Accident investigation reports: Writing, reading, learning? An in-depth case study of barriers to organizational learning from accident investigation reports within a Dutch Safety Region

(1)

Institute for Security and Global Affairs

Leiden University – Faculty of Governance and Global Affairs

Master Thesis Crisis and Security Management

Accident investigation reports: Writing, reading,

learning?

An in-depth case study of barriers to organizational learning from accident

investigation reports within a Dutch Safety Region

Program:

Master Crisis and Security Management

Author:

Mireille Francis van Abeelen

Student number:

S1216635

Date of admission:

08-06-2018

Word count:

29.266 (excl. bibliography & appendices)

59.753 (incl. bibliography & appendices)

Thesis supervisor:

Dr. S.L. Kuipers

(2)

Page 1 of 81

[Document subtitle]

“I do not speak as I think, I do not think as I should, and so it

all goes on in helpless darkness.”

(3)

Page 2 of 81

Finishing this master thesis would not have been possible without the help of the following

people, to whom I express my sincerest gratitude.

Thank you Drs. M. Kowalski for establishing the first contacts with the much needed

interviewees.

I would like to thank the respondents of the Safety Region Amsterdam-Amstelland and the

Inspectorate of Justice and Security for participating in this research.

My sincerest appreciation goes out to Dr. S.L. Kuipers for advising me and guiding me

through the process of writing a master thesis and thank you W.G. Broekema for your wise

words during our meetings.

Last, but certainly not least, I would like to thank my fellow students in the capstone

‘Learning from Crises and Safety Investigations’, as it was a pleasure to share and complain

(4)

Page 3 of 81

1. Introduction

1.1 Problem outline

The Dutch high-voltage grid had a supply security of 99,9999 per cent in 2016. While this is a positive number, reassuring us that there will always be light, a dark side is looming. This high-level security is at the same time a vulnerability. The Netherlands is a country that relies blindly on the millions of capillaries of a finely branched network that delivers electricity every second of the day to homes and businesses. It is therefore completely dependent on it1_{. Therefore, it is not surprising that the Dutch} National Coordinator for Security and Counterterrorism (NCTV) assesses the impact of a large-scale power outage, which is in essence a disruption of the country’s vital infrastructure, as severe. Especially since a power outage leads faster to larger societal problems in today’s digital society. Moreover, the longer a power outage lasts, the more problems occur due to the so-called cascade effects (Inspectie Veiligheid en Justitie, 2017, p. 3). The National Safety Profile of 2016 speaks of cascade effects such as a disruption of the telecommunication and public transport as well as the management of waste water and drinking water (Analistennetwerk Nationale Veiligheid, 2016, p. 109). It is therefore essential that society in general, and emergency services and various partners in the security domain in particular, are prepared for such an incident.

The severity of such an impact becomes immediately clear when looking at the large-scale power outage that occurred in March 2015, the largest one in Dutch history, which affected the provinces of Noord-Holland and (small parts of) Flevoland. Approximately one million households were out of power, and the consequences for daily life were far-reaching. Trains could not depart, traffic lights stopped working, and the air traffic on Amsterdam Airport Schiphol was disrupted. In addition, hospitals could not perform surgeries, and parts of mobile networks and other telecommunication networks stopped working (Inspectie Veiligheid en Justitie, 2016, pp. 11-12).

The Inspectorate of Justice and Security (Inspectorate), whose aim it is to signal risks and encourage improvement within the organizations of the Ministry of Justice and Security by supervising and investigating the quality of the performance of tasks2_{, conducted an accident investigation afterwards.}

1_{Stokmans, D. & Logtenberg, H. ‘Hoe één verroest draadje de halve Randstad platlegt’. NRC, July 7, 2017}

https://www.nrc.nl/nieuws/2017/07/07/kortsluiting-hoe-een-verroest-draadje-de-halve-randstad-plat-legt-11727393-a1565893

2_{Website Inspectorate of Justice and Security, Mission and Vision}

(7)

Page 6 of 81

Their findings gave way to make recommendations to the concerned Safety Regions and telecom providers in order to learn and improve their crisis management and response.

Only two years later, in January 2017, another large-scale power failure occurred, again in the province of Noord-Holland, which concerned more or less the same Safety Regions as in 2015. Again, the impact on society and daily life was quite severe. Not only were there enormous traffic disruptions, there was also a hospital in Amsterdam whose emergency generator failed and therefore had to evacuate vulnerable patients as the most critical systems, such as respiratory equipment, stopped working. Again, many of the transmission masts went down, which hindered crisis communication. Many citizens were literally left in the dark and had no chance of reaching the emergency services, due to on the one hand the failure of mobile networks and on the other hand an extreme peak in emergency calls and messages (from those who somehow did manage to have a working telephone connection) to the emergency control room (Inspectie Veiligheid en Justitie, 2017, pp. 3-4; 18). Especially in times of crises and calamities, a citizen rightly expects the emergency services to be accessible and to provide assistance in acute distress (Inspectie Veiligheid en Justitie, 2013, p. 7). Hence, the unreachability of the emergency control room and news reports about the tragic deaths of persons in need during the power outage led to the request of the Safety Region Amsterdam-Amstelland to the Inspectorate to investigate this incident.

Strikingly, the Inspectorate concluded in its findings that several recommendations for improving the reachability of the emergency control rooms and crisis communication in general were already made in earlier reports directed to among other the Safety Region Amsterdam-Amstelland (Inspectie Veiligheid en Justitie, 2017, p. 10). This leads to the question as to why these earlier recommendations were not adopted and acted upon. Why were the ‘lessons learned’ not learned?

1.2 Research question

In light of the above, this research attempts to answer the following research question: “What factors

hindered organizational learning from accident investigation reports within the Dutch Safety Region Amsterdam-Amstelland in the period between the power outage in 2015 and the power outage in 2017?”

1.3 Academic relevance

There is an extensive amount of literature on (barriers to) learning, both individual and organizational. However, most of the research has focused upon, in the case of organizational learning, companies. Therefore, this research contributes to the academic field in that it examines potential barriers to learning in public and governmental organizations. Moreover, there is not much empirical evidence on Dutch case studies and Dutch Safety Regions specifically have not been subject to substantial prior

(8)

Page 7 of 81

research before, in particular not in terms of learning. This research might yield interesting outcomes concerning the latter, especially since this research combines barriers coming from the literature on organizational learning with barriers to learning from accident investigation reports. Thus, not only is looked at what internal barriers originate within the organization itself, but also at what external barriers arise out of the characteristics of the accident investigations and the resulting reports. So far, the academic literature on accident investigations has solely focused on the public inquiry process itself, which is a process of knowledge acquisition, thus leaving out how the knowledge is subsequently transferred to and assimilated in the particular organization (Elliott & McGuinness, 2002, pp. 14-15). This thesis aims to fill this gap in the academic literature as it offers an interesting insight in the complete process of organizational learning from accident investigation reports in which two actors participate – the Safety Region as the organization and the Inspectorate as the accident investigator.

