Learning From Crises: Why Some Fail and Others Prevail

(1)

(2)

Learning from Crises: Why Some Fail and Others Prevail

Crisis Learning in a Medical Institution

A thesis presented to Crisis & Security Management MSc Leiden University

Rick Garnier

Student number: 2471795

Thesis supervisor: Dr. Cabane Second reader: Dr. Wolbers Date of submission: 26-05-2020

(3)

ACKNOWLEDGEMENT:

Every thesis can be regarded as, and should form, a challenge. Yet, the time at which this research was conducted forms even a larger challenge. Right now, the entire world faces the danger of the COVID-19 virus that has led to many protective measures that constrain each and every one of us in our daily lives. These measures are necessary to protect us. Yet, it calls on everyone’s strength, patience and effort to get through this.

Before I start this thesis, I want to praise our hero’s in the medical institutions and thank them for their incredible hard work to keep us alive and safe. Further, I especially want to thank the participants of this research because, despite that COVID-19 requires so much of their attention, these participants were willing to help me by being flexible and making time in their busy schedules so that they could be interviewed. Due to your efforts, in multiple ways, I was able to write this thesis.

But let us not forget the many other hero’s in every other sector. The people who keep our economy going and the people who are fighting to remain standing up right in these difficult times. We will get there. There is a future ahead of us.

(4)

1. Introduction:

1.1 Facing potential crises in hospitals:

Whether people like it or not, crises occur (Van der Zalm, 2008, p. 1; Roper, 1999). Organizations also experience these crises. These type of crises can be as destructive and dangerous as societal crises (Pearson & Clair, 1998). This research dives into how organizations learn or fail to learn from internal crises. Before explaining the research further, the next paragraph offers a short example of what kind of crises this research focusses on.

The example is a Dutch hospital (hospital A)2 that was struck by a crisis in June 2018. In this hospital, a water pipe broke without anyone’s knowledge. Within hours, several parts of this hospital were filled with large amounts of water and hot steam. This caused a power blackout of both regular and emergency power. Furthermore, it damaged several critical (electronic) hospital systems. Some patients relied on these systems and thus their safety was endangered. Therefore, some parts of the building had to be evacuated. Besides the danger for the patients, the business continuity was disrupted as several parts of the hospital were unable to operate for days. It took three days for the hospital to fully recover.

Hospitals face a broad range of potential crises, ranging from fires to disease outbreaks and stolen data (Onderzoeksraad voor Veiligheid, 2008, p. 5; COT, 2016, p. 1; NOS, 2020; Nolan, 2019). Such crises can cause the cancellation of hundreds of appointments for people who need acute care, damage to critical systems, and most importantly, threaten people’s lives (Mansfield-Devine, 2017, p. 16; Leupen, 2015; Healthcare & Public Health Sector Coordinating Councils, n.d.).

1.2 The problem of not learning from crises:

Besides knowing what kind of crises this research looks into, it is important to explain that there is a problem that requires research.

2_{This hospital will form the main case of this research. More explanation on this follows in this chapter. Yet, it is}

important to note that the hospital wants to remain anonymous. Therefore, this hospital is from now on referred to as ‘Hospital A.’

(7)

Despite that crises themselves are problematic, the academic field of organizational science argues that learning from these crises is a challenge in itself (Mitki & Herstein, 2011; Elliott, 2009; Choularton, 2001; Barnett & Pratt, 2000; Donahue & Tuohy, 2006). Several authors in this field argue that it is important to learn from crises to be able to prevent similar ones in the future (Kim, 1998; Ulrich, Jick, & von Glinow, 1993; Wang, 2008; MaximaMC, 2018, Broekema, 2018). The problem is, however, that research done by Shell (2004) and Smith and Elliott (2007) concludes that organizations do not always learn from crises. By understanding what causes the difference in learning outcomes, organizations can discover how to learn better from crises (Broekema, 2018).

Two fields of organizational science, being organizational (crisis) learning and organizational change, suggest some influential factors that might explain the difference in organizational learning outcomes (Kotter & Cohen, 1995; Kim, 1998; Weick, 1993; Choularton, 2001; Roux-Dufort, 2000; Elliott, 2009; Wang, 2008). Different authors in organizational science indicated the following factors: the level of urgency to learn, organizational culture which is open for change, the presence of a leader to promote learning and the willingness of organizational units to develop, the organizations level of sense-making of the crisis, and contextual factors (Kotter & Cohen, 1995; Kim, 1998; Ulrich, Jick, & von Glinow, 1993; Boyce, 2003; Choularton, 2001, Elliott, 2009, Lagadec, 1997; Roux-Dufort, 2000; Fiol & Lyles, 1985; Weick, 1993; Wang, 2008; Schilling & Kluge, 2009; Stern, 1997; Smith & Elliott, 2007). The organizational perspective on these factors is specifically relevant because this research looks into crises and learning within organizations.

Yet, some of the authors repeatedly and explicitly state that there has not been enough empirical research on whether and to what extent these factors play a role. Therefore, they call for more practical research and examples to either prove or modify their insights (Roux-Dufort, 2000, p. 26; Schilling & Kluge, 2009, Wang, 2008; Kim, 1998; Broekema, 2018, p. 17).

This research answers that call by offering empirical examples. The examples are four cases of crises within an organization to limit the reliance on one, possibly exceptional, case3. The examples come from one organization, hospital A. The research focuses on one organization because researching more would cause too much divided attention within a limited timeframe. This research selected Hospital A because it experiences several (small) crises a year4.

(8)

Therefore, it is possible to select crises within the same organizational framework. This leads to less contextual variables that can influence the difference and thus make the comparisons more trustworthy (Bennett, 2004, p. 31).

In two cases, the hospital learned relatively much. In the other two, the hospital did not learn (or less). By analysing these cases, the research attempts to both explain to what extent the theory is correct and to get a better understanding of what causes the difference between learning and failing to learn. This research thus aims to answer the following research question:

Why did hospital A learn more from certain organizational crises with similar causes, disruptive potentials, and time of occurrence, but not (or less) from other crises within that same scope?

1.3Purpose of researching the learning from crises:

This section explains what the purpose of the current study is. The research question provides an indicator for the purpose as it asks why hospital A learns from some crises but not from others. The research thus aims to determine why the learning outcome differs in each case. This leads to the main purpose of the research, which is to provide insight into what requires more effort to enhance the amount of learning. An enhanced amount of learning could lead to more crisis prevention or better crisis preparation in the future (Cooke & Rohleder, 2006, p.1; Broekema, 2018, p. 14-15). Yet, as it is difficult to generalize from a few case studies, a secondary purpose is to see to what extent the organizational science framework can explain these differences.

