Accidental Partners: A Comparative Analysis of Learning as a Form of Accountability in Dutch Accident Investigations

(1)

Accidental Partners:

A Comparative Analysis of Learning as a Form of Accountability in Dutch Accident Investigations

Research Question

To what extent did investigations by multiple forums into large accidents in the Netherlands affect learning as a form of accountability?

Abstract

Accountability is a concept that is often used in the context of ideas such as good governance and virtuosity. Accountability’s most important role however, is arguably its contribution to organizational learning in both public and private spheres and it is in this sense that accountability may be evaluated. In this thesis, accountability was conceptualized as the mechanism by which organizations are induced to learn from failures. Three cases were selected in which either the Inspectie Openbare Orde en Veiligheid, IOOV, and the Onderzoeksraad Voor Veiligheid, OVV were both present, or only one was present. These cases are the 2009 Poldercrash at Schiphol, the 2010 Bushfire in the Strabrechtse Heide, and the 2014 Monster Truck accident in Haaksbergen. Using the absence or presence of key learning objectives, the lessons to be learned and recommendations made by the IOOV and OVV were evaluated. Investigations by both the IOOV and OVV were found to exhibit less variance between indicators and higher total scores than the investigations by only one forum. It was therefore concluded that investigations by multiple forums affect learning as a form of accountability in a positive manner and to a slight extent.

Program: MSc. Crisis & Security Management Author: Bernard Schalk Jansen

Student Number: s1604821

Date of Submission: 10th of June 2018 Word Count: 18124

Thesis Supervisor: Dr. S. L. Kuipers Second Reader: W.G Broekema, MSc

(2)

2

Acknowledgements

Without the help and guidance from Dr. Sanneke Kuipers, this thesis would never have materialized from the pandemonium that is the author’s mind. The author wishes to express his everlasting gratitude to Dr. Kuipers for all her assistance and hopes that they were not all too much trouble.

The author also owes a debt of gratitude to all the other participants of the capstone on Learning from Crisis and Safety Investigations as well as Wout Broekema and Michael Kowalski for their comments and contributions to the roundtable discussions.

The author would also like to express their gratitude to their family and girlfriend, without their undying love and support, this thesis would not have been possible.

So Long, and Thanks for All the Fish

(3)

3

Introduction

Investigative Proliferation

One of the many reoccurring issues in investigations following accidents is that once an investigation is announced, begun, or otherwise announced to the general public, there is a large chance that a number of other actors may also begin an investigation, thereby creating a proliferation of investigative efforts. Indeed, an organization that finds themselves being held accountable by a governmental investigative organization will often feel a need to perform their own internal investigation. The Dutch Onderzoeksraad Voor Veiligheid, or Dutch Safety Board, identifies this problem in their 2018 publication on independent investigation (OVV 2018).

Pointing to the 2011 fire in Moerdijk at Chemie-Pack as an example of a proliferation of investigative efforts, the OVV notes that such a great number of separate investigations ultimately led to a highly questionable division of labor between the various investigative agencies as well as a lack of cohesiveness. The OVV does acknowledge however, that in specific cases multiple investigations may in fact be beneficial. For example, ins some cases the organization responsible for an incident performs their own investigation in order to identify measures that may prevent a repetition of the incident whilst an independent investigative organization performs a simultaneous investigation to identify the wider lessons that may be learned from the incident (OVV 2018, pp. 46-48).

The Netherlands is host to a number of investigative agencies. With respect to general safety and security, the Inspectie Openbare Orde en Veiligheid or the Inspectorate for Public Order & Safety reports on the annual state of affairs in the Dutch safety regions, as well as inspecting threats to public safety when requested, such as when the Kennemerland Safety Region and the municipality of Haarlemmermeer requested an investigation into the Turkish Airlines crash at Schiphol in 2009. The independent OVV however, has a strong mandate to conduct an investigation on any crisis it selects as being important or relevant for learning. However, the OVV may also be approached by public bodies such as the police or a municipality to conduct an investigation. These two organizations form the main investigative organizations following accidents in the Netherlands. It should be noted however that as of January 2012, the IOOV was restructured into the Inspectie Veiligheid en Justitie, or

(7)

7 Inspectorate for Safety & Justice (IJenV 2017, p. 4; IOOV 2009, p. 4; Kuipers & Boin 2014, p. 25; Wijkhuis & van Duin. 2012, p. 31).

This problem with investigative proliferation is effectively a problem with redundancy. If it is indeed beneficial to have multiple investigations, then this redundancy may have positive effects on the eventual outcome of the investigative effort as a whole in terms of the lessons that may be learned. If however, such a multiplicity of investigations does not lead to organizations achieving the desired outcomes, the need for multiple investigations into a single crisis in the Netherlands by the OVV and the IOOV becomes somewhat unclear. (Bovens & Schillemans 2011, p. 26-27).