1.4 Societal relevance

Many authors have stressed the importance of learning after crisis, thereby implicitly explaining the linkage between the concept of organizational learning and crisis management. Broekema et al. (2017) explain that “it is of the utmost relevance that organizations learn from crises in order to prevent or adequately respond to future ones, because the consequences of crises are severe and the tolerance for mistakes is low” (p. 326). Carley and Harrald (1997) also state that human beings would like to live in a failure-free world and “learning is expected to be one of the key mechanisms through which organizations come to prevent and minimize the impact of disasters” (p. 310). However, as explained above, the Safety-Region Amsterdam-Amstelland had the opportunity to learn from the accident investigation reports published by the Inspectorate, but did not adopt and implement several of the recommendations. Hence, the findings of this research are of value to society as insights in the Safety Regions’ difficulties and barriers to organizational learning from accident investigation reports could raise awareness and therefore remove these barriers. Once these barriers are removed, it enables the organization to actually learn and enhance its preparedness for a future large-scale power outage. The outcome of this research is also useful for the other twenty-four Safety Regions in The Netherlands, as they too might have to face a similar incident and need to know whether their organizations have difficulties in learning. As indicated earlier, the severity and likeliness of such a disaster make it necessary that the Safety Regions and other security partners are well prepared. Additionally, the outcomes provide an insight in how the Safety Region receives, perceives, and deals with an accident investigation report from the Inspectorate of Justice and Security. The Inspectorate can possibly use these insights in order to improve their reports and recommendations to better facilitate the learning process.

(9)

Page 8 of 81

1.5 Thesis outline

The introduction of this thesis has explained its focus and relevance. The rest of the thesis is divided into chapters. Chapter 2 presents the theoretical framework in which this research situates itself. It draws upon the literature on organizational learning and learning from accident investigation reports and combines the identified barriers to these types of learning into a new comprehensive theoretical model. Hereafter, Chapter 3 will discuss the methodological framework that is used in this research to find an answer to the main research question. Subsequently, the choices will be justified and its limitations explained. The research findings are presented in Chapter 4, in which the results are analyzed. Chapter 5 is the last chapter of this thesis, which contains a discussion of the findings and a conclusion that links the findings together in order to provide an answer to the research question. The bibliography and appendices constitute the final pages of this thesis.

(10)

Page 9 of 81

2. Theoretical framework

The previous chapter has demonstrated the problem that is central to this research. The Safety Region Amsterdam-Amstelland experienced similar crises and had the opportunity to learn from the ‘lessons’ identified by the Inspectorate in the accident investigation reports that followed these crises. However, as evidenced by the accident investigation report on the power outage in 2017, several recommendations from previous reports have not been adopted and acted upon. This observation is key and seen in this research as an outcome of non-learning. However, it is of importance to note that there is no normative value judgment attached to this observation. The Safety Region Amsterdam-Amstelland may very well have legitimate or understandable reasons for why these recommendations have not been implemented. The question remains though why this has not happened and why the organizational learning process has not been completed. This research attempts to explain this by looking at factors or variables that hinder or interrupt the organizational learning process, or, in other words, barriers to organizational learning.

Therefore, the following chapter will look at the existing body of literature on organizational learning to find possible explanations (barriers) for why the Safety Region Amsterdam-Amstelland did not implement some of the previous made recommendations. This chapter will concentrate on the most important concept in this research – organizational learning. The concept will first be explored with a specific focus on barriers that hinder or prevent organizations from learning in general. Subsequently, organizational learning is linked more specifically to the literature on learning from accident investigation reports, as these reports are of great importance in this research. The position and role of accident investigations and public inquiries is discussed, again with a focus on barriers that might hinder learning from these type of reports. Lastly, a comprehensive theoretical framework is presented that combines the barriers to organizational learning with the barriers to learning from accident investigations that are relevant to this specific research.

2.1 Organizational learning

Learning is an ability all human beings naturally possess. Even daily, we acquire new knowledge and obtain skills through experience. However, it is deemed that non-living entities such as organizations can learn as well, given the extensive amount of academic literature on organizational learning. In 1963, Cyert and March were the first who proposed the idea that an organization could learn and that knowledge could be stored over time (Easterby-Smith & Lyles, 2011, p. 11)

Argyris (1999), well known for his work on learning organizations, explains why individuals can think, act, and learn on behalf of an organization. If a collectivity is able to make collective decisions, delegate

(11)

Page 10 of 81

authority for action to an individual in the name of that collectivity, and distinguish between who is and who is not a member of that collectivity, the collectivity becomes organizational (Argyris, 1999, p. 9). It follows that it “makes conceptual sense to say that on behalf of an organization individuals can undertake learning processes (organizational inquiry) that can, in turn, yield learning outcomes as reflected in changes in organizational theories of action and the artifacts that encode them” (Argyris, 1999, p. 9).

Schilling and Kluge (2009) define organizational learning as “an organizationally regulated collective learning process in which individual and group-based learning experiences concerning the improvement of organizational performance and/or goals are transferred into organizational routines, processes and structures, which in turn affect the future learning activities of the organization’s members” (Schilling & Kluge, 2009, p. 338). It is of the essence to note that this definition views organizational learning as a process. Schilling and Kluge (2009) explain that learning can be defined as a relatively permanent change in knowledge or skill resulting from experience, which emphasizes the dual nature of learning as process (perceiving and processing information, i.e. experience) and result (modified knowledge or skill) (Schilling & Kluge, 2009, p. 338). This is consistent with the way in which this research approaches organizational learning. In this case, it is assumed that somewhere in the process of perceiving and processing information, one or multiple barriers emerged that prevented achieving the desired result of modifying knowledge or skill. As explained before, some of the earlier made recommendations were not implemented by the Safety Region Amsterdam-Amstelland and it can therefore not be demonstrated that there is modified knowledge or skill.