1.4 Social and academic relevance of research on learning from crises: To justify this research, it is important to further specify its social and academic relevance. The social relevance refers, in this research, specifically to the relevance for the society in the field of crisis and security management. This research attempts to reveal weak spots in the process of learning from crises. By identifying these weak spots, organizations can put additional efforts into removing them (Wybo, 2004, p. 30-32). This can improve the organization’s safety in the future and thus holds specific relevance in the field of crisis and security management (Cooke & Rohleder, 2006, p.1; Broekema, 2018).

(9)

Furthermore, this research has academic relevance. Roux-Dufort (2000), Schilling and Kluge (2009), Wang (2008), and Kim (1998) mention that studies in crisis management lack enough empirical research on organizational learning from crises (Roux- Dufort, 2000, p. 26). This is because many studies assume that organizational learning always follows a crisis (Smith & Elliott, 2007, p. 519). This research makes an empirical contribution to the field of organizational learning from crises and drops the assumption about automatic learning. This can create new insights into the field of organizational learning from crises (Broekema, 2018).

1.5 The research structure:

This section explains what one can expect from the remainder of this research. The next chapter describes the theoretical insights on organizational learning from an organizational perspective. Subsequently, chapter three explains the research methods including the description of the four cases. Chapter four first analyzes the four cases individually and then offers a cross-case analysis. Finally, chapter five offers a conclusion based on this analysis and makes recommendations for medical institutions and future studies.

(10)

2. Theoretical framework – Learning from crises:

This chapter reviews the literature in organizational science about learning (from crises) in organizations. The review seeks to determine what influences learning in order to answer the research question ‘Why did hospital A learn more from certain organizational crises with

similar causes, disruptive potentials, and time of occurrence, but not (or less) from other crises within that same scope?’ The organizational perspective forms the basis of the theoretical

framework because the research question is about an organization. Therefore, taking a specific organizational angle at learning seems the most suitable (Boyce, 2003). Moreover, Pearson and Clair (1998, p. 60) and Boyce (2003) mention that the organizational perspective has much importance when studying organizational learning. Further, the focus of the literature review is on factors that influence learning because these might identify why learning outcomes differ.

2.1 Organizational learning and organizational crises:

Before jumping directly into what organizational science says about influential factors on learning, this section explains the basics of what organizational learning and crises are and how they relate.

2.1.1 Organizational learning:

To know what influences organizational learning, one should understand what organizational learning is. The first step should be to look at the definition of learning. This is later specified into the meaning of organizational learning.

In the academic literature there are several perspectives on such a definition. De Houwer, Barnes-Holmes, and Moors (2013) discuss two of these perspectives. The first perspective is that leaning is ‘an effect of experience on behavior’ (p. 361). Yet, they find this too simplistic for it does not involve changes in the person that learns. Therefore, they contrast it with the definition ‘changes in the organism that result from experience.’ However, they believe that both are not practical enough and therefore provide an alternative: ‘changes in the behavior of an organism that are the result of regularities in the environment of that organism’ (p. 633). In other words, all kinds of factors can cause changes within an organism, and that is learning. Besides these, there are many other possible definitions. Alexander, Schallert, and Reynolds (2009) did a study in which they attempted to find common ground amongst this diverse range

(11)

of definitions. They found that learning is both a process and a product, that it is interactional, and above all, that it includes change (p. 178).

Although there are many definitions of learning, it has become clear that it at least involves a change in behavior due to experiences or environmental stimuli. This forms a basic understanding of learning. Now, organizational learning can be further specified.

Organizational learning is a specific form of learning. The concept first came up in managerial literature somewhere in the early 1980s (Mitki & Herstein, 2011, p. 455). Since then, the idea of learning in, from, and of organizations continues to grow much in popularity as a tool for improvement (Wang, 2008, p. 431).

Despite its growing popularity, the concept remains somewhat vague. The first definition described it as the ‘detection and correction of error’ (Argyris & Schon, 1978, p. 2). Yet, Levy (1994) finds that the concept is much more complex, which makes it harder to define. Therefore, he labels learning as a ‘conceptual minefield’ (p. 280).

In addition to the definitions above, Hedberg (1981) defines organizational learning as organizations that use knowledge to adjust so that they fit better in their environment. Another definition is that an organization creates shared values and knowledge based on the experiences of individuals within that organization (Friedman, Lipshitz, & Overmeer, 2001). Whereas Mitki and Herstein (2011) mention Gravin’s definition that refers to the creation, acquisition, interpretation, transfer and holding on to knowledge. Further, Boyce (2003) adds that

“acquiring and practicing new competencies are aspects of organizational learning (…)” (p.

125).

All of these definitions have different angles. Yet, Stern (1997) argues that there is agreement on what it means in the basis. He perceives this agreement as ‘generating acquisition of new knowledge, skills, ways of thinking or modes of social organization’ (p. 69). Mitki and Herstein (2011) give a similar definition. However, they add that it is ‘the internal capacity of organizations to learn from experience, to examine and adopt new ideas and transform them into policy and action plans’ (p. 455). It is thus both a process and a result (Koh, 1998). Kirwan (2013) puts it as: ‘Organizational learning makes organizations do things differently’ (p. 49). Boyce (2003) agrees and says ‘(…) literature demonstrates that organizational change and organizational learning are inextricably linked’ (p. 133). Thus, the concept is very similar to

(12)

However, organizational learning, in contrast to regular learning, takes place at two different levels: individual and organizational. Although it are mainly the individual members within an organization who acquire knowledge and change behavior, organizational learning is not simply the sum of their individual learning (Kim, 1998; Stern, 1997; Wang, 2008). The process of creating knowledge involves the communication and spreading of the newly acquired knowledge to others in the organization. This is the basis for integrating it into the organization’s memory and actions (Kim, 1998). The learning of one individual within an organization is therefore not enough. Multiple people are needed to translate the learning into common knowledge and action (Attewell, 1992, p. 5-6; Ulrich, Jick, & von Glinow, 1993). Also, it must be noted that there are two forms of learning: single-loupe and double-loupe learning. Single-loupe learning involves small changes that fit into the institutional framework. For example, a machine does not work as intended and is reset. No further investigation is done why it behaved differently and the organization goes on as it always did. This is learning because information (notice wrong behavior) causes a change (the reset). However, for more permanent learning, double-loupe learning is required (Boyce, 2003, p. 126). This involves a need to understand why something happened and based on that cause change in underlying assumptions. It occurs when an organization understands that it cannot go on the way it does. Afterwards, the organization can start to perform and behave differently from before. This causes a more permanent change and thus more permanent learning (Boyce, 2003, p. 127). In his research, which is relatively similar to this one, Broekema (2018) also considered the many definitions and took into account the ‘conceptual minefield.’ Based on all these definitions, he came up with the following: ‘[Organizational learning is] the process of

acquiring new knowledge and understanding (cognitive) and the transposing of this new knowledge and understanding into improved organizational actions (behavioral)’ (p. 24).