Moreover, whilst the primary function of inspectorates and other governmental bodies concerned with supervision and evaluation is to learn from incidents and crises and to provide an account of the crisis or incident in question, these goals may sometimes form the opposite ends of a spectrum. Indeed, if the primary goal of such an investigative effort is to hold those involved in the process being evaluated accountable for flaws in regulations or in the execution of routines, a conflict of interest may arise if those being evaluated are allowed to unduly influence or interpret the data being collected. As both the OVV and IOOV differ in this regard, a comparison on the effects of their presence, or combined presence, on accountability from accident investigations is warranted (Foster, Hartley, & Leslie 2017, pp. 64-66).

From A Count to Accountability

As Bovens (2007) identifies, investigative practices that focus on learning in order to improve the effectiveness of an organization can be seen as process of accountability. According to Bovens (2007), processes of accountability provide public office holders and other agencies and organizations with feedback-based inducements to increase their effectiveness and efficiency regarding the achievement of socially desirable outcomes. This occurs through interaction by the agent, the organization or individual being investigated, with the investigative forum, the organization or agency performing the investigation. The word accountability finds its etymological origin in the Norman conquest of England in 1066, when William I ordered a count to be made of all property possessed by his vassals in 1085. Whilst extremely useful for taxation purposes, such censuses also enabled the foundations of royal governance. Since then, broad conceptualizations of accountability have formed it into

(8)

8 somewhat of an umbrella term. Indeed, accountability is often used interchangeably with concepts such as virtuosity and good governance. Most importantly however, accountability, and the processes by which it is rendered, forms a considerable contributor to organizational learning in both public and private spheres (Bovens 2007, pp. 448-450; Bovens 2010, pp. 948-951; Schillemans & Smulders 2015, p. 248).

As such, this thesis seeks to explore the relationship between accountability processes and organizational learning using case studies into accident investigations in the Netherlands. By comparing the investigative efforts by both the OVV and IOOV in the Poldercrash case, to those of the OVV and IOOV in the cases of the Haaksbergen Monster Truck Demonstration and the fires at the Strabrechtse Heide respectively, the effects of the number of investigative forums on accountability after accidents will be identified, compared, and evaluated.

Research Question

To what extent did investigations by multiple forums into large accidents in the Netherlands affect learning as a form of accountability?

Societal & Academic Relevance

As a result of the persistent threat of water to its territorial integrity, the Netherlands has a longstanding history of dealing with crises and has consistently chosen to invest in disaster mitigation. However, crisis such as the aforementioned fire in Moerdijk have shown that the Dutch civil security system is not without its faults. Issues such as the coordination and communication difficulties between regional and national authorities, as well as the ill-defined role of the central government when a crisis starts at a local level, have revealed that the Netherlands is by no means a perfect example of crisis management. Moreover, as Wijkhuis & van Duin (2012) identify, the Netherlands is characterized by its large number of investigative efforts following a crisis, often ranging between 5 to 10 reports for a single incident, indicating the possibility of evaluation overloads. As such, this thesis seeks to provide a much needed framework for comparing the process of accountability following such crises in the Netherlands, and assess whether the Dutch tendency to produce multiple reports is indeed expedient (Kuipers & Boin 2014, pp. 2, & 24-28; Wijkhuis & van Duin 2012, p 50).

(9)

9 Moreover, it should be noted that the provision of safety and security is one of the core tasks of the democratic state. Independent external assessment of crises and the response efforts to crises by civil society, inspectorates, or even academics can provide the necessary feedback policy makers need on the legitimacy and efficacy of their policies. In this manner, accountability provides the mechanism by which policy makers may learn from crises and subsequently adapt their policies. Furthermore, the public nature of the investigative process after disasters forms a valuable source for learning for those individuals in similar positions as those being held to account in an investigation (Bovens 2007, pp. 463-464; Bovens 2010, p. 954).

Thesis Structure

This thesis is structured by chapter. In the first chapter, the problem to be investigated is introduced, along with the main theoretical concept of accountability. Moreover, the research question as well as the societal and academic relevance are presented. In the second chapter, the relevant theoretical literature for constructing a set of indicators to study accountability process as organizational learning is given. In the third chapter, the research design and method chosen to answer the research question is elaborated upon. Next, the operationalization of the concepts in the theoretical framework is propounded. In the fourth chapter, findings from the four different case studies are presented. These are subsequently analyzed in the fifth chapter to ascertain to what extent the actors held accountable were stimulated to learn. The conclusions that may be drawn from the case studies and the results and limitations of this analysis and investigation are also discussed in the fifth chapter. Finally, some tentative recommendations based on the conclusions drawn are given. The works cited to construct this thesis may be found in a separate section at the end.

(10)

10 Chapter 2

(11)

11

Theoretical Framework

In this theoretical framework, a conceptual discussion on accountability is firstly presented, after which the relationship between learning and accountability is discussed. Finally, the key objectives of organizational learning as well as the conditions that inhibit organizational learning in accountability processes are propounded. These objectives are operationalized into indicators for learning as a form of accountability in the subsequent chapter on methodology.