Additionally, the definition given by Schilling and Kluge (2009) on organizational learning is in line with the description of Argyris (1999), as they both incorporate the individuals within the organization as well as the visibility of learning in changed routines or structures. Furthermore, this definition concentrates on the organizational psychological perspective on organizational learning, rather than concentrating purely on technological (such as IT systems for storing knowledge) or economical (such as intellectual capital) aspects (Schilling & Kluge, 2009, p. 338). This perspective is similar to the one taken in this research. Moreover, Schilling and Kluge (2009) developed a thorough theoretical framework, integrating existing theory and evidence on barriers to organizational learning by analyzing the existing theories, concepts, and evidence (Schilling & Kluge, 2009, p. 338). Hence, their definition is encompassing as it is derived from and based on the existing amount of literature on organizational learning. Clearly, the definition on organizational learning provided by Schilling and Kluge (2009) is valuable to this study for multiple reasons, which are listed above, and will therefore be adhered to in this research.

(12)

Page 11 of 81

2.1.1 Barriers to organizational learning

Most of the literature on organizational learning is optimistic, implying that once a learning process is initiated, it will progress naturally and smoothly. However, even though learning is something that we are all capable of doing, the reality shows that learning is neither an effortless nor an automatic or easy process (Berthoin Antal, Lenhardt, & Rosenbrock, 2003, p. 865). Barriers, defined as those factors either preventing organizational learning or at least impeding its practicability, stand in its way (Schilling & Kluge, 2009, p. 337).

Berthoin Antal et al. (2003) found that the literature on organizational learning was not able to provide a systematic overview of the barriers that hinder the learning process. They made a valuable contribution by providing such an overview and subsequently illustrating them with examples from four cases. They categorized the barriers into three groups: interrupted learning processes, psychological and cultural barriers to learning, and barriers related to organizational structure and leadership (Berthoin Antal, Lenhardt, & Rosenbrock, 2003, p. 865).

The first group is based on scholars who portrayed organizational learning as a process or cycle. According to Berthoin Antal et al. (2003), these authors provided the most systematic discussion of barriers to organizational learning. For example, Berthoin Antal et al. (2003) explain that March and Olsen (1975) built a model of learning that highlighted the linkages between individual beliefs, individual action, organizational action and environmental response. Subsequently, they identified several types of interruption to the learning cycle, such as when the connection between individual beliefs and individual action is interrupted if individuals are limited by their role in the organization and therefore unable to act on their learning. Another interruption happens when individuals do change their behavior but cannot persuade others to do the same (Berthoin Antal, Lenhardt, & Rosenbrock, 2003, p. 865). In other words, organizational learning is a process that can be interrupted at several stages, which prevents the newly acquired knowledge or practices from being transferred to the next stage.

The second group consists of barriers identified by writers that draw predominantly on findings from psychology relating to individuals and groups. Argyris (1991) wrote on the topic of defensive routines that individuals naturally develop to protect themselves from threatening situations and feeling incompetent, embarrassed or vulnerable (p. 8). Translated to the organization, this prevents an employee from critically examining its own role in the organization, thereby blocking the ability to learn or see and do things differently. Other academics focused for instance on the impact of success and failure on learning processes. There exist contradictory views in the literature regarding this impact. Some found that success becomes a barrier because it leads to managerial overconfidence in the ability to foresee risks, while others warn for ‘failure traps’ when one idea after another is tried

(13)

Page 12 of 81

out and then abandoned before enough experience has been accumulated for it to be used successfully (Berthoin Antal, Lenhardt, & Rosenbrock, 2003, p. 867). Clearly, these barriers arise out of personal or individual experiences and feelings explainable by psychological phenomena. While they provide for useful explanations for why learning may not occur, they are less easy to categorize than the barriers discussed above, as they cannot be divided per phase of the organizational learning process.

The last group focuses on barriers arising out of organizational structures and leadership. An example of a structural barrier is the existence of departmental structures within an organization since these can inhibit organizational learning by focusing the attention of employees on parochial rather than organization-wide problems. With regards to barriers related to leadership, it is considered problematic if a leader shows behavior that is not conducive to learning. For instance, leaders can impede organizational learning if they behave as though knowing were a greater virtue than learning or if they relegate people to followership, thereby limiting their ability to participate in leadership and learn (Berthoin Antal, Lenhardt, & Rosenbrock, 2003, p. 869). The previous group of authors looked at impediments emerging out of the individual or psychological level, while the scholars in this group seek explanations at the organizational and structural level. The three categorizations presented by Berthoin Antal et al. (2003) thus complement each other.

2.1.1.1 The ‘4I Model’

While the overview provided by Berthoin Antal et al. (2003) is useful, it is criticized by Schilling and Kluge (2009) for having no clear theoretical foundation (p. 338). The authors do acknowledge that since 1995 progress has been made in terms of addressing the barriers that work against or hinder organizational learning. However, a theoretical framework to describe and explain impediments to organizational learning still needs to be developed, which is precisely what Schilling and Kluge (2009) attempted to do (p. 337).

They developed an appropriate model for the organizational learning process in order to describe and explain the potential barriers in a systematic and well-organized manner. They used an existing model, designed by Crossan et al. (1999) and further extended by Lawrence et al. (2005). The ‘4I model’ describes four processes by which the different levels of organizational learning (individual, group and organization) are connected (Schilling & Kluge, 2009, p. 339-340).

The first process is called ‘Intuiting’, in which new insights are developed within an individual based on personal experience. One must note that this process takes place at the individual level and the process itself could therefore not be considered as organizational learning. However, as explained by both Argyris (1999) and Berthoin Antal et al. (2003), it is necessary for individuals to undertake learning

(14)

Page 13 of 81

processes on behalf of the organization. Thus, the first process of ‘Intuiting’ is a prerequisite for organizational learning and therefore incorporated in the ‘4I model’, which together with the following processes or phases constitutes the complete process through which organizational learning takes place. The second process is called ‘Interpreting’, in which the individual tries to convey its insights to others through words or actions. The third process is ‘Integrating’, in which a shared understanding among individuals is achieved that allows for coherent and collective action within the organization. This process happens at group level. The fourth process is ‘Institutionalizing’, in which the shared understanding is finally implemented in organizational systems, structures, procedures, rules, strategies, and eventually visible in organizational action (Schilling & Kluge, 2009, p. 340). These stages are important as Schilling and Kluge (2009) use them to classify the barriers described in the literature. Each stage has its own barriers and in this way, the dynamic aspect of organizational learning is taken into account (Schilling & Kluge, 2009, p. 340). One should note that the model is consistent with the first category identified by Berthoin Antal et al. (2003). Schilling and Kluge (2009) also present organizational learning as a process consisting of four stages, which allows for a clear and systematic analysis of barriers to organizational learning.

