders may be tempted to appear decisive. Quick promises can erode a leader’s credibility when it turns out that leaders cannot deliver on them (Boin, 't Hart, Stern, & Sundelius, 2005, p. 82).

2.2.3.3 Meaning-making strategies

Crisis leaders can make use of certain moves to influence public opinion and emotions (Boin, 't Hart, Stern, & Sundelius, 2005, p. 82). Boin, ‘t Hart, Stern and Sundelius (2005, p. 82) discuss engaging in framing, using crisis rituals, and masking information that is related to the crisis at hand.

2.2.3.3.1 Framing

Framing involves the selective use of data, arguments, and historical analogies and the formation of coalitions with like-minded groups (Boin, 't Hart, Stern, & Sundelius, 2005, p. 83). Even labeling something as a ‘crisis’ is a way of framing, because using such a term is an assessment of its seriousness itself. These kinds of terms also lead to certain expectations by the public. On the other hand, policy makers can de-politicize an event by framing certain events in judicial or scientific language (Boin, ‘t Hart, Stern, & Sundelius, 2005, p. 83).

(20)

Successful framing efforts can temporarily give a policy maker more freedom. When people perceive an event as a crisis, they are more easily persuaded that ‘the end justifies the means’. Policy makers can put this to political use (Boin, 't Hart, Stern, & Sundelius, 2005, p. 84).

2.2.3.3.2 Rituals

A ritual is defined as “symbolic behavior that is socially standardized and repetitive” (Boin, ‘t Hart, Sundelius & Stern, 2005, p. 84). Boin, ‘t Hart, Stern and Sundelius specifically mention a visit to the relevant sites, and to victims and operational staff as an essential symbolic strategy for leaders in response to a crisis (2005, p. 85). The importance of this cannot be exaggerated and failure to publicly display compassion is sure to bring intense public-relations problems (Boin, 't Hart, Stern, & Sundelius, 2005, p. 85). Opponents of crisis leaders can also engage in using rituals, for example during anti-police demonstrations.

2.2.3.3.3 Masking

Masking is a term used to describe the actions taken by crisis leaders to conceal the reality of a crisis. They may not tell the full story, downplay the seriousness, or hide aspects of their own actions (Boin, 't Hart, Stern, & Sundelius, 2005, p. 87). Although authoritarian regimes are infamous for masking efforts, it is also done by democratic governments. If it succeeds, it can buy the public actor time or political credit. However, when discovered or not followed up by real measures that deal with the problem, masking efforts can backfire and erode credibility (Boin, 't Hart, Stern, & Sundelius, 2005, p. 87).

2.3 Summary table

Table 1: Summary of the theoretical framework. Crisis:

A situation wherein policy makers “experience a serious threat to the basic structures or the fundamental values and norms of a system, which under time pressure and highly uncertain circumstances necessitates making vital decisions” (Boin, ‘t Hart, Stern & Sundelius, 2005, p. 2).

Crisis management tasks Factors influencing the performance on crisis management tasks Sense making:

The process of recognizing that something

There are organizational barriers that hinder effective sense making:

- Many organizations are not designed to look for trouble. - Some organizations suffer from

their rationality and efficiency, allowing mistakes to travel through an organization.

(21)

extraordinary is developing.

- For people within an organization, sharing

information with other people or organizations is not necessarily their first concern.

There are individual barriers that hinder effective sense making.

- Because of the characteristics of threat perception, sometimes a potential threat fails to get recognized as such.

- Stress can hinder effective sense making.

- Other individual constraints that hinder effective sense making. Decision making and

coordination: Decision making and coordination entails that crisis response leaders have to address pressing issues, decide how to act and what to prioritize and coordinate this with other

organizations.

Decision making There are vulnerabilities of crisis teams responding to crisis situations.

- Newness and conformity, as the crisis response team will not always be familiar to one another or have experience of working together as a group. - Internal conflicts can hinder

effective decision making in the crisis response team.

Conditions for successful decision making in crisis situations.

- Experience in working together is a factor for success.