Event and the Day After

The contemporary Marxist philosopher Slavoj Žižek (2014) describes an event as “something shocking, out of joint, that appears to happen all of a sudden and interrupts the usual flow of things” (Žižek 2014, p. 2). As Žižek (2014) deliberates, events may come in the form of celebrity scandals, the introduction of new forms of cinema, the triumph of the people over a brutal military dictatorship, a chemical fire, a plane crash, or in the shape of devastating natural phenomena such as hurricanes. Events and crisis in particular, may not be so much things that occur within the world we collectively inhabit, but rather a more fundamental change in the frame through which we perceive and engage with the ever-more interconnected world around us (Žižek 2014, pp. 21-32).

Whilst disastrous events may destroy decades of accumulated capital and development, they also open up opportunities in the period following the event. It is in this ‘day after’ period after the event when learning occurs, and when those held responsible must account for what has come to pass (Wilkinson, Twigg, & Few 2018, p. 3).

Disambiguating Accountability

As discussed in the introduction, accountability has evolved into somewhat of an umbrella term in colloquial discourse. Accountability is often associated with concepts such as transparency, trustworthiness, and fair and equal governance. However, in order to understand how process of accountability may lead to organizational learning, an analytical and narrow definition of accountability is needed. Accountability in an analytical and narrow sense may be understood as “a relationship between an actor and a forum, in which the actor has an

(12)

12 obligation to explain and to justify his or her conduct, the forum can pose questions and pass judgement, and the actor may face consequences” (Bovens 2007, p. 450).

Under this definition, an actor is defined as an individual official, private or public institution, or an organization, whereas the forum can be defined as an individual superior or organization such as a parliament, inspectorate, or court. There are of course a number of different relationships between actors and forums. In general terms, the relationship between the actor and the forum consist of at least three constituents. Firstly, the actor must obliged to inform the forum about their conduct, through providing data on the outcomes or performances of tasks and procedures. Secondly, there must exist a possibility for the forum to question the actor on the adequacy of the information and its legitimacy. Thirdly, the forum may pass judgement upon the conduct of the actor, in both a positive or negative sense. The possibility of sanctions or other negative repercussions is the defining difference between non-committal provision of information and actually being held to account. The actor may therefore face consequences. The obligations upon the actor to do all of the above may be formal or informal. After disasters however, public officials are often under formal obligation to provide an account to specific forums, such as courts or parliamentary committees. In the case of accident investigations in the Netherlands, the IOOV and OVV form the forum to whom account is to be rendered (Bovens 2007, pp. 450-452; Bovens 2010, pp. 950-955).

Bovens & Schillemans (2011) note that there are a number of standing issues when agent are held to account to multiple forums. Firstly, there may be conflicting expectations from the different kinds of forums. In legal cases for example, a prosecution is sought, which is not necessarily the case with accountability processes within NGOs for example. Secondly, there are a number of transaction and opportunity costs associated with accountability processes, such as time, wages, and staff. An increase in the number of forums implies an increase in all of these associated costs. However, the scholars’ research into accountability processes in the Netherlands found that these issues were in fact rather moderate (Bovens & Schillemans 2011, pp. 6-11 & 30).

Organizational Learning as Accountability

Accountability may therefore be seen as a tool by which government institutions may make and keep effective policies, and as such conceptualized as a process of organizational

(13)

13 learning. The purpose of accountability is therefore to induce learning in the executive branch, with the possibility of sanctions acting as a motivator. Indeed, seen in this perspective, accountability forms a crucial aspect of pluralist democracy, in that it creates an incentive for intelligence and learning in policy-making because of the public nature of accountability processes. Ultimately, accountability forms the insurance for the citizenry, that the governing state body above them is in fact legitimate. Processes of accountability can thus promote the acceptance of government authority and confidence in the state by the citizenry. Moreover, after crisis, accountability processes help provide public catharsis, as was the case in South Africa’s TRC. Taken from this perspective, accountability can therefore be further defined as the mechanism by which actors are provided with feedback-based inducements, such as inspectorate reports, to foster organizational learning (Bovens 2007, pp. 463-464; Bovens 2010, p. 954).

Schilling & 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 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. 388). For an organization to learn, it must therefore create new rules, policies, or regulations based on its lived experience.

As Schilling & Kluge (2009) observe, the mutual dependency between individual and organizational forms of learning occurs as a result of individuals learning whilst being the representatives of their particular organization. The knowledge that they acquire in this capacity, must be stored in the form of documents, processes, or structures if it is to be retained within the organization. Organizational learning should not however, be taken as a simplistic aggregate of individual learning. Rather, organizational learning should be conceived of as a collective development that engages and changes fundamental perceptions and beliefs within the organization’s members (Greiling & Halachmi 2013, pp. 385-386; Schilling & Kluge 2009, pp. 388).