It is of great importance to understand here that not all processes are equally relevant to this research. The Inspectorate of Justice and Security has already identified the problems and lessons to learn in their accident investigation report about the power outage in 2015 and the report ‘1-1-2 onder de loep’ of March 2013. Moreover, these reports present solutions to these problems in the form of recommendations so that the Safety Region can implement and actually learn its lessons. Therefore, it is irrelevant to look at stages in which barriers occur that prevent organizations from identifying the causes and problems and coming to a shared understanding of what happened. Hence, the barriers to ‘Institutionalizing’, which is the last process of the ‘4I model’, are now discussed in more detail, since this stage is concerned with implementing the shared understandings (the ‘lessons learned’ and recommendations) in organizational systems, structures, procedures, rules, strategies, which become eventually visible in organizational action (Schilling & Kluge, 2009, p. 340).

2.1.1.2 ‘Institutionalization’

Schilling and Kluge (2009) recognize four blocks of barriers that impede institutionalization and adoption of innovations, or, in this case, recommendations. These blocks are: (A) a lack of trust in the innovation itself, (B) deficient skills and knowledge to adopt the innovation on part of the teams and employees, (C) a lack of management skills to provide consistent and systematic implementation, and (D) counteractive and opportunistic behavior from the organizational units and members that results in the rejection of an innovation (Schilling & Kluge, 2009, p. 355). Each of the four blocks will be discussed seperately in order to shed light on the different barriers that together compose a block.

(15)

Page 14 of 81 A. A lack of trust in the innovation itself

The first set of barriers arise from a lack of trust in the proposed innovation (Schilling & Kluge, 2009, p. 355). One barrier arises when organizations conclude that certain learning results are simply irrelevant for future purposes and do not need to be stored. Especially organizations that deal with rapid technological changes, such as those in the telecommunication or ICT branch, might face the problem that an innovation is already obsolete and outdated once its implementation is finally completed. This fosters the organization’s belief that institutionalizing innovations is unnecessary (Schilling & Kluge, 2009, p. 352). In the context of this research, it could also be that the ‘shared understandings’ are not implemented because the understandings acquired by the Inspectorate are simply not shared by the Safety Region. It is plausible that if the Safety Region does not agree with the findings of an external party, that being the Inspectorate, they will also disagree with the proposed solutions and thus have a lack of trust in the innovation or recommendation.

An additional barrier emerges when organizations “try to implement innovations and knowledge from a different culture”, as they are likely to “experience specific difficulties in finding appropriate ways to adopt and communicate it” (Schilling & Kluge, 2009, p. 353). It must be noted that this barrier has originally been identified by Kuznetsov and Yakavenka (2005), who found that international knowledge transfer is constrained by factors associated with national culture, such as for example linguistics (p. 1;15). As both the Inspectorate and the Safety Region are Dutch organizations, this barrier is not applicable to this research and will therefore not be taken into account.

The last barrier in this block occurs when there is a high degree of emerging management trends. Organizations may be seduced into relying more strongly on improvements suggested by external consultants than by their own employees in order to keep up with emerging trends to impress stakeholders and public (Schilling & Kluge, 2009, p. 353). Ironically, the improvements suggested by external consultants are in this research key to organizational learning since the Inspectorate is an external party. Therefore, this impediment is not regarded as a barrier in this context and will therefore not be used.

B. A lack of skills and knowledge

The second block of barriers that prevent innovations from being implemented and institutionalized are related to deficient skills and knowledge to adopt the innovation on the part of the teams and employees. Needless to say, a lack of knowledge, skills and abilities is an important obstacle to closing the learning cycle from innovation to practice (Schilling & Kluge, 2009, p. 354-355). But what factors exactly contribute to the existence of such a lack?

(16)

Page 15 of 81

First, it is fostered by a lack of time, space and organizational resources to support the implementation process. Examples given in the literature are the absence of appropriate employee training and development to prepare for the institutionalization, as well as poor communication methodologies. The latter are specified by Beer and Eisenstat (2000), who explain that difficulties related to poor vertical communication within the organization and an ineffective management team that refuses to cooperate out of fear to lose power, lead to cynicism among employees. Additionally, workplace deficiencies that hinder the execution of the innovation, such as physical workplace conditions that prevent teamwork or inflexible software that cannot be adapted, are mentioned (Schilling & Kluge, 2009, p. 354; Beer & Eisenstat, 2000, p. 32).

Then, a high employee turnover might form a second barrier, “as it causes discontinuity and disruption of the organizational memory” (Schilling & Kluge, 2009, p. 354). Learning results are not always tangible assets such as raw materials and lessons learned are nothing that organizations can possess independent of specific individuals, as also noted by for example Argyris (1999) (Schilling & Kluge, 2009, p. 354). One could argue that the lessons learned cannot be forgotten once they are written down in an official and public accident investigation report. Following that line of reasoning, this barrier would not be interesting or relevant to this research. However, one could also imagine that the significance of the report diminishes over time when certain employees who witnessed the event itself and the publication of the report leave the organization. Newcomers cannot remember the impact or even the importance of the report. It may also be that they feel less responsible as they were not employed yet during the crisis and its aftermath. Hence, it is too ambiguous to rule this specific barrier out by stating that the lessons learned can simply not be forgotten. Therefore, a high employee turnover might be a potential explanation for why the Safety Region did not implement some of the recommendations and will thus be used as a barrier in this research.

C. A lack of management skills

While the second block of barriers hinders the employee’s skill to implement structural and process innovations, the third block constitutes barriers arising out of a lack of management skills on part of the management team. Top management might hinder organizational learning “by not making necessary changes in policies and practices that are needed to implement new routines” (Schilling & Kluge, 2009, p. 354).

The first barrier arises out of a laissez-faire management style, which prevents systemic implementation as it assumes that the innovations or lessons will somehow find their way into organizational practice. In addition, inadequate down-the-line leadership skills prevent innovative

(17)

Page 16 of 81

structures, processes and practices from coming to life (Schilling & Kluge, 2009, p. 354; Beer & Eisenstat, 2000, p. 32).

Another characteristic of this leadership style is a lack or diffusion of responsibility among members of the organization, which is assumed to lead to a lack of personal involvement. This prevents an innovation from being implemented successfully and is therefore regarded as a barrier (Schilling & Kluge, 2009, p. 354).