This research will work with that definition for two reasons. First of all, it includes all important aspects of the other definitions. Secondly, Broekema performed a similar research to the current study but focused on governmental organizations only. By sticking to the same definition, it is easier to compare research outcomes and build a base that other researchers can use in the future.

(13)

2.1.2 The value of learning:

The research question aims to answer why hospital A learns in one case but not (or less) in another. This implies that there is some importance in learning. This section explains why it is important for an organization to learn in the first place.

To start off with, a failure to learn can lead to crises within organizations (Smith & Elliott, 2007, p. 519). Even organizations that have had impressive successes over the years can disappear to the background because of a failure to adapt and thus a failure to learn (Appelbaum, 1997). In the modern world, environments are more complex, uncertainties grow, and small mistakes can lead to larger problems (Lagadec, 1997; Ulrich, Jick, & von Glinow, 1993; Koh, 1998). For the organization to survive, adapting and thus learning is more important than ever (Barnett & Pratt, 2000, p. 75-77).

In addition, learning can determine an organization’s survival because it updates the organization’s knowledge. Based on this, organizations make decisions about their action, values, and priorities. Moreover, learning can introduce new skills that make adaptation easier (Kim, 1998, p. 508). Both equip organizations with a more complete toolset for decision-making and action-taking which helps organizations to solve or avoid problems (Kim, 1998; Ulrich, Jick, & von Glinow, 1993; Wang, 2008, Broekema, 2018).

Finally, learning can make organizations more effective and efficient. This provides them with a competitive advantage (Mitki & Herstein, 2011, p. 455; Ulrich, Jick, & von Glinow, 1993, p. 55; Gray & Gonsalves, 2002). More importantly, a failure to learn often leads to costs, differing from financial to social and political (Elliot, 2009). If an organization does not learn, it is not able make the best-informed decisions. This positions them behind competitors and makes them fit less in their environment (Mitki & Herstein, 2011), which makes them less likely to survive on the long term.

2.1.3 Crises offer learning opportunities:

The research question is about crises within hospital A and thus about crises within an organization. Before this research can explain the relationship between learning and crises, it must explain what organizational crises are.

(14)

effect, and means of resolution, as well as by a belief that decisions must be made swiftly’ (p.

60). Whereas this is a relatively detailed definition, Nystrom and Starbuck (2004) offer a shorter one, namely: ‘a situation that seriously threatens and organization’s survival’ (p. 101). In contrast, Hermann (1972) provides a longer definition again: ‘a situation that threatens the high

priority goals of the organization, restricts the amount of time available for response, and surprises decision-makers by its occurrence, thereby engendering high-levels of stress’ (p. 13).

These definitions all highlight a threat to the organization. Yet, on other accounts, they differ. Therefore, this research seeks a combination between these definitions and the definition that hospital A has itself. Internal documents of hospital A refer to organizational crises as events that threaten, damage or harm the organization’s business continuity, critical systems, or persons. The combined definition that this research provides is, therefore: Unexpected events

with possibly ambiguous cause that threatens, damages or harms the organization’s business continuity, critical systems, or persons and needs a quick response to limit the threat, harm or damage.

Then how do these threats connect to organizational learning? According to Roux-Dufort (2000), organizational crises offer opportunities (p. 26). Crises reveal weak spots and previously hidden causal connections within organizations. These weak spots and causal relations are not always easy to discover (Broekema, 2018). A little ‘push’ is often necessary for people to see them. Crises can be such a push. Organizations also learn under ‘normal’ circumstances. Yet, learning often accelerates after a crisis ‘push’, since there is sudden attention for the need to change (Stern, 1997). This can create momentum for taking action (Kim, 1998; Wang, 2008). Simon and Pauchant (2000) therefore call a crisis ‘a powerful trigger for change and learning’ (p. 6).

Carley and Harrold (1997) refer to this shift in attention as a window of opportunities. They state that after a crisis, a window ‘opens’ with the opportunity to make changes. Yet, as time passes, these windows ‘close’ because people’s attention to the need of change shifts away. After that, learning is less likely (p. 312).

So, by being aware, people can take measures to prevent a similar crisis in the future or limit the potential damage (Wybo, 2004, p. 30-32). This relates to the definition of organizational learning because crises can cause the acquisition of new knowledge and offer the opportunity to take improved actions.

(15)

Crises can thus offer the perfect opportunity to learn and change. Unfortunately, organizations do not always learn (Roux-Dufort, 2000). Lagadec (1997) mentions that organizational science makes assumptions on what factors are influence organizational learning from crises (Lagadec, 1997). These are discussed in the following section.

2.2 Facilitating and inhibiting factors for organizational learning:

The previous section described the theoretical background of organizational learning and crises. This section delves deeper into the organizational perspective on organizational (crisis) learning. The reviewed literature comes from organizational change, (crisis) management, and (crisis) learning journals and books. This scientific field is recommended by Boyce (2003), who did a previous literature review on what influences organizational learning (p. 121-123).

The reviewed literature identifies at least five main factors that could influence organizational learning. Although other factors might have some influence as well, not all variables can be measured (Bennett, 2004, p. 30-31). Therefore, this research focuses on the five (broad) factors that cover the most often mentioned elements in the reviewed literature. The factors come mainly from organizational science because this angle allows for a specific focus and it seems most relevant to the topic (Boyce, 2003).

Kotter & Cohen (1995) specifically mention three of these five factors as being influential. These factors are (1) the level of urgency to learn, (2) an organizational culture that is open for change, and (3) the presence of both a leader to promote change and organizational units that are willing to follow. All these factors are also discussed by Lagadec (1997) and further supported by other authors such as Fiol and Lyles (1985, p. 804), Elliott (2009), Roux-Dufort (2000), Kim (1998), Schilling and Kluge (2009), and Sullivan and Nonaka (1986).