- Good relations within the crisis response team prior to the crisis are a factor for success.

(22)

Coordination

There are a number of problems in coordinating a crisis response.

- The mobilization of the members may be slow, incomplete or simply impossible.

- The organizations may be required to perform tasks that they were not designed for. - It is not always clear when

upscaling should happen. - Ineffective communication or

conflicts can lead to problems. Factor that can foster

effective crisis coordination.

- The formalization of the coordination process.

Meaning making: The process of reducing uncertainty and confusion and providing the public with answers to the questions of what is going on, why it is happening and what needs to be done.

There are a number of factors determining the effectiveness of governmental crisis communication efforts.

- The degree of preparedness. - The coordination of outgoing

information. - The degree of

professionalization.

- The degree of credibility that public actors possess. Crisis leaders can make

use of certain meaning making strategies to influence public

perceptions and emotions.

- Framing: the selective use of data, labeling.

- Rituals: symbolic behavior. - Masking: concealing the reality

of a crisis.

(23)

Ch. 3: Methodology

In this chapter, I will discuss the research design and the use of comparative case study design and argue why this research design is suitable for answering the research question posed in this thesis. I will also provide an overview of the cases that are studied and argue why these cases have been selected. In the second part of this chapter, I will elaborate on the operationalization of the variables and the chosen research method.

3.1 Research design: comparative case studies

Yin (2003, p. 5) states three conditions that must be met for using a case study research design to conduct research. The first is the type of research question posed, this must be a “how, why?”-question. In my research, I am looking for an explanation. Why did this happen? How did this happen? To what extent was the performance on crisis management tasks by the mayor and other crisis response leaders an important factor? A case study design is a fitting design to answer such questions. The second condition is that the research must not require the control of behavioral events. This condition is met in this thesis. I cannot control behavioral events, but I have formulated scope conditions to ensure that these cases are in fact comparable to each other. These scope conditions are further defined in the next paragraph. The third condition is that the research must focus on contemporary events. This condition is also met in this thesis, since I am looking at a contemporary phenomenon that has been taking place over the last 6 years and is still persisting. Since all conditions are met, I conclude that a case study is a suitable design for conducting this research.

3.1.1. Case selection

Rohlfing (2012, p. 15) distinguishes six generic types of case studies. The case study design that I will employ is a cross-case comparison. I will look at the cases defined in paragraph 3.1.2 and look for differences and similarities between these cases.

As a means for case selection I have made use of a typical distribution with a difference in kind (Rohlfing, 2012, p. 77). I look at the outcomes of the events and determine to what extent the social event turned into a crisis. The dependent variable therefore is an answer to the question: to what extent did this event turn into a crisis? The definition of a crisis has been established in the body of knowledge. The independent variables are derived from the crisis management tasks framework by Boin, ‘t Hart, Stern and Sundelius (2005): these are the crisis management tasks and these are regarded in this thesis as factors influencing whether the social media event escalated into a crisis or not. I am seeking variance in the dependent variable and try to keep maximum similarity with other cases on the independent variables by setting scope conditions (Rohlfing, 2012, p. 89). In doing so, I am able to compare the performance on crisis management tasks by mayors and other crisis leaders, or the crisis response leaders, in similar crisis situations. The scope conditions for the cases are:

(24)

• Event on Facebook and/or other social media; • Number of people invited;

• Taking place in a medium sized town or city;

• Reported by traditional media, but escalating on social media first;

All these cases are similar on these scope conditions. I am interested in factors that influence the performance on crisis management tasks and lead to a different crisis outcome. Boin, ‘t Hart, Stern and Sundelius argue that policy makers can limit the damage by responding well to a crisis, but they can also increase the impact of a crisis when they fail. By keeping the scope conditions similar and identifying relevant factors influencing the performance on crisis management tasks by mayors and other crisis leaders while influencing the dependent variable (the crisis outcome), I will study this relation.

3.1.2 Definition of the cases

For my research, I will be looking at factors influencing the performance on crisis management tasks

by crisis response leaders during a ‘Project X’-event in a city in the Netherlands. These cities are:

Haren, Schiedam, Arnhem, and Enschede.