These leadership deficiences are related to a number of personal and structural or organizational barriers. On the personal level, the manager’s perception of the skills and willingness of employees to implement innovations and low levels of trust towards them reduces the expectation of the management team concerning the success of the implementation. Therefore, the management’s commitment to institutionalize the innovation deminishes (Schilling & Kluge, 2009, p. 354).

The last barrier of this block originates from negative experiences or even conflicts during past learning transfers, as these can also decrease the commitment to implement innovations on the part of the management team (Schilling & Kluge, 2009, p. 354).

D. A lack of compliance

Lastly, the fourth group of barriers hinders the adoption of the innovation, thereby blocking the learning process of institutionalization. One of the barriers arises when organizational units and their members reject the innovation by acting counteractive and showing opportunistic behavior (Schilling & Kluge, 2009, p. 355). Kim (1993) explains that employees or teams might purposely try to bypass the standard procedures of an organization by defying the organization’s widely shared mental models (values and cultures), as these individuals can have aspirations that are not compatible with the new innovations they are supposed to implement, therefore seeing them as a threat that needs to be counteracted (p. 46; Schilling & Kluge, 2009, p. 355).

Two factors can aggravate this counteractive behavior. First, a negative attitude or cynicism towards change in general reduces the openness of employees to new ideas. Furthermore, contradictions in the innovation such as inconsistencies between the initial goals of the innovation and success criteria to evaluate it reduce the employee’s willingness to implement an innovation (Schilling & Kluge, 2009, p. 355).

Secondly and lastly, a high level of decentralization in the organization proves to be another barrier in implementing innovative ideas and keeping them alive, as departmental structures divert the attention of their members away from organization-wide problems (Schilling & Kluge, 2009, p. 355).

(18)

Page 17 of 81

By now, it has become clear that the literature offers a wide array of barriers that hinder the organizational learning process in the ‘Institutionalization’ phase. Schilling and Kluge (2009) grouped the barriers into four blocks. The first concerned barriers related to the lack of trust in the innovation itself. The second is constituted by barriers that create a deficiency in skills and knowledge necessary to adopt the innovation. The third block is related to a lack of management skills and the fourth block to counteractive behavior from the organizational members and units that results in the rejection of an innovation.

To conclude, the work of Schilling and Kluge (2009) is of value to this research for a significant reason. It categorizes the barriers mentioned in the literature more systematically and provides a clear theoretical foundation for them, thereby creating the thorough theoretical framework for which Berthoin Antal et al. (2003) provided a basis. Schilling and Kluge (2009) listed the identified barriers to organizational learning in tables for each of the four processes of the ‘4I model’. This allows for selecting the barriers applicable and of relevance to this research, which are those from the ‘Institutionalization’ phase.

So far, the focus has been on barriers to organizational learning originating in the organization itself. This can be explained by the fact that the model by Schilling and Kluge (2009) is based on an organizational learning process that takes place entirely in the organization itself. However, a significant detail about this whole endeavor is that an external party, the Inspectorate, goes through the process of acquiring the knowledge and lessons. This knowledge is subsequently published in the form of an accident investigation report and it is the task of the Safety Region to, in a way, learn that knowledge and institutionalize it in the organization by means of implementing the recommendations made by the Inspectorate. Therefore, as the reports from the Inspectorate are a vital aspect in this research, it is necessary to look at literature that deals specifically and more thoroughly with this type of learning and the potential barriers that arise from learning from accident investigation reports in the next section of this chapter.

2.2 Accident investigation reports and public inquiries

Even though the words ‘accident investigation report’ and ‘public inquiry’ have been mentioned several times already, it seems appropriate to provide a little more context to the concept. While it is hard to generalize about the public inquiry, usual characteristics are impartiality, data collection and analysis, and public reporting. Its central aim is the collection of evidence and fact-finding and the report is considered a step towards action and ultimately the raising of standards. In the case of disaster investigations, an additional purpose is added: learning lessons for the future (Elliott & McGuinness, 2002, p. 15). Efforts to identify lessons from failure and to avoid their repetition are institutionalized in the form of the public inquiry (Elliott, 2009, p. 157). With regards to the

(19)

Page 18 of 81

investigation of disasters, public inquiries have a long history and go back already to the eighteenth century. Inquiries are triggered by high profile disasters that raise fundamental questions about a breakdown in existing arrangements or systems. They provide an opportunity to make sense of complexity and uncertainty and are a step in fulfilling the desire to learn the lessons from a crisis event (Elliott & McGuinness, 2002, pp. 15-16).

2.2.1 Fantasy documents?

While the above seems promising in terms of organizational learning, the literature appears less convinced of the effectiveness of accident investigation reports on learning. In general, many authors express quite a pessimistic view on effective learning from a crisis investigation.Elliott and McGuinness (2002), Elliott (2009) and Birkland (2009) all have their reservations, of which the latter is the most negative.

Birkland (2009) calls accident investigation reports ‘fantasy documents’, as they really focus on ‘lessons observed’ instead of actual ‘lessons learned’. He furthermore elaborates that ‘fantasy learning documents’ predominantly serve rhetorical purposes, reassuring that the authoritative actor has ‘learned its lessons’ and will not replicate its errors (Birkland, 2009, pp. 146-147). The model of event-related policy learning he later on proposes, takes a political perspective, focusing mostly on agenda setting and group mobilization, rather than on the role of investigation reports specifically. While these observations are interesting, other authors present barriers to learning from accident investigations that are more specified and take an approach closer to the one taken in this research. It is those barriers that this chapter turns to now.

2.2.1.1 Criticisms, criticisms, and criticisms

Elliott and McGuinness (2002) have considered the efficacy of the public inquiry as a tool for learning from disaster and critique three broad areas: the process, the underlying aims, and the impartiality of the public inquiry process (p. 14). They present barriers for each of the three areas, which will be addressed subsequently.

First, the critiques on the process are particularly interesting and serve as barriers to learning from an investigation report. The first barrier is related to the scope of the investigation. Elliott and McGuinness (2002) explain that many public inquiries focus on micro issues and the immediate circumstances that caused the crisis. Wider organizational, political, social, and economic contexts are not considered. This may lead to either too broad or too narrow recommendations that often deal with technical issues to the exclusion of all else. This is seen as a fundamental weakness of the public inquiry since disasters are socio-technical failures (Elliott & McGuinness, 2002, p. 18). In addition, the public inquiry often focuses upon one incident, thereby restricting organizational learning to the implementation of

(20)

Page 19 of 81

specific recommendations. Learning may be encouraged when one or more similar failures or incidents are taken into consideration. Related to this is the so often ‘left censored’ scope of the investigation, which focuses on learning from failure rather than success (Elliott & McGuinness, 2002, p. 18; 20). Another barrier that is presented is the lack of a legal requirement to act upon the findings in the accident investigation report. Studies have indicated that even where regulation is introduced, compliance does not follow quickly or easily and it is unlikely that organizations will comply without the force of law (Elliott & McGuinness, 2002, p. 19).