Some of them offer two additional factors. Weick (1993), Lagadec (1997, p. 28), Donahue and Tuohy (2006, p. 10-17), and Wang (2008, p. 430), amongst others, mention that (4) forms of sense-making plays a role as well. In addition, Schilling and Kluge (2009) and Stern (1997) argue that learning can also be influenced by (5) the organization’s context, such as the organization’s form.

(16)

review, she indicated and explicitly mentions the influence of organizational cultures, leadership, sense-making, and context (in the form of resources and organizational structure). Also, she implicitly mentions urgency by saying that commitment and continuous attention is required (p. 121-129). Ekboir et al. (2009) make similar statements based on previous research, as does Broekema (2018).

Although other factors might be influential too, these five are repeatedly and explicitly mentioned. Further, these factors are broad and can include ‘other’ factors. For example, communication can be important for sense-making (Qureshi, Kamal, & Keen, 2009). Therefore, these five will form the basic background for the research. Yet, the research allows for the opportunity to find other influential factors as well.

This research created a conceptual model that shows what factors influence learning (figure 1). Organizational learning from organizational crises is the dependent variable. The left side of the model shows the independent variables which can determine the variability in learning and thus possibly answer the research question. The model does not imply that all factors must be present for learning to take place.

(17)

Figure 1. Conceptual model of organizational learning from organizational crises

The following sections describe the influence of each factor.

2.2.1 Sense-making:

This section describes how sense-making can influence organizational learning from crises. Weick (1993) wrote an article on how sense-making can collapse during and after a crisis. He used an example to show that crises can sometimes feel completely senseless because they do not fit in the worldview of the organization. If that happens, the organization might have a difficult time understanding what is going on and how to deal with that. Wolbers, Boersma, and Groenewegen (2017) mention several examples in which sense-making turned out to be difficult, such as after 9/11 and a chemical spill in Sweden (p. 1525).

(18)

Cornelissen, Mantere, and Vaara (2013) expand on this. They explain that if a crisis does not fit the worldview of the organization, the organization might still approach it in a way that suits their ‘normal’ form of sense-making. Yet, when these do not align, one can draw wrong conclusions. This inhibits solving the crisis and can even escalate it.

So, to be able to take actions and to learn from a crisis, organizations must know what happened (Kovoor-Mistra, Clair, & Bettenhausen, 2001; Boyce, 2003, p. p.123). Without fully understanding the causes of a crisis, one cannot draw the right conclusions and therefore not learn (Weick, 1993; Broekema, 2018). Understanding and thus making sense is the first step toward learning (Smith & Elliott, 2007, p. 521-522).

Yet, understanding is not self-evident. After a crisis, there are often several interpretations of what happened (Smith & Elliott, 2007, p. 521; Choularton, 2001, p. 68-69; Kuchinke, 1995, p. 313; Broekema, 2018). Therefore, a well-constructed review session is required, in which an analysis takes place, to discover what can and should be learned from the crisis. If this does not happen, lessons are often forgotten (Lagadec, 1997, p. 28; Donahue & Tuohy, 2006, p. 10-17). Unfortunately, crisis reviews do not always happen and if they do, they sometimes remain on a superficial level (Choularton, 2001, p. 64; Smith & Elliott, 2007, p. 521-522; Broekema, 2018). This causes organizations to not find the root causes or solely focus on one problem while other things played a role as well. If that happens, organizations do not learn the real lessons (Wang, 2008, p. 430).

Besides misinterpretations, also disinformation can play a role. Crises can reveal things which some people do not want to be revealed. This sometimes causes people to hold back information (Lagadec, 1997, p. 28; Elliott, 2009; Elliott, Smith, & McGuinness, 2000).

Also, limited availability of information and routine thinking can distort the level of sense-making because it limits the consideration of alternatives (Smith & Elliott, 2007, p. 527; Wolbers, Boersma, & Groenewegen, 2017, p. 1536; Elliott, 2009).

In short, making sense of all aspects of the crisis is important to be able to learn from it. The wrong interpretation might cause people to believe that it does not reoccur in a later stage, while it might (Roux-Dufort, 2000, p. 26).

(19)

2.2.2 Urgency:

This section describes the influence of urgency on organizational learning. Yet, for the sake of clarity, this section first briefly mentions what this research means with urgency. In the context of this research, urgency refers to the feeling that action or change is necessary (Donahue & Tuohy, 2006; Choularton, 2001). In this way, urgency is linked to feeling the need to learn. Sullivan and Nonaka (1986) support this and state that a certain degree of urgency is essential for people to decide that learning is necessary. Without urgency, they claim, learning often fails (p. 130).

For example, if one interprets a crisis as a one-time event, people might not feel the urgency to take actions and thus to learn. This happened to airplane manufacturer McDonnell Douglas. Several of their planes had to crash before they understood that the occurrence of a certain problem with their planes was not a chance of one in a million (Choularton, 2001, p. 66-67). Roux-Dufort (2000, p. 25) says that most organizations react in a similar way. Organizations often want to return to the normal situation as quickly as possible, rather than learning from the crisis and make actual changes. However, as soon as the situation reaches ‘normality,’ the urgency to change is often gone. Postponement of learning from crises is thus dangerous as it lessens the level of urgency (Carley & Harrold, 1997). A crisis report then slips to the bottom of the drawer and becomes another statistic from which organizations forget to learn (Lagadec, 1997; Broekema, 2018, p. 151).

Unfortunately, people sometimes tend to minimize dangers and rationalize situations in simplified ways (Roux-Dufort, 2000, p. 26). If an organization takes standpoints such as ‘that would not happen here’ or ‘that only happens once,’ they lack the urgency to learn (Smith & Elliott, 2007, p. 532; Boyce, 2003, p. 129; Ekboir et al., 2009; Madsen & Desai, 2010, p. 457). When an organization believes it is not vulnerable, it is unlikely that it takes measures and therefore fails to learn (Pearson & Clair, 1998, p. 69; Choularton, 2001).

According to Kim (1998), it requires effort to solve problems. The effort put in solving the problem depends on the amount of energy that an organization is willing to put into the solution. The intention of the organization and the priorities it has thus influence their level of urgency to learn (Elliott, 2009).

(20)

Smith and Elliott (2007) name one other way in which urgency can influence learning. They draw on the reasoning from Turner (1976) that decoy problems could play a role. In other words, organizations can be distracted from learning because of other problems, which some believe to be more pressing, surface (Carley & Harrold, 1997).