3.1.3 Selected cases in further detail

I have selected different ‘Project X’-situations in the Netherlands as cases that will be studied in this multiple case study design. I have looked for similarities and differences across these different cases. As shown, there are a number of scope conditions that allow me to compare the different cases. Additionally, factors influencing the performance on crisis management tasks by the crisis response

leaders is the independent variable. To analyze this, the theoretical framework of ‘crisis management

tasks’ will be used. The dependent variable, whether a ‘Project X’-situation develops into a crisis or not, differs per case.

(25)

Table 2: Description of the cases.

Scope conditions Independent

variable Dependent variable Project X Large number of people invited Taking place in medium sized town or city Event on Facebook and/or other social media Reports by traditional media, but escalating on social media first Factors influencing performance on crisis management tasks by crisis response leaders Crisis outcome

Haren YES1 _YES2 _YES3 _YES4 _CRISIS

Schiedam YES5 _YES6 _YES7 _YES8 _{NO CRISIS}

Arnhem YES9 _YES10 _YES11 _YES12 _{NO CRISIS}

Enschede YES13 _YES14 _YES15 _YES16 _DAMAGES

AND ARRESTS

1_{Comissie 'Project X' Haren. (2013). Twee Werelden. Den Haag: Ministerie van Justitie en Veiligheid.} 2_Ibid.

3_Ibid. 4_Ibid.

5_{Algemeen Dagblad. (2012, September 23). Zeven arrestaties na oproep 'Project X' in Schiedam. Opgehaald van} Algemeen Dagblad:

https://www.ad.nl/binnenland/zeven-arrestaties-na-oproep-project-x-in-schiedam~a8b603cd/ 6_Ibid.

7_Ibid. 8_Ibid.

9_{Van Gelder, H. (2017, November 22). Politie verijdelt project X feest Arnhem. Opgehaald van de Gelderlander:} https://www.gelderlander.nl/arnhem-e-o/politie-verijdelt-project-x-feest-arnhem~a28cc9edd/

10_Ibid. 11_Ibid. 12_Ibid.

13_{Meesterburrie, A. (2017, October 14). Vernielingen en arrestaties bij 'Project X' Enschede. Opgehaald van} NRC: https://www.nrc.nl/nieuws/2017/10/14/vernielingen-en-arrestaties-bij-project-x-enschede-a1577289 14_Ibid.

15_Ibid. 16_Ibid.

(26)

3.1.4 Collection of data

The six most commonly used sources of evidence case study designs are: documentation, archival records, interviews, direct observations, participant-observation, and physical artifacts (Yin, 2003, p. 85). According to Yin (2003, p. 85) documentary information is likely to be relevant to every case study topic. Another important and much used source of information for case studies is the interview. This research will make use of both of these sources. Extensive reports have been written about ‘Project X’ Haren. These reports provide sufficient information to analyze the case of ‘Project X Haren’. However, less information is available about the ‘Project X’-situations that did not turn into a crisis. I have conducted interviews with involved government officials responsible for managing these crises to gather the data needed to study the cases of ‘Project X Schiedam, Arnhem, and Enschede. Additional data is retrieved from news reports and a letter from the mayor of the municipality of Schiedam to the city council of Schiedam.

3.1.5 Reflection on generalizability of the research design

A frequently heard criticism on case studies is the limited generalizability of research done using this method (Babbie, 2013, p. 340). However, this differs for the different types of case study designs and the chosen case selection method. Rohlfing (2012, p. 203) explains that for a distribution-based selection strategy with typical cases, the generalizability is limited to similar cases. This is not necessarily problematic for my research as I have made it explicit that I am looking into these ‘Project X’-events specifically. I am not claiming that I will be able to say something about the behavior and actions of crisis response leaders in all crisis situations. I can, however, use this research to say something about factors influencing the performance on crisis management tasks by crisis response leaders during the escalation of social media events in the Netherlands. In this thesis, I argue that this topic and this research is relevant to both society and academics.