The reports of the Inspectorate indeed focus on one incident, the power outage, and it is also true that there are officially no legal obligations to act upon the recommendations. Therefore, these two barriers presented by Elliott and McGuinness (2002) are useful for this research, as they might explain (in part) why the Safety Region did not implement all of the recommendations made.

Subsequently, the critiques on the impartiality of the public inquiry are discussed. These are however less relevant, since the critiques focus on the government appointing members from the judiciary to the investigation team and appointing individuals in general. Among other the extent to which judicial neutrality can exist given innate prejudice regardless of whether it arises from a particular judge’s background or from the rigidity of the process in which they are trained is questioned (Elliott & McGuinness, 2002, p. 17). The government does not appoint anyone to the investigation team of the Inspectorate, as this is done internally, and therefore these criticisms do not apply and cannot serve as a barrier in the context of this research.

The same goes for the criticisms on the third and last broad area – the underlying aim of the inquiry. Elliott and McGuinness (2002) explain that public inquiries may scapegoat and apportion blame to individuals or human error, which “is a key impediment to organizational learning” (p. 20). However, the specific reports from the Inspectorate referred to in this research do not focus upon individuals or human error and the recommendations are made to organizations in general, such as the Safety Region(s) and telecom providers.

2.2.1.2 Theory versus practice

In one of Elliott’s (2009) later works, several observations are made concerning the public inquiry process. Certain observations are rather similar to the ones discussed by Elliott and McGuinness (2002), such as those related to the often narrow scope of the investigation and its findings. They are therefore not repeated. Moreover, they are subordinate to Elliott’s (2009) main argument, which provides a valuable contribution to the list of barriers in this research.

(21)

Page 20 of 81

Elliott (2009) also presents organizational learning as a process, in which three stages are distinguished. The public inquiry itself is positioned in the first stage of knowledge acquisition. Subsequently, the stages of knowledge transfer and knowledge assimilation follow respectively. The main problem is that not enough attention is devoted to actually effectuating change and bridging the gap between policy and practice during the knowledge transfer stage. Implicitly, it is assumed that new guidelines or regulations emerge from a public inquiry and naturally flow into practice. Moreover, public inquiries appear to make specific recommendations without fully comprehending how these should be implemented (Elliott, 2009, p. 161;163). As put by Elliott (2009), “the central problem is that inquiries identify lessons to be learned, make recommendations and prepare guidelines and regulators regulate, but often to little avail or impact” (p. 165). He therefore recommends that the range of experts conducting the inquiry should include individuals with an understanding of how to develop and sustain a context for change at both field and organization levels.

Paradoxically, the literature that specifically deals with organizational learning from accident investigation reports and public inquiries has shown that certain characteristics of these reports could function as a barrier to learning from crisis events, which is contradictory to the original purpose of these documents. Three additional barriers have been identified as being relevant to this research. The first concerns the scope of the public inquiry and its findings. The lack of legal requirements for the organization to act upon the findings constitutes a second barrier to organizational learning. The last barrier is explained by Elliott (2009), stating that the public inquiry does not pay enough attention to ensuring the translation of knowledge to the organizational or practical level, which hinders the implementation of knowledge.

2.3 Theoretical framework on barriers to organizational learning from accident

investigation reports

In the next subdivision, the barriers identified in the literature on organizational learning and learning from accident investigations and public inquiry processes will be combined into a comprehensive theoretical framework, provided that they are relevant to this specific research.

Drupsteen and Guldenmund (2014) explain the necessity of creating such a theoretical framework that combines the barriers coming from two different, although related, strands of literature. In comparison with the literature on organizational learning, aspects about the information to learn from incident and analysis are much more specified in the safety literature, while the organizational learning theory gives more details about sub-processes such as “learning lessons, sharing, storing and applying lessons (…) and factors that influence these processes are listed (…)” (Drupsteen & Guldenmund, 2014, p. 81). Elliott and McGuinness (2002), who stress that the public inquiry process is broader than solely capturing and publishing the acquired knowledge, also underline this. When seeing organizational

(22)

Page 21 of 81

learning as a process, the first step of knowledge acquisition by means of an accident investigation is followed by a second step that concerns the means by which the knowledge is transferred and disseminated and a third step that concerns the organizational conditions that influence the receptiveness to knowledge and changes to operating norms and practices (Elliott & McGuinness, 2002, pp. 14-15). The literature on learning from accident investigation reports and the public inquiry emphasize that these reports are a means of knowledge acquisition, thereby focusing solely on step one of the public inquiry process (Elliott & McGuinness, 2002, pp. 14-15). However, in this research, the second and third step are accounted for by looking at the barriers to organizational learning. Hence, combining the barriers from the different strands of literature on organizational learning provides a fuller picture, which justifies and legitimizes the theoretical framework below, as it is used in this research as a tool to find an explanation for why the recommendations made by the Inspectorate have not been adopted.

As mentioned above, only barriers relevant to this research are incorporated in the theoretical framework. Barriers are considered to be relevant when they are able to provide a possible explanation for why the solutions and recommendations made by the Inspectorate are not acted upon and implemented by the Safety Region Amsterdam-Amstelland. Barriers that hinder identifying the problems and causes of a crisis and therefore hinder the learning process are thus not applicable, as the investigation reports from the Inspectorate already identified the lessons that need to be learned, as well as the solutions and recommendations. Hence, it is logical to select those that focus on impeding the implementation stage or institutionalization stage.

Table 1 on the next page visualizes the theoretical framework on barriers to organizational learning from crises and accident investigation reports. It presents two sets of barriers, distinguished by type. The first set constitutes the internal barriers that emerge out of the weaknesses within the organization itself. Schilling and Kluge (2009) provided a thorough overview of such internal barriers, which will be adhered to in this research. Their work thus serves as the main theoretical foundation for this framework and the preceding subdivisions have shown and justified which of their barriers are appropriate to adopt and integrate into the theoretical framework that is used for this study. The second set composes the external barriers that are related to the weaknesses arising out of the characteristics of an accident investigation and the resulting report, which plausibly hinder organizational learning. These barriers come from the literature on learning from accident investigation reports and the public inquiry process. The barriers presented by Elliott and McGuiness (2002) serve as the basis for this set of barriers and the ones selected for this research are appropriate for reasons discussed in the sections before.