Urgency can therefore influence the level of learning in several ways.

2.2.3 Organizational culture:

This section explains how organizational culture can influence the amount of organizational learning. Similar to the previous section, this section first explains what is meant with organizational culture.

Briefly explained, the organizational culture is ‘the way we do it here.’ In a more structured way, it is the underlying assumptions, beliefs, and values that are shared by the individuals in an organization. This directs the way in which the organization thinks, prioritizes and acts (Schein, 1990, p. 115; Boyce, 2003, p. 122; Ekboir et al., 2009).

Culture can influence learning in several ways. The first way is indirect. As the culture partly determines what the organization finds important and prioritizes (including safety and security), it also determines the urgency and sense-making of the organization (Fiol & Lyles, 1985, p. 804).

Secondly, culture can directly influence learning because it determines how open the organization is to change (Fiol & Lyles, 1985, p. 804; Carley & Harrold, 1997, Broekema, 2018). With that, organizational culture can be both a facilitator and a barrier to organizational learning. On the one hand, if an organization is open for change, learning might be easier as there is the potential willingness to do things differently (Elliott, 2009). On the other hand, resistance to learning often has its roots in culture (Lagadec, 1997, p. 29). Learning a certain lesson might not fit in the organizational routines. If people then stick to their cultural routines, culture forms a limiting barrier to learning (Schilling & Kluge, 2009; Elliott, 2009; Elliott, Smith, & McGuinness, 2000). Therefore, sometimes old habits should be unlearned before new ones can be learned (Nystrom & Starbuck, 2004; Elliott, 2009).

If organizations want to learn and adjust behavior, they must challenge existing ideas. Listening to ‘dissent’ might help to steer away from rigid beliefs that are incorporated in the culture

(21)

(Choularton, 2001, p. 64; Nystrom & Starbuck, 2004; Ulrich, Jick, & von Glinow, 1993, p. 53). However, challenging routines can cause uneasy feelings within an organization and therefore lead to resistance (Wang, 2008, p. 434).

Culture can therefore influence learning because it either opens doors for change or keeps them shut out of feelings of fear, uneasiness, or because the lessons do not fit the organization’s view.

2.2.4 Leadership and ownership:

A fourth factor is the presence of both leadership and ownership. This section describes how the presence or absence of these can influence the amount of organizational learning.

The main way in which leadership can influence learning is by making sure that learning actually happens. Learning involves changes and changes require time. However, as explained, urgency generally fades away as time passes. Therefore, learning requires a person or a group of people to keep up the feeling of necessity to learn and make sure that both evaluations and actions are completed (Kotter & Cohen, 2012, p. 1-14; Heorhiadi, la Venture, & Conbere, 2004, p. 7-9; Ulrich, Jick, & von Glinow, 1993, p. 61; Kim, 1998; Elliott, 2009; Boyce, 2003; Ekboir et al., 2009; Broekema, 2018). Lagadec (1997) finds that a small taskforce, that has the mandate to bring change and support others in bringing about this change as well, is the ideal form to do this. Unfortunately, most organizations do not have such a taskforce (p. 27-29).

Although leadership and urgency are linked, they are not the same. Urgency is necessary for someone to be willing to take the lead. Yet, similarly, a leader is necessary to keep up the urgency.

Besides leadership, also ownership can influence the amount of learning. If a certain department requires change, the people in that department must also see themselves as part of both the problem and the solution. They should accept that they play a role. However, if people do not want to be or do not regard themselves to be accountable, this generally results in bargaining. If a department does not want to change but they have to, offers are made for smaller forms of change. Unfortunately, bargained changes are often not optimal. Therefore, this inhibits learning (Stern, 1997).

(22)

2.2.5 Organizational structure and context:

Finally, this section describes how organizational structure and context might influence organizational learning. Although context can be wide, some specific elements are mentioned in the reviewed literature.

The first important part of the context is the structure of the organization (Fiol & Lyles, 1985, p. 804; Elliott, 2009; Kim, 1998, p. 508; Boyce, 2003; Carley & Harrold, 1997; Broekema, 2018). An example of how the structure can influence learning can be seen in how hierarchy works in an organization. A strict or clear top-down hierarchy makes learning dependent on a specific group of people. In such a structure, one must often convince the highest ranks of employees in an organization that learning is required before actions can be taken. If these employees are not convinced, there might not be enough support for learning to take place (Schilling & Kluge, 2009).

Other contextual factors can be laws, regulations and available resources. These can play an important role in both a facilitating and inhibiting manner. However, they are usually difficult to influence. As long as laws and regulations do not prohibit certain measures, they will probably not form an obstacle. Yet, they can also make certain forms of learning mandatory. This is, however, never as efficient as learning on the organization’s own initiative (Roux-Dufort, 2000, p. 26-27).

Resources are also important. Time and money are regularly involved in learning. If they are not available, bargains must be made about what lessons can and cannot be implemented (Stern, 1997, p. 71; Boyce, 2003, p. 121; Ekboir et al., 2009; Carley & Harrold, 1997), or learning cannot take place at all.

In conclusion, learning is vital for the organization’s survival and crises offer opportunities to learn. The literature review on organizational science indicates that at least five factors have an influence on the amount of organizational learning. These are sense-making, urgency, learning facilitating culture, leader- and ownership, and the organizational structure and context. Although other factors might also play a role, this research focusses on these factors because they are repeatedly mentioned as being important.

Therefore expectation of this research is: The amount of learning from crises is affected by

sense-making, urgency, organizational culture, leader- and ownership, and structure and context.

(23)

However, Roux-Dufort (2000) mentions that the conditions for learning have rarely been proven and are thus assumed (p. 26). The same counts for the information which is available about organizational learning barriers (Schilling & Kluge, 2009). Empirical research on this is thus missing (Schilling & Kluge, 2009, Wang, 2008; Kim, 1998; Elliott, Smith, & McGuinness, 2000).

Smith and Elliott (2007) specifically stated:

‘(…) research in this area suffers from a lack of empirical evidence. (…) there is a need for further testing of the main theoretical perspectives outlined within the literature. In particular, research is needed to assess the role of the main barriers in preventing learning from occurring’ (p. 534).

This research will partially fill that gap in the theory. It provides empirical research that might even put the theory to the test. So, an added expectation is: Besides the five facilitating and

inhibiting factors for learning, other factors influence the amount of learning as well.