As Foster, Hartley, & Leslie (2017) observe, inspectorates and other governmental evaluator bodies often takes accountability and the generation of knowledge, colloquially referred to as learning, as their primary goals. Moreover, it is vital to note that organizational learning from accountability is a highly collaborative process, with the evaluator often taking the role of a critical friend or coach rather than an organizational superior. In practice however,

(14)

14 these two goals often form opposite ends of a spectrum. If the primary goal of an evaluation is to hold those involved in the process being evaluated accountable for flaws in the process, a conflict of interest may arise if those being evaluated are allowed to unduly influence or interpret the data being collected (Foster, Hartley, & Leslie 2017, pp. 64-66).

Organizational learning can be conceived of as occurring in the form of cycles, in which the organization continuously generates shifts in knowledge and skills based on its historical experiences. Learning may also occur at a technical level, systemic, or strategic level, and the lessons learned at each level may influence behavior and policy at the other levels. Moreover, the value of lessons learned is dependent on time and organizational learning pathologies. Such pathologies refer to those fundamental concepts upon which an organization is built. These fundamental concepts may be as simple as a common understanding on the linearity of time or more complex, such as measuring success by the absence of failure (Greiling & Halachmi 2013, p. 386).

Organizations that are successful at learning are those that manage to encode historical deductions into their daily practices and operations. However, at the time of writing there have only been empirical indications that accountability may contribute to learning, rather than on what agents learn or on when they are likely to learn. Schillemans & Smulders (2015) found that when executives do learn, those lessons are quite practical and operational in nature and that fundamentally new insights on problems were rather scarce. The scholars also found that executives were more likely to learn

when the forum they were accountable to was able to remain critical on past behaviors, whilst the forum simultaneously maintained a focus on future actions and plans (Ebrahim 2005; Greiling & Halachmi 2013; Schillemans & Smulders 2015, p 250 & 266-267).

The link between

accountability processes and organizational learning rests upon the retrospective nature of such

Anticipation Phase Information Phase Debate Phase Judgement Phase

Figure 1: Learning From Accountability Cycle

Adapted from “Learning From Accountability?! Whether, What, and When” by T. Schillemans & R. Smulders, 2015, Public Performance & Management Review, 39(1), p. 254. Copyright 2015 by the Taylor & Francis Group

(15)

15 processes. As discussed above, actors will need to provide information about their behavior to a forum, and juxtapose this behavior with standards and expectations. This process should lead to reflection, the identification of opportunities for improvement, and in turn, learning. Schillemans & Smulders (2015) make use of this retrospective flow of information to develop a cyclical model of learning from accountability, see figure 1. In the information phase, the presence of information gathering and provision routines must be present, come at approximately the halfway point of decision processes, and be set in a relatively open environment. In the debate phase, reflection and evaluation of the agent’s actions and performance must be performed by a reputable forum in an ongoing and substantive manner. In the judgment phase, the forum and actor must maintain open exchanges whilst acknowledging real political risks, such as forced resignations. Moreover, the possibility of sanctions must be strong enough to induce anticipation of such sanctions, whilst the forum and agent must simultaneously maintain a safe environment so as to minimize defensive behaviors. In the anticipation phase, the actor must demonstrate a commitment to improving their practices by focusing on outcomes, and the forum must place a greater weight on innovation and achieving goals than on control and regulation. These stages echo those present in Bovens’ (2007) conceptualizations of accountability discussed above (Bovens, Schillemans, & ‘t Hart 2008, pp. 232 & 238; Greiling & Halachmi 2013, pp. 387-389; Schillemans & Smulders 2015, pp. 249 & 253-254).

Conditions For Organizational Learning

Moreover, Schillemans & Smulders (2015) identify three aspects that form the conditions under which organizational learning will be optimized. The substantive focus of the forum forms the first condition. In a strict principal-agent relationship, learning may be curtailed as a result of its focus on controls and regulations. Moreover, under such strict foci, there is an incentive for those under review to manipulate the process in their own self-interest. However, by placing the focus of the accountability process on long-term organizational goals, organizational learning is much more likely to occur, as this aforementioned incentive is minimalized. Moreover, there is some evidence that organizational learning is more likely to occur when errors are treated as opportunities for learning rather than the absence of success (Schillemans & Smulders 2015, pp. 254-255).

(16)

16 The second condition is that of interpersonal trust. As organizational learning is to some extent the result of both formal and informal relationships and interactions between an actor and its associated forum, the way that a forum approaches and interacts with a given actor matters. As Schillemans & Smulders (2015) observe, learning is often more likely when the actor and forum are in a more informal relationship, as this will decrease tensions and the likelihood of defensiveness, and the resultant withholding of information, in the accountability process (Schillemans & Smulders 2015, p. 254-255).

Although already mentioned, the third condition for learning to occur from the accountability process is the availability of possible sanctions or other forms of punishment. Whilst there is some disagreement on whether these sanctions should be formal or informal, there is a clear indication that at least some form of sanctioning must be available to the forum. Moreover, it should be noted that the possibility of moderate sanctions might push actors to learn from critical feedback from the accountability process, whereas harsher sanctions may lead to strategic behavior such as withholding information, as discussed above (Schillemans & Smulders 2015, p. 255).