(23)

Page 22 of 81

Table 1. Theoretical Framework on Barriers to Organizational Learning from Crises and Accident Investigation Reports

Barriers to organizational learning from crises and accident investigation reports

Source

Internal: Barriers within organizations

Barriers to implement (technological) innovations (A):

1. A lack of trust in the innovation

Barriers that hinder the employee’s skill to implement structural and process innovations (B):

2. A lack of time, space and organizational resources

3. A high employee turnover

Barriers that hinder institutionalizing learning experiences due to a lack of management skills (C):

4. A laissez-faire management style 5. A lack of clear (personal) responsibilities 6. A lack of trust in the skills and willingness of the employees

7. Experience of conflicts during past learning transfers

Barriers that lead to the rejection of the innovation by organizational units and their members (D):

8. Counteractive behavior 9. A high level of decentralization

(Schilling & Kluge, 2009)

External:

Barriers to learning from accident investigation reports

10. The (narrow) scope of the accident investigation and its findings and recommendations

11. A lack of force of law and legal requirements to act upon findings

12. A lack of attention to effectuating change after report is published

(Elliott & McGuinness, 2002) (Elliott, 2009)

(24)

Page 23 of 81

3. Methodology and case description

A review of the literature on organizational learning and learning from accident investigation reports has shown that there are multiple barriers that could hinder the organizational learning process. These are merged into a theoretical framework in order to understand and assess what barriers influenced the Safety Region’s learning process. The following chapter will discuss the methodological framework underpinning this research and describe the case under investigation. The choices for the research design and case selection will be explained and justified, as well as the way in which data is gathered. Lastly, the validity and reliability of this research is reflected upon.

3.1 Research design

This study makes use of a single case study design, taking a process tracing approach to qualitatively study the Dutch Safety Region Amsterdam-Amstelland and its (hampered) learning process. This type of design provides the possibility to analyze a case at much greater depth, so that the imagined theoretical explanatory logic and the proposed detailed causal mechanisms become visible (Toshkov, 2016, p. 291). It is important to focus on researching causal mechanisms rather than outcomes, as the outcome in this case is already fixed and identified: non-learning by the Safety Region Amsterdam-Amstelland. Hence, a single case study design allows for visualizing the analytical story that theoretically is supposed to connect a set of variables and bring evidence on how the mechanisms operate in reality (Toshkov, 2016, p. 291).

The choice for this research design is furthermore appropriate because, as nicely described by Broekema et al. (2017), it does “justice to the complexity of the process of organizational learning in relation to crisis, with (potentially) multiple factors at play which are strongly embedded in the specific crisis context” (p. 331). This is similar to this research, as the crisis context is important and multiple barriers might have influenced the learning process in a negative way. Therefore, it is the quest to “derive inductively from the case information about possible causal ideas that can account for the particular case but also become the building blocks of new theories or the scaffolding for extensions of existing ones” (Toshkov, 2016, p. 292).

3.2 Variables

Toshkov (2016) defines a variable as something that, quite literally, must be able to vary or, in other words, to take different values. A distinction can be made based on their role in theoretical models, such as the theoretical framework presented in Table 1 at the end of Chapter 2. The variable that captures the outcome the researcher is interested in describing, explaining or predicting is called the ‘outcome variable’ or ‘dependent variable’, whereas the variable that captures the main hypothesized causal factor is called the ‘explanatory factor’ or ‘independent variable’ (Toshkov, 2016, p. 115).

(25)

Page 24 of 81

In this research, the purpose is to explain, rather than describe or predict, the outcome and thus the dependent variable. As described in the beginning of Chapter 2, the already determined outcome that needs to be explained is the non-learning of the Safety Region Amsterdam-Amstelland. The Inspectorate of Justice and Security demonstrated in their accident investigation report on the power outage in 2017 that the Safety Region did not implement several recommendations that were already suggested in earlier reports. This observation is seen as an indicator or proxy for non-learning in this research. Hence, the outcome of non-learning constitutes the dependent variable, which is thus fixed and determined.

The independent variable can, as its synonym already suggests, potentially explain the outcome or dependent variable. Each barrier serves as an explanatory factor that captures a hypothesized causal relation. In other words, each barrier can possibly explain why the Safety Region Amsterdam-Amstelland did, repeatedly, not implement the recommendations made by the Inspectorate. Hence, there are as many independent variables in this research as there are barriers.

3.3 Case selection

The case under study is the Safety Region Amsterdam-Amstelland. The reason for selecting this case is twofold. First of all, it provides for the unique opportunity to look into the learning process of a public organization in the Netherlands that experienced similar incidents with recurring problems, while having the lessons to learn identified by the Inspectorate of Justice and Security and therefore the opportunity to learn from crises and accident investigation reports. The Safety Region Amsterdam-Amstelland has experienced several failures or malfunctions in the reachability and continuity of its emergency control room during several incidents of which the large-scale power outages in 2015 and 2017 are the most significant. As one can see, the incidents occurred in a limited period of time, long enough to allow for learning, while short enough to keep other factors that could play a role, such as a completely different organizational culture, constant. The Inspectorate of Justice and Security has conducted research into all incidents and published their findings and recommendations publicly. However, as explained before, the report on the power outage in 2017 stated that the same vulnerabilities were detected as in previous investigations, and that several recommendations already made in earlier reports have not been adopted and implemented by the Safety Region Amserdam-Amstelland. The succession of similar incidents and the subsequent accident investigation reports thus allows for determining that non-learning has taken place within the Safety Region Amsterdam-Amstelland. As explained before, the outcome of the 2017 report is seen in this research as a form of non-learning, without attaching a normative value to this observation. In sum, the circumstances and uniqueness of this case almost resemble a laboratory experiment, however now in the ‘real world’,

(26)

Page 25 of 81

which makes it therefore an excellent case to research what barriers hinder organizational learning from crises and accident investigation reports.

Secondly, the Safety Region Amsterdam-Amstelland is at an advanced stage in the learning process. Keeping Elliott’s (2009) stages of the public inquiry process in mind, the phase of knowledge acquisition is accounted for since the Inspectorate already identified the lessons learned and recommended solutions in order to eradicate the observed vulnerabilities. What follows are the stages of knowledge transfer and knowledge assimilation. Research into the Safety Region Amsterdam-Amstelland could provide valuable insights into what factors impede transferring and implementing the knowledge gained from accident investigation reports. As argued by Elliott (2009), studies on learning from crisis at the policy or the organization level have given little attention to the processes of knowledge transfer and knowledge assimilation (p. 161). Hence, selecting the Safety Region Amsterdam-Amstelland as a single case in this research is justified as it is able to contribute to the existing literature.