In this way, the research makes an attempt to not exclude the possibility of other influential factors.

2.3 Operationalization of the key concepts of the research:

As Roux-Dufort (2000), Schilling and Kluge (2009), Wang (2008), Kim (1998) and Smith and Elliott (2007) mentioned, one needs to empirically test the influence of the five factors. This section operationalizes these factors so that one can measure them. The theory above forms the basis for the operationalization. Besides the five influential factors, this section also operationalizes organizational learning and organizational crises.

Table 1 provides an overview of the indicators of the main concepts. These indicators can show to what extent a certain concept is present or not.

(24)

Table 1. Operationalization of key concepts.

Concepts Positive indicators Negative indicators Sources

Organizational learning

1. New knowledge or skills are acquired;

2. The acquired knowledge or skills are integrated at all relevant organizational levels; 3. The acquired knowledge or

skills are translated into action;

4. The actions taken lead to a more positive outcome for the organization.

1. No new knowledge or skills are acquired;

2. The acquired knowledge or skills are not integrated at all relevant organizational levels;

3. The acquired knowledge or skills are not translated into action;

4. The actions taken do not lead to a more positive outcome for the organization. (Working Knowledge, n.d.; Stern, 1997; Mitki & Herstein, 2011; Kirwan, 2013; Kim, 1998; Wang, 2008; Boyce, 2003; Broekema, 2018) Organizational crisis

1. The event is unexpected; 2. The event threatens, damages

or harms the organization’s business continuity, critical systems, or persons; 3. A quick response is required

to limit the threat, harm, or damage.

1. The event is expected; 2. The event does not threaten,

damage or harm the organization’s business continuity, critical systems, or persons;

3. No quick response is required to limit the threat, harm, or damage. (Pearson & Clair, 1998; Nystrom & Starbuck, 2004; Hermann, 1972; Broekema, 2018; internal document of hospital A5₎

Sense-making 1. All crisis causes are known; 2. The crisis is extensively

evaluated;

3. The organization knows what actions to take;

4. People understand one another.

1. Not all crisis causes are known;

2. The crisis is not or limitedly evaluated;

3. The organization does not know what action to take; 4. People do not understand

each other.

(Sullivan & Nonaka, 1986; Lagadec, 1997; Smith & Elliott, 2007)

Urgency 1. Change is deemed necessary; 2. Change is deemed necessary

immediately;

3. Creating change is more or equally important to other issues.

1. No change is deemed necessary;

2. Change is not necessary (now);

3. Change has no priority over other issues. (Donahue & Tuohy, 2006; Choularton, 2001; Roux-Dufort, 2000; Turner, 1976; Shetye, 2013) Positive organizational safety culture6 1. Organization is willing to change;

2. Mistakes are discussed; 3. Safety and security are always

top priority.

1. Organization is reluctant to change

2. Mistakes are rather not discussed

3. Safety and security can be traded off for other values.

(Halaj, 2017; Choularton, 2001)

5_{The internal document is the organization’s crisis, incident and calamity plan. This is not open for public view.} 6_{Culture is in this case not only organizational but also per department, surrounding a specific crisis.}

(25)

Leadership and ownership

1. Someone leads the knowledge

gathering and change implementation;

2. Leader puts efforts into

creating change on a daily basis;

3. Parties involved in the crisis

accept that they actively play a role in the learning process.

1. Nobody takes the lead in the learning process;

2. Leader puts no or limited effort into creating change; 3. Parties involved in the crisis

do not take part in the learning process. (Kotter & Cohen, 2012; Stern, 1997) Structure and context

1. Necessary measures are financially viable; 2. Necessary measures are

lawful;

3. The organization’s structure poses no barriers to learning.

1. Necessary measures are financially unviable; 2. Necessary measures are

unlawful;

3. The organization’s structure poses barriers to learning.

(Donahue & Tuohy, 2006; Schilling & Kluge, 2009)

An important note must be made about this table. In reality most of the concepts in the table will not likely occur in an extreme form of either being there or not. The positive and negative indicators are therefore only a point of reference. As the indicators can be in between the two extreme outcomes, table 2 offers a more nuanced form of measuring that applies to all of the concepts:

Table 2. Concept indicator levels. Concept level Description

Positive Nearly all, or all, of the data shows processes that are in line with the

positive indicators.

Somewhat positive The data shows processes that are mainly in line with the positive indicators

but not in the extreme form. Possibly some negative indicators are present but the positive indicators are dominant. For example: Only some change was deemed necessary. Most people found it necessary, but not all.

Neutral The data shows processes that line up with both positive and negative

indicators. None of these indicators is dominant.

Somewhat negative The data shows processes that are mainly in line with the negative indicators

but not in the extreme form. Possibly some positive indicators are present but the negative indicators are dominant. For example: Most people do not deem change necessary, but a small group does.

Negative Nearly none, or none, of the data shows processes that are in line with the

positive indicators.

(26)

3. Methodology:

This research wants to answer the question ‘Why did hospital A learn more from certain

organizational crises with similar causes, disruptive potentials, and time of occurrence, but not (or less) from other crises within that same scope?’ The question refers to the comparison

between several crises. Therefore, this research has a comparative case study design. It compares four crisis cases7 to find answers to the research question. The comparison analyzes at least the five factors found in the literature. The remainder of this chapter describes how the research is conducted, why in this way, and what its limitations are.

3.1 Comparative case study design:

This section explains what the design of this research is and which cases it includes. Yet, the section first explains why a comparative case study is relevant for this research question. The research question seeks to find factors that cause differences in the outcomes of organizational learning. To understand differences, it is useful to make a comparison between cases’ similarities and differences (Yin, 2014). Furthermore, as there is not much empirical evidence on influential factors yet (Roux-Dufort, 2000), some unexpected factors might be relevant as well. By spending much time on a limited amount of cases, it is more likely to discover these unexpected factors than if one must spend that time on many cases (Verhoeven, 2015, p. 151-155; Sreejesh, Mohapatra, & Anusree, 2013, p. 39; Rich & Ginsburg, 1999, p. 372; Hammarberg, Kirkman, & Lacey, 2016, p. 499). Therefore, conducting a limited amount of comparative case studies is important for answering the research question.

This research analyses four different cases. The next section explains how this research design and the selected cases are relevant for answering the research question.

3.1.1 The selection of hospital A and the four crisis cases:

This section explains the selected cases and how they are relevant for the research. Yet, before that, it explains briefly why this research selected hospital A as the overall case.