Learning Objectives for Accountability

Having established under what conditions learning from accountability may occur, the objectives of such learning must now be discussed Schillemans & Smulders (2015) identify three key objects of learning from accountability for those held to account. Firstly, it is vital for those being held to account to obtain new knowledge of what is happening inside the organization, both in operational and in strategic terms. Secondly, those being held to account must gain new skills. Finally, those being held to account must to distinguish between the lessons learned as either being operational and strategic lessons (Schillemans & Smulders 2015, pp. 255-256).

The acquisition of knowledge in this sense is not only about accumulating facts, but also about developing new and greater understandings of what may or may not happen in specific contexts. Wagenaar, Hudson, & Reason (1990) distinguish between such new forms of knowledge on what went wrong or right as being either slips or mistakes. Slips occur at a skill-based level, and are errors in the application of routine actions. In these cases, regulations are unlikely to change, as they are identified as being incorrectly applied rather than faulty in and

(17)

17 of themselves. Mistakes however, occur as a result of rule-based errors. This refers to situations in which actors believe that they have a sound policy or solution to a problem, only to find that this policy or solution is not based on a full analysis of the problem. Mistakes may also be knowledge-based, such as in cases where new and emergent problems confront actors. Wether rule-based or knowledge-based, mistakes require the adoption of new rules and procedures (Schillemans & Smulders 2015, pp. 255-256; Wagenaar, Hudson, & Reason 1990, p. 283).

Whilst the skills that exceutives may learn may be rather banal skills and resources, such as executives learning how the accountability process works and what is expected of them, they should ideally improve the organization’s capacity to tackle complex issues such as emergency response. Ideally however, organizations’ should find themselves having improved their organization’s strategies for tackling complex issues such as disaster relief as well as acquire knowledge on how to procure and mobilize the sets of skills, resources, and provisions associated with such efforts (Schillemans & Smulders 2015, pp. 255-256).

Finally, lessons must be distinguished between being at an operational level or at a strategic level. Brändström & Kuipers (2003) distinguish between incidents and crises being treated as incidents and as symptoms. When treated as incidents, and therefore operational, outcomes are often focused upon the technical and operational drivers and of crises, such as the vulnerability of particular types of chemical containers or organizational mishaps. These operational drivers are corrected through containing the crisis to the organization itself, as well as processes of scapegoating. They may also be corrected through the adjustment of operational rules and procedures, such as how emergency responders are expected to conduct radio communications. When treated as symptoms, the outcomes are often focused upon the higher-level aspects of crises, such as failures in regional safety policy or other large systems, or the failures of policy makers. Such failures invite overhauls and other forms of major policy change, whether focused upon individuals or wider systems and policies. An operational lesson to be learned may therefore be considered as those that are intended to fine tune procedures as a result of errors made. Strategic lessons may be considered as those that are intended to overhaul wider policies, such as Safety Region policy, changes to municipal policies, or the organizational structures of emergency response branches (Brändström & Kuipers 2003, pp. 302-304; Schillemans & Smulders 2015, pp. 255-256).

(18)

18 The absence or presence of these key learning objectives may therefore be used to determine to what extent the actors held to account in Dutch accident investigations were enticed to learn from accountability. If all three objectives are present, learning from accountability may be considered to be present to a greater extent than if only one or two are present.

(19)

19 Chapter 3

(20)

20

Methodology

In this chapter, the research design and method chosen to answer the research question are elaborated upon. Firstly, the comparative case study design chosen is discussed, as well as the means by which cases were selected. The process by which data will be collected and analyzed, as well as the theoretical expectations on outcome are also presented. Finally, the operationalization of the concepts discussed in the theoretical framework above is given.

Research Design

The research question to be answered in this thesis is: To what extent did investigations by multiple forums into large accidents in the Netherlands affect accountability

Accountability therefore forms the dependent variable in this research. Whereas the number of investigative forums present forms the independent variable of interest.

This research question will be analyzed through a cross-case analysis of four cases. A case, as identified by Rohlfing (2012), is “a bounded empirical phenomenon that is an instance of a population of similar empirical phenomena” (Rohlfing 2012, pg. 24). In this thesis, the empirical phenomenon being studied in each case is the investigated crisis. In each crisis, the findings, recommendations, and lessons to be implemented by the actors held to account are analyzed to ascertain to what extent account was rendered in each case using the key objects of learning as identified Schillemans & Smulders (2015), Brändström & Kuipers (2003), and Wagenaar, Hudson, & Reason (1990).

Whilst the precise natures of the crises themselves may differ, for the purposes of this thesis the crises themselves are secondary to the process of investigation and learning from accountability following the crisis itself, and as such form comparable cases. As this thesis is focused on the underlying mechanics by which actors are held to account, a case study design was deemed appropriate (Stake 1978, p.6; Stoecker 1991, p. 93).

Using three case studies, see Table 1, investigations by the OVV and IOOV will be evaluated using the key objects of learning as identified by Schillemans & Smulders (2015), Brändström & Kuipers (2003), and Wagenaar, Hudson, & Reason (1990) to qualify how the

(21)

21 presence of the OVV, the IOOV, or both investigative forums affected accountability in each case.