3.2. Research method

3.2.1 Research method: directed content analysis

To analyze the data, I am making use of a directed content analysis approach. The themes, categories and codes that are used to systematically analyze the data, are derived from the theoretical framework constructed in chapter 2 (Hsieh & Shannon, 2005, pp. 1281-1283). This allows me to determine if the factors described in the theoretical framework are relevant to the performance on crisis management task in the escalation of social media events. Based on the theory, the relevant factors are then analyzed in more detail to discover similarities and differences between the factors that influence the performance on crisis management tasks, and the outcome of the potential crisis in those cases. The outcome of this analysis will be shown in a summary table, providing an overview. This method also allows to add relevant factors to the framework, by identifying factors that do not fit the pre-determined codes.

(27)

3.2.2 Operationalization of the theoretical framework

The theoretical framework as described in Chapter 2 has been used to gain a deep understanding of the crisis management tasks and the factors that influence performance on this task. Here, the general theory has been adapted to be used as a suitable framework to analyze ‘Project X’-events in the Netherlands. Table 3: Operationalization of the theoretical framework.

Crisis: Indicator for

differentiation in crisis outcome:

A situation wherein policy makers “experience a serious threat to the basic structures or the fundamental values and norms of a system, which under time pressure and highly uncertain circumstances necessitates making vital decisions” (Boin, ‘t Hart, Stern & Sundelius, 2005, p. 2).

The amount of damages and incidents. Crisis management tasks

Factors influencing the performance on crisis management tasks

Theme Category Factors (CODES) Indicators to assess

presence/absence of this factor in the case Factors influencing sense making. Organizational barriers that hinder effective sense making.

- Many organizations are not designed to look for trouble. (LOOKING FOR

TROUBLE)

- Some organizations suffer from their rationality and efficiency, allowing mistakes to travel through an organization. (TRAVEL THROUGH ORGANIZATION) - For people within an organization, sharing

information with other people or organizations is not

necessarily their first concern.

Systems in place to monitor social media, active scanning of social media, contacts with people that can warm them

Not relevant as discussed in analysis

Contacts with other organizations, information from the police,

(28)

(SHARING INFORMATION) Individual barriers that hinder effective sense making.

- Because of the characteristics of threat perception, sometimes a potential threat fails to get recognized as such.

(THREAT PERCEPTION) - Stress can hinder effective

sense making. (STRESS)

- Other individual constraints that hinder effective sense making.

(OTHER FACTORS HINDERING SENSE MAKING)

Agreement within the crisis response team on threat, actions taken upon receiving first signals, informing other people, setting up a crisis response team

Mentioned in interview, feelings of fear, uncertainty, inexperience

Brought up in interviews, apparent in text from context

Factors influencing decision making and coordination: Decision making Vulnerabilities of crisis teams responding to crisis situations.

- Newness and conformity, as the crisis response team will not always be familiar to one another or have experience of working together as a group. (NEWNESS AND

CONFORMITY)

- Internal conflicts can hinder effective decision making in the crisis response team. (INTERNAL CONFLICTS)

Previously worked together before, regular meetings, know each other, known organization

Disagreements about actions that need to be taken, responses from the interviewees that indicate conflict (context), records of fights or disagreements

(29)

Conditions for successful decision making in crisis situations.

- Experience in working together is a factor for success.

(EXPERIENCE WORKING TOGETHER)

- Good relations within the crisis response team (prior to the crisis) are a factor for success. (GOOD RELATIONS) - It must be clear who leads.

(CLEAR LEADER)

Previously worked together before, regular meetings, know each other, known organization

Responses from the

interviewees that point at good relations, explicit statements about this, positive experiences in the past, pleasant

cooperation

Formal rules, authority, procedures, explicit statements from the interviewees

Coordination Problems in coordinating a crisis response.

- The mobilization of the members may be slow, incomplete or simply impossible.

(MOBILIZATION) - The organizations may be

required to perform tasks that they were not designed for. (NOT DESIGNED FOR TASK)

- It is not always clear when upscaling should happen. (UPSCALING

PROCEDURES)

- Ineffective communication or conflicts can lead to problems.