(27)

The research focusses on hospital A because by focussing on one organization, external factors, such as location, remain the same and thus have less influence on the results. Thereby, Hospital A faces several crises a year. This increases the number of potential cases compared to organizations with few crises.

The research analyses four cases because looking into several cases lowers the possibility that the results are coincidental (Yin, 2014). In two of these cases, the organization learned. In two other cases, it did not learn or at least less. Due to this division, the research can show what factors played a role in each case and hence answer the research question.

Further, there are additional reasons for selecting these four cases. Namely, factors that are outside the scope of the research should be as similar as possible. Otherwise, these factors could influence the results (Yin, 2014). Therefore, the research selected four crises in hospital A that are somewhat similar, yet diverge in their outcome. Yet, cases are virtually never completely similar in nature. This needs to be taken into account.

Also, the research only selected cases from which can be determined if learning has taken place. Because the cases have similarities but different outcomes, these can serve as examples to find an explanation for the differences and thus answer the research question.

Table 3 shows an overview of the similarities and differences between the selected cases. The differences, with exception of the outcome, need to be taken into account in the analysis as possible influential side-factors.

(28)

Table 3. Case similarities and differences.

Case 1 Case 2 Case 3 Case 4

Location Hospital A Hospital A Hospital A Hospital A

Frequency Once Repeatedly Once Repeatedly

Date Past 3 years Past 3 years Past 3 years Past 3 years

Cause Technical and human Technical and human Technical and human Technical and human Department in charge of learning Safety and environment; Technical

Technical Safety and

environment; Technical Safety and environment Involvement of external parties in the learning process No No No No Involvement of the board Yes No Yes No Media coverage

Some to limited None to limited None to limited None to limited

Impact potential Disrupted business continuity and decreased human safety Disrupted business continuity and decreased human safety Disrupted business continuity and decreased human safety Disrupted business continuity and decreased human safety

Learning Limited/ none Limited Mainly Yes Yes

The remainder of this section describes the four cases and why they are relevant.

Case 1 – Electricity blackout:

In June 2018, a waterpipe in the basement of hospital A broke without anyone’s knowledge. Water, flowing from the pipe, damaged a hot steam pipe and power generators. Within hours, parts of the hospital were filled with water and hot steam, causing an power outage. This caused some critical systems of the hospital to go offline. Some of the hospital’s patients rely on these systems for their health and they therefore had to be evacuated. Because the situation developed rapidly, it was difficult to find the cause. Yet, a quick response was necessary to limit the damage. In the end, parts of the hospital were out of business for three days. Internal documents show that after the crisis new knowledge has been gathered, but that virtually no actions have been taken based on that knowledge. No significant measures were implemented to prevent a similar crisis in the future.

(29)

Based on the operationalization in chapter 2.3, this case was an organizational crisis. The event came unexpectedly, threatened the business continuity, damaged infrastructure, and endangered people’s health. In addition, it required a quick response to limit all this. Yet, as new knowledge has not been implemented, the hospital still faces the same risks and therefore did not learn (enough) from the crisis.

Case 2 – Contaminated water damage:

In 2018, hospital A introduced a new waste processing system. Employees can put their waste in a grinder at their department. The waste is then transported, with the hospitals wastewater, though pipelines to a tank where it is degraded. Since 2018, the hospital repeatedly experienced that its new pipelines were clogged. Due to the clogging, contaminated water (from for example toilets), repeatedly flooded departments at unexpected moments. At the start of 2020, contaminated water filled the surgery rooms. Although the cause is relatively simple, it resulted in repeated small crises because each time the water damaged (electrical) infrastructure. Moreover, as it is contaminated water, it posed health risks and this causes flooded departments to be closed for several hours. To limit the water from spreading to more areas, quick responses are required. This is in line with the conceptualization of an organization crisis.

Unfortunately, internal documents show that the organization implements only temporal solutions to end the crisis. However, the organization takes very limited preventive measures. This shows because the same crisis keeps repeating itself. Therefore, the organization did not learn (enough) from this case.

Case 3 – Radioactive and biological infectious waste:

In hospital A, human excreta from patients, who are treated with radioactivity, normally goes through shredders behind the toilets directly to specially secured containers where the radioactivity of the waste can erode so that it can be safely transported to external processors later. The shredders, however, were new but unsuitable for toilet paper. In mid-2018, the shredders broke down and caused the radioactive material to be rerouted to less safe containers. Unfortunately, these containers were not suitable for the material and started to leak waste which was dangerous for people’s health. Thereby, the waste includes viruses and bacteria, which causes a threat of infections. Therefore, the storage area had to be sealed off, patients treated with radiation were called off, several departments had to close and an emergency team

(30)

again. Internal documents show that the some departments in the organization used this experience to create new protocols and make constructional adjustments to the building to prevent a similar crisis in the future. The organization thus learned because they implemented preventive measures based on new experiences.

This case was an organizational crisis because it was unexpected, threatened people’s health and the business continuity and required a quick response to limit the danger to the organization.

Case 4 – Chemical and cytostatic spills:

At the start of 2017, hospital A experienced several chemical and cytostatic spills. These chemicals and cytostatic are dangerous for people’s health and can be easily spread through the air and underneath shoes. Therefore, the organization needs to clean the spill up quickly to limit the danger. Unfortunately, the hospital prepared inadequately for these spills because it lacked enough material to clean them up safely and quickly. Therefore, several departments had to be closed for hours and people were not allowed to leave the location of the spill (to prevent spreading). The latter caused people to be exposed to the chemicals longer than healthy. These spills often come unexpectedly, require a quick response to prevent escalation and threaten people’s health and the business continuity. Therefore, they can be seen as small organizational crises.

After several of these spills, the organization implemented measures at all relevant departments to enable a quick clean up and with that prevent an accident (spilling a chemical) from becoming a crisis by disrupting the business continuity and threatening people’s health. Implemented measures are (amongst others) new protocols, new training, and new material to prevent escalation. The organization thus learned from this case.

3.2 Collecting data from the cases:

This section explains how the research collected data from the cases through interviews and documents to answer the research question.

(31)

3.2.1 Interviews:

To answer the research question, one must determine what causes the different outcomes in each case. Therefore, the research requires data on what happened after the crisis and specifically on how the learning process went. Yet, there is only a limited number of people and documents available that have knowledge or data on it.