As this thesis seeks to qualify the effects of the number of forums on accountability in each case, a within-case comparative analysis was selected. Such an analysis calls for each individual case to be examined and evaluated independently in terms of Schillemans & Smulders (2015) learning objectives for accountability. Findings from these within-case analyses may subsequently be compared across the four different cases to establish to what extent accountability may have been affected by the presence of multiple forums (Rohlfing 2012, pp. 12-15; Yin 1981, p. 103).

Case selection was performed through selection upon the independent variable. By discriminating on the independent variable of interest, a diverse case selection was achieved. In such a case selection, cases are selected so as to represent the full possible variation upon the independent variable of interest. In this thesis, the independent variable of interest is the number of investigative forums in each case. In the Poldercrash case, both the IOOV and the OVV investigated the crises at hand, whereas in the other two cases the crisis was investigated by either the IOVV or the OVV, see Table 1 (Gerring & Cojocaru 2016, p. 300; Seawright & Gerring 2008, pp. 300-301; Rohlfing 2012, pp. 82-83).

Table 1: Cases to Be Studied

Case Investigated by

OVV & IOOV

Investigated by IOOV Investigated by OVV Schiphol Poldercrash 2009 (IOOV 2009; OVV 2010) X Strabrechtse Heide 2010 (IOOV 2011) X Haaksbergen 2014 (OVV 2015) X

(22)

22 The Netherlands was selected as it has a constitutional and legal framework that fragments responsibility for crisis management between different agencies, including the responsibility to investigate after accidents. Moreover, partially because of its history of dealing with water & crisis management, the Netherlands has a strong professional investigative culture, and incidents are thoroughly investigated. Moreover, both the OVV (2018), as well as Wijkhuis & van Duin (2012), point to the problem of accountability overloads, in the Netherlands. In this Dutch context, the IOOV and OVV are the main investigative actors after accidents. As such, cases were selected based on the presence of either the IOOV, the OVV, or the presence of both (Kuipers & Boin 2014, pp. 25 & 34; OVV 2018, pp. 46-48; Wijkhuijs & van Duin 2012, pp. 31 & 50).

The OVV is marked by its notable focus on both the direct and indirect causes of the crises it investigates. Moreover, the OVV takes learning from crises to be a key aspect of their core mission and goals, and as such, investigations are focused on preventing or improving the response to future crises rather than enforcing rules or regulations. Furthermore, the OVV’s interaction with those it holds to account after crises is rather formal. It does not have recurring structural interactions with the actors it investigates. The OVV’s investigative powers are codified in the Rijkswet Onderzoekraad voor Veiligheid (n.d.), and interactions between the OVV’s investigators and those held to account are to be along the lines stipulated in this law. Investigators may for example request insight into private and business records and make copies of these. Moreover, any persons being investigated are under a legal obligation to cooperate with the OVV under Article 40 of this law. However, under Article 69, none of the materials acquired by the OVV to produce its reports, as well as the reports themselves, can be used for legal purposes. As such, the OVV does not have a strong power to impose sanctions upon those they hold to account in their investigations (OVV n.d., pp. 3, 5, & 7; “Rijkswet Onderzoeksraad voor Veiligheid” n.d.).

The IOOV is characterized by its role as a supervisory body within Dutch policymaking, forming the crucial link between the execution of policies and policy revisions and changes. It is from this standpoint of policy supervision and regulation that the IOOV investigates crises. The IOOV’s interaction with those it holds to account is quite formal, as they are part of the Dutch Ministry of Interior & Kingdom Affairs and the wider policy process of making and reshaping policy. Whilst the IOOV has some recurring interactions with the actors it investigates, these are within the context of yearly evaluations as required by law. Moreover,

(23)

23 whilst the IOOV does not have any direct sanctioning power ascribed to it, as a result of its position within the Ministry of Interior & Kingdom Affairs, the IOOV can make use of the Minister’s sanctioning powers. Whilst indirect, this does mean that the IOOV has a considerable ability to sanction those it holds to account (IOOV 2010, pp. 5, 7, &11-14).

Schillemans & Smulders (2015) observe the extent to which actors may learn from accountability processes is dependent upon the forum to which they are held to account possessing a particular set of conditions, as discussed above. These may range from a fixation on regulations and controls in a formal environment with heavy sanctions to a focus on longer-term organizational goals in an informal environment with no sanctions available. The OVV and IOOV display variation upon two of the three learning conditions identified by Schillemans & Smulders (2015) for learning from accountability, see Table 2, and as such, their investigations and subsequent reports constitute comparable units from which the extent to which accountability was affected by the presence of each forum may be analyzed (Schillemans & Smulders 2015, pp. 253-255).