Explicit statements about mobilization problems in reports or transcripts, shortages of police officers at the location,

Not relevant as discussed in analysis

The existence of scenarios, agreements about upscaling, formal plans and rules, agreements about upscaling

Conclusions from the reports, statements by the respondents that they had trouble

communicating, problems with telephones or communication

(30)

(INEFFECTIVE COMMUNICATION)

systems

Factor that can foster effective crisis

coordination.

- The formalization of the coordination process. (FORMALIZATION)

Procedures, protocols, GRIP-protocols, rules Factors influencing meaning making: There are a number of factors determining the effectiveness of governmental crisis communication efforts.

- The degree of preparedness. (PREPAREDNESS MEANING MAKING) - The coordination of outgoing

information. (COORDINATION OF COMMUNICATION) - The degree of professionalization. (PROFESSIONALIZATION) - The degree of credibility that

public actors possess. (CREDIBILITY)

Employees ready, communication plans ready, capacity, knowledge Agreements with the other parties involved in the crisis response, communication plan, common communication, joint press releases

Professional communication employees, communication departments, spokespersons

Statements about credibility, previous track record, no lies, misleading

Crisis leaders can make use of certain meaning making strategies to influence public perceptions and emotions.

- Framing: the selective use of data, labeling.

(FRAMING)

- Rituals: symbolic behavior. (RITUALS)

- Masking: concealing the reality of a crisis.

(MASKING)

Selective use of data, statements about framing efforts in interviews, conclusions from the Commission investigating Haren

Mayor or alderman visiting the crisis site, press conferences, interviews

Misleading, hiding facts, incomplete communication, statements about masking in interviews, transcripts, records, Commission reports

(31)

Ch. 4: Analysis

In this chapter, I will first provide a description of the cases that are being researched in this thesis. I will then analyze the cases, using the insights from the theoretical framework constructed in chapter two and the research method described in chapter three.

4.1 Description of the cases

4.1.1 Project X Haren

Project X Haren took place on September 21, 2012 in Haren (Commissie 'Project X' Haren (1), 2013, p. 5). The event was first posted on Facebook, on September 7. It quickly escalated on social media, with 3500 people saying they would be attending the event within a day (Commissie 'Project X' Haren (1), 2013, p. 9). Eventually, between 3000 – 5000 youngsters gathered in Haren to celebrate a party that never was. Project X Haren resulted in riots which lead to wounded people, a million euro in damages, looting, and violence against the police (Commissie 'Project X' Haren (1), 2013, p. 5).

4.1.2 Project X Schiedam

Project X Schiedam took place on September 22, 2012 in Schiedam. Employees of the safety department of the municipality of Schiedam discovered that people were calling for a Project X Schiedam on Twitter (Appendix B, p. 1). The event escalated on social media and was retweeted thousands of times. It is unclear how many people intended to show up in Schiedam, as many people got stopped from going to Schiedam in the first place (Interview 2, 2018, p. 15). A couple hundred made it to Schiedam for the event (Interview 2, 2018, p. 15). Project X Schiedam resulted in five arrests, but no other incidents or damages (Appendix B, p. 2).

4.1.3 Project X Arnhem

Project X Arnhem took place on September 28, 2012 in Arnhem. The municipality of Arnhem discovered the event on Facebook, where it quickly escalated (Interview 3, 2018, p. 27). According to news reports, thousands of people said they would be attending the event (Klis, 2012, September 28). The police arrested 9 people, but there were no other incidents or damages reported (Suse, 2012, October 10).

4.1.4 Project X Enschede

Project X Enschede took place on October 13, 2017 in Enschede. The event was discovered on Facebook, where it quickly escalated. About 150 people showed up in Enschede for the event. Project X Enschede resulted in disturbances, arrests and 10,000 euro in damages (Jansen, 2018, October 4).

Factors influencing the performance on crisis management tasks in dealing with escalating social media events

Institute of Security and Global Affairs

Faculty of Governance and Global Affairs – Leiden University