One must note that the case itself also constitutes the unit of analysis in this research. The level of analysis is the level at which the analysis is performed and, importantly, at which the conclusions are pitched (Toshkov, 2016, p. 116). The Safety Region Amsterdam-Amstelland is analyzed as an organization and the conclusions on what factors contributed to the non-learning of this organization are thus directed at the organizational unit Safety Region Amsterdam-Amstelland.

3.4 Case description

The next section provides more background and context to the case studied in this research. First, the Safety Region Amsterdam-Amstelland is discussed. Hereafter, the focus shifts to the accident investigation report of 2017, as this report is central to this research since it shows the Safety Region’s non-learning by stating the earlier made recommendations that have not been implemented. The work of the Inspectorate of Justice and Security is shortly reflected upon too, as they wrote the accident investigation report. Logically, the report is about a specific incident, which is in this case the power outage in 2017. Therefore, the incident itself will be elaborated upon as well to gain a better understanding and a fuller picture of the context surrounding this case. Lastly, the actual findings and recommendations made by the Inspectorate relevant to this research will be discussed.

3.4.1 The Safety Region Amsterdam-Amstelland

The Safety Region Amsterdam-Amstelland is taken as a case study in this research. But what exactly is a Safety Region and what are its tasks and powers? This section provides an answer to this question and, importantly, shows that the Safety Region is an organization and can therefore engage in organizational learning.

(27)

Page 26 of 81

The ‘Wet Veiligheidsregio’s’ of 2010 divides the Dutch territory in twenty-five Safety Regions. The law was a response to the increasing need for a multidisciplinary collaboration and comprehensive approach in the field of crisis management and disaster response, due to new and complex types of threats that transcend the local municipal level. Examples are terrorism, as well as a threatening flu epidemic and large-scale disasters such as the fireworks disaster in Enschede in 2000 (Ministerie van Veiligheid en Justitie, 2013, p. 5).

The law requires the municipalities belonging to one of the twenty-five regions to engage in a partnership with the relevant (emergency) services in the field of crisis management and disaster response. Each region sets up a ‘Gemeenschappelijke regeling’ or common arrangement, which serves as a legal basis for the aforementioned partnership and thereby establishes a public body called ‘Safety Region’. By means of Article 9 of the Wet Veiligheidsregio’s, the tasks and powers laid down in the Wet

Veiligheidsregio’s are transferred to and become the responsibility of the general management board

(Safety Board) of the Safety Region, consisting of the mayors of the participating municipalities (Wet Veiligheidsregio's, 2010). In addition, each Safety Region has a daily management, which is in the hands of an executive director and his or her staff (Veiligheidsregio Amsterdam-Amstelland, 2016). This will be discussed later on, as this subdivision first turns to the common arrangement that constitutes the Safety Region Amsterdam-Amstelland specifically.

In this case, the board of the Safety Region Amsterdam-Amstelland consists of the mayors from the following municipalities: Aalsmeer, Amstelveen, Amsterdam, Diemen, Ouder-Amstel and Uithoorn. The common arrangement requires them to meet at least four times a year, together with the Commander of the Fire Brigade, the Director of Public Health, the Police Chief of the Regional Unit Amsterdam, the coordinating officer of the Safety Region, as well as other officers whose presence is important in relation to the subjects to be discussed (Veiligheidsregio Amsterdam-Amstelland, 2016). The board is invested with the following tasks, as laid down in Article 10 of the Wet Veiligheidsregio’s: a). the inventory of the risks of fires, disasters and crises;

b). advising the competent authority on the risks of fires, disasters and crises in the cases designated by or pursuant to the law and in the cases determined in the policy plan;

c). advising the mayor and aldermen on the fire brigade service;

d). preparing for combating fires and organizing emergency response and crisis management; e). setting up and maintaining a fire service;

f). setting up and maintaining a GHOR (medical assistance organization); g). providing a functioning emergency control room;

(28)

Page 27 of 81

i). the establishment and maintenance of the information provision within the services of the Safety Region and between these services and the other services and organizations involved in the tasks mentioned under d, e, f, and g.

By executing these tasks, the Safety Region Amsterdam-Amstelland provides for a decisive crisis management, fire brigade service and medical assistance with the aim of anticipating risks as well as possible, helping those affected as well as possible, limiting material damage and returning to a normal situation as soon as possible3_.

Additionally, Article 17 of the common arrangement allows the board to establish a Safety Bureau, which serves as the civil organization that is responsible for managing the multidisciplinary regional cooperation in the field of crisis management. To this end, the Safety Bureau will support and advise the Safety Board, as well as facilitate and coordinate multidisciplinary tasks and projects (Veiligheidsregio Amsterdam-Amstelland, 2016). The Safety Bureau thus functions as the executive branch of the Safety Region Amsterdam-Amstelland.

Note here that the fourth chapter of the common agreement itself constitutes and speaks of ‘the civil

organization’. Furthermore, Article 3 of the same document states that there is a legal entity

possessing a public body called Safety Region Amsterdam-Amstelland, which is located in Amsterdam. This indicates that the Safety Region Amsterdam-Amstelland can be seen as a legally competent and full-fledged organization, which means that it should be able to engage in organizational learning, as explained by Argyris (1999). This is an important observation and necessary to make, as this thesis focuses on the question why the Safety Region Amsterdam-Amstelland did not complete its learning process by not implementing several recommendations made by the Inspectorate. If it cannot be demonstrated that the Safety Region Amsterdam-Amstelland indeed functions as an organization that can learn, there would be no point in researching why it did not.

3.4.2 Accident Investigation Report: ‘Onderzoek naar de stroomstoring Amsterdam en

omstreken 17 januari 2017 – Bereikbaarheid en continuïteit van de meldkamer’

It is now time to take a closer look at the accident investigation report published by the Inspectorate of Justice and Security in response to the large-scale power outage in January 2017, since it plays a pivotal role in the organizational learning process from crisis in this research.

To provide a little more context and background to the report itself, the institution that wrote the relevant report (the Inspectorate) will be discussed briefly first, as well as the incident (the large-scale power outage in 2017) that the report is about. Lastly, the findings and observations made in the report

3_{Website Veiligheidsregio Amsterdam-Amstelland, Organisation}