Therefore, this research conducted eight semi-structured interviews with people who were involved in the learning process after the crises. The research conducted (at least) two interviews for each case to limit the reliance on the memory of one person. The selection of interviewees was based on their direct involvement in and knowledge of the learning process (Weiss, 1994, p. 1-2; Verhoeven, 2015). Each of the interviews lasted 30-85 minutes and were recorded. Yet, to protect the interviewees from possible repercussions, all are treated confidentially.

The purpose of the interviews was to collect information on what factors influenced the learning process and how they influenced it. The main part of the data is information on how the five factors, identified in the literature review, played a role. Yet, the interviews also aimed to collect information on how other potential factors might have influenced the process. The interview guide in appendix A contains all the questions.

To answer the research question, the interviews collect data on (1) factual information about how the learning process went, and (2) factual information on the extent to which potential influential factors were present. One can later compare this data for each case to find the differences that the research question is looking for.

3.2.2 Documents:

The research also uses a limited amount of internal documents from hospital A. The data collected from them is the same as from the interviews. These documents can verify (triangulate), add details, and provide a background to the interviews. None of these documents are publicly available. Appendix D contains a list with the names and release dates of the documents.

(32)

3.3 Method of analysis:

After collecting the data, this study required an analysis to answer the research question. The analysis determines how the learning process of each case went by focusing on the five influential factors. The analysis of each case went through several stages. These stages are explained in this section.

The first step of the analysis involved the latent coding of the data. The codes are based on the theory (Yin, 2014, p. 136) and are therefore equivalent to the operationalization in table 2 and 38. However, some additional codes were constructed during the research. The full coding scheme can be found in appendix C.

The second step was to go through each case individually. The research provides a schematic overview of each case. These tables show what codes were present for each indicator of each factor. By coding all influential processes found in the data, one can determine whether these indicators are mainly positive, neutral, or negative. Based on that, one can see to what extent each of the influential factors was present or not9_.

The information in these tables shows which factors (mainly) positively or (mainly) negatively influenced the learning process10. These tables will be put in context by describing the related information that the research collected.

After this is done for each case, a third step follows. In this step, the analysis combines the schematic overviews of each case into one table. In this table, one can see for each case if the factor had a mainly positive, neutral or a mainly negative indicator. Based on this table, one can look for patterns (Yin, 2012, p. 149). For example, one might find that both cases with a limited learning have a negative indication on the level of urgency, while both cases from which was learned have a positive indication. This shows a difference that might explain the different learning outcomes11. Yet, when all cases have a positive indicator for urgency, it most likely does not explain the difference because it is the same in each case12. Based on the similarities

8_{Thus, codes are ‘positive level of understanding, ‘negative level of understanding,’ etc.} 9_{These will most likely not be extremes but nuanced findings.}

10_{See appendix B for an example of how such a table looks like.}

11_{The tables serve as an overview only. The context provided by the interviewees determines the extent to which}

it had influence.

12_{This does not mean that the factor had no influence. This only means that it probably did not make the difference}

(33)

and differences conclusions can be drawn about which factors influenced the amount of learning.

Finally, a fourth step follows. Based on the comparison in step three, one can compare the outcomes to the suggestions made in theory. This shows to what extent these cases can support the theory.

3.4 Research limitations, reliability & validity: This section describes the limitations of this research.

The first limitation is its reliability. The research partly relies on the memories of interviewees and the researcher’s interpretation of the collected data. This can cause less objectivity and thus less reliability (Verhoeven, 2015, p. 302). However, this research takes some measures to enhance its reliability. Firstly, it documents and explains each step to make the research easier to replicate. If other researchers would take the exact same steps, it is likely that the conclusion will be similar to that of this research (Verhoeven, 2015, p. 189; Sacred Heart University Library, n.d.). Also, the research interviews multiple people per case to limit the dependency on one person’s memory. This dependency is also limited by triangulating the data with internal documents. The more sources confirm the data, the more reliable it is (Verhoeven, 2015, p. 303-304).

The second limitation concerns the external validity. The research looks into only four cases, all within the same organization. This is not representative for all organizations and the research is therefore unable to generalize to organizations outside hospital A. Yet, the theoretical insights do generalize. If this research’s conclusions are in line with the theoretical insights, the research can strengthen the generalizing theory. Furthermore, this research has an exploratory nature because of the limited conducted empirical research. In explorative research, it is acceptable to draw conclusions on one case and make assumptions for wider populations. Future research can then study these assumptions (Sacred Heart University Library, n.d.; Verhoeven, 2015, p. 187-197; 304).

On the other hand, the construct validity is relatively large. Case studies allow for a research to go in-depth and measure exactly what it intends to measure. This is because there is room for identifying unexpected variables (Verhoeven, 2015, p. 305). Thus, even though this research

(34)

focusses on only five factors, other factors are not excluded. This provides the research with a relatively strong internal validity (Verhoeven, 2015, p. 305-306).

(35)

4. Analyzing the four crisis cases:

This chapter individually discusses the collected data for each of the cases concerning electricity blackout, contaminated water, radioactive and biological infectious waste, and chemical and cytostatic spills. Consequently, a cross-case analysis determines the similarities and differences in the influential factors. With that analysis, one will be able to answer the research question ‘Why did hospital A learn more from certain organizational crises with

similar causes, disruptive potentials, and time of occurrence, but not (or less) from other crises within that same scope?’

4.1 Case 1 – Electricity blackout:

This section analyzes the electricity blackout case of June 2018, from which hospital A did not or limitedly learn. The analysis includes data from interviews with one of the organization’s safety managers, an involved crisis investigator, and a coordinator of the in-house emergency response unit. Furthermore, it includes data from an internal investigation report (Appendix D, document 4)13. The analysis describes how the learning process proceeded.

In the summer of 2018, a broken water pipe in the basement of hospital A caused an electricity blackout and water damage to critical electronical systems. Therefore, some of the hospital’s departments were unable to operate. It took the hospital’s crisis team several days to close off the water, clean everything up, and restart business as usual.

Several days after solving the crisis, the crisis team14_{and the hospital’s board came together.} Both the in-house emergency response unit and the crisis coordinator delivered an observation report to provide insight into what happened. In addition, according to the safety manager, the board stated that they wanted to know more about the crisis causes and how to prevent a similar crisis in the future. Following that statement the board appointed three impartial (internal) investigators and ordered them to conduct a detailed investigation. The crisis investigator says that this is unique because it was the first time to investigate a non-medical crisis this way. The ordered investigation indicates that the board was willing and found it necessary to find out

Learning From Crises: Why Some Fail and Others Prevail