Table 2: Case Variance in Organizational Learning Conditions

Condition OVV IOOV

Substantive Focus of The Forum (Schillemans & Smulders 2015)

Focused Upon Longer-Term Organizational Goals

Focused Upon Controls & Regulations

Relationship Between Forum & Actor (Schillemans & Smulders 2015)

Formal Relationship Formal Relationship

Availability of Sanctions (Schillemans & Smulders 2015)

(24)

24 The presence of accountability forums in the form of either the OVV, the IOOV, or both the IOOV and OVV following the crisis, the presence of the learning conditions specified by Schillemans & Smulders (2015) in the investigating forum, and the country in which the investigation took place formed the scope conditions for case selection.

Data Collection & Analysis

Using document analysis, reports from the IOOV and the OVV, as well as the reactions to the reports from the forums being held to account in each case, will be analyzed using the indicators displayed in Table 3. This will enable an evaluation of each investigation in terms of its accountability outcome, either high or low.

Based on the three conditions under organizational learning is facilitated as identified by Schillemans & Smulders (2015), it is clear that the two forums to whom account was rendered in the cases vary in terms of their substantial focus, and ability to enforce sanctions, but not in their relationship to the actors held to account, see Table 2. Based on this variation, it is expected that the forums being analyzed in this thesis, the OVV and IOOV, will display similar outcomes in terms of holding actors to account.

By comparing the investigative efforts by the OVV and IOOV in the Poldercrash case, to those of the OVV and IOOV in the cases of the Haaksbergen Monster Truck Demonstration and the fires at the Strabrechtse Heide respectively, the effects of the number of investigative forums on accountability after accidents will be identified, compared, and evaluated using the set of indicators described in the section on Operationalization and displayed in Table 3.

Operationalization

Having discussed the relevant theoretical concepts in the previous section, these concepts must now be operationalized into observable indicators. Firstly, drawing from Bovens (2007) accountability is defined as “a relationship between an actor and a forum, in which the actor has an obligation to explain and to justify his or her conduct, the forum can pose questions and pass judgement, and the actor may face consequences” (Bovens 2007, p. 450).

Accountability is subsequently further conceptualized as the mechanism by which actors are provided with feedback-based inducements, such as inspectorate reports, to foster

(25)

25 organizational learning. As such, the key objects of learning as identified Schillemans & Smulders (2015), Brändström & Kuipers (2003), and Wagenaar, Hudson, & Reason (1990) may be used to determine to what extent account was rendered. Subsequently, analysis of the investigations by the OVV and IOOV to determine whether the presence of these investigative forums resulted in higher or lower amounts of accountability is enabled.

Firstly, those being held to account should obtain new understandings of what is happened within the organization. A positive indicator of this is the implementation of new procedures after old procedures were found to be insufficient, whereas a negative indicator of this is that old procedures were found to be sufficient, but simply not implemented.

Secondly, those being held to account are expected to acquire new skills and resources. A positive indicator for this is the presence of recommendations in the investigations by the OVV and IOOV on the need for new skills, for example in the form of expert knowledge. A negative indicator for this is the absence of such recommendations.

Thirdly, those being held to account must be able to distinguish between strategic and operational lessons learned. A positive indicator of this is the presence of both strategic and operational recommendations in the investigations by the IOOV and the OVV, whereas a negative indicator for this would be the presence of one but the absence of the other. The lack of any specification at all, in that it cannot be discerned if the recommendation is too implemented at a strategic or operational level, is also considered to be an indication of a negative score.

Using these concepts and indicators, an assessment of accountability in terms of low, medium and high is facilitated. If all three objectives are found to be present, accountability may be considered high. Accountability is considered as being medium if only two are present, and low if there is only one objective present. See Table 3 for an overview.

After having applied these indicators to each case study, each investigation will be qualified using these indicators to ascertain what extent the presence of multiple forums affect accountability. If similar results, in terms of high, medium, and low accountability, are present for each of the four cases analyzed, the presence of multiple accountability forums can be determined as having a negligible effect on accountability in accident investigations.

(26)

26 Table 3: Operationalized Indicators to Be Applied To Each Case

Variable: Understandings Gained (Schillemans & Smulders 2015)

Indicator Example Implementation Of New Procedures As Old Procedures Were Found To Be Insufficient

Recommendation that GRIP and VOS levels be automatically coupled rather than manually entered by centralists in the alarm center (IOOV 2009, pp. 25-28).

Old Procedures Were Found To Be Sufficient, But Not

Implemented

Failure by the municipality of Haarlemmermeer to register victims using registration cards as prescribed in Dutch policy (OVV 2010, pp. 87-89, 92, 96-98, & 120- 121).

Variable: Skills Learned (Schillemans & Smulders 2015)

Indicator Example

New Skills/Resources Recommended

Recommendation that a public information coordinator be added to the municipal ACs’ staff (IOOV 2011, pp. 97-100).

No New Skills/Resources

Recommended

Despite the identification of the C2000 radio system as a significant issue during the emergency response, the IOOV did not make concrete recommendations as to how this was to be addressed. Rather, the IOOV recommended that the issues be researched further so that the cause of these issues may be determined (IOOV 2009, pp. 49-50, 54, 59, & 65).

Variable: Distinguish Between Strategic & Operational Lessons Learned (Schillemans & Smulders 2015)

Indicator Example

Both Strategic & Operational Recommendations

Made

Strategic

Recommendations made to the Kennemerland Safety Region on the creation of a supra-regional crisis policy for the Schiphol area involving both Schiphol and the four safety regions that border it. (OVV 2010, pp. 68, 122, & 125).

Operational

Recommendations made to the Minister of Interior & Kingdom Affairs to improve the efficiency and discipline on the C200 channels (OVV 2010, pp. 99-107, & 121).

Either Strategic or Operational Recommendations

Made Recommendations Made But The Level At Which This Is To Be Implemented Is

Unclear

Recommendation to improve the technical use of Cedric and the creation of overviews as prescribed in law through additional training (IOOV 2011, pp. 78, 82-84 & 91).

(27)

27 Validity & Reliability

Yin (2009) describes four measures for judging the quality of case study research designs. Construct validity, or the identification of the correct operationalized measures for the concepts being studied. Internal validity, or how well a causal relationship is distinguished from other spurious relationships. External validity, or the extent to which findings may be generalized. Finally, there is the measure of reliability, which is that the procedure followed should replicate the same findings if repeated on the same cases at a later period by a different researcher (Yin 2009).

According to Yin (2009), two steps are necessary to ensure construct validity. Firstly, the phenomenon to be investigated must be defined in terms of specific concepts and related to the objective of the research. Secondly, operational measures must be identified that match these concepts. In order to ensure this, accountability is defined as the mechanism by which forums provide actors with feedback-based inducements to foster organizational learning. Each case must therefore contain the elements of a forum, an actor, and some indication of organizational learning. Consequently, the operationalization was focused on indicators for the objectives and inhibitors of organizational learning, in order to determine if similar or different outcomes were present if the number of investigative forums changes.

Yin’s (2009) second measure of internal validity is concerned with the extent to which the relationship identified as causal by a researcher, may actually have been influenced by different and unforeseen factor. Moreover, it is concerned with the extent to which rival explanations have been considered. To satisfy this requirement, pattern matching between the expected pattern and observed pattern will be applied in the section on analysis. The pattern expected from the theory, namely that the IOOV and OVV will display similar outcomes in terms of learning as a form of accountability, should ideally match the pattern observed from the indicator scorings in order for internal validity to be assured (Almutairi, Gardner, & McCarthy 2014; Yin 2009).

Yin’s (2009) third measure of external validity. In the context of case studies, this specifically refers to analytical generalization of a particular set of results to a broader theory. In this thesis, the generalizations that may be inferred from the cases studied are limited to the Netherlands and as such, any inferences made should not be considered to hold for other countries. A replication of the research in a different country may help to build a more general answer to the perceived problems of accountability overload (Yin 2009).

(28)

28 Yin’s (2009) final measure is that of reliability. In order to ensure that the results of this thesis are replicable, the various steps taken that lead to the findings will be clearly stated throughout. Findings will be substantiated with evidence, as well as counter-evidence, from publicly available sources (Yin 2009).

Generalizability

As identified by Seawright & Gerring (2008), the diverse case selection strategy holds a stronger claim of representativeness than other small-N samples such as a typical case selection, aiding in the eventual generalizability of any results found. Owing to the diverse case selection method, causal inferences may be generalized to other cases with units that exhibit similar variation on Schillemans & Smulders ‘s (2015) learning conditions, see Table 2, and similar variation on the independent variable, 1 or 2 investigative forums. Eventual findings from this thesis will however not be generalizable to cases with higher degrees of variation on the independent variable (Rohlfing 2012, pp. 201-202; Seawright & Gerring 2008, pp. 297 & 300-301).

(29)

29 Chapter 4

(30)

30

Results

In this chapter, case studies on the Poldercrash at Schiphol, the Bushfire in the Strabrechtse Heide, and the Monster Truck at Haaksbergen are presented. The Poldercrash at Schiphol was investigated by both the OVV and the IOOV, whereas the Bushfire and the Monster Truck were investigated by the IOOV and the OVV respectively. In each case, the findings regarding the lessons to be learnt and recommendations made are firstly presented. These lessons and recommendations are subsequently categorized using the indicators for Schillemans & Smulders’ (2015) key objectives for learning from accountability; see Tables 4, 5, 6, and 7. These scores are compared, analyzed, and discussed in depth in the fifth and final chapter.

The Poldercrash at Schiphol: 25th_{of February 2009}

Following the deaths of four passengers and five members of the flight-crew after Turkish Airlines Flight 1951 crashed during a landing at Schiphol Airport, investigations into the causes of the crash and the response by emergency services were launched by both the OVV and the IOOV. The OVV investigation and subsequent report focused on the longer-term goals of drawing lessons from the crisis, namely how emergency services such as the fire department may improve their response. The IOOV investigation was performed in conjunction with the Inspectorate for Health and was focused upon identifying lessons that may be learned from the response to the Poldercrash, this time in terms of infrastructure, the medical assistance rendered, and the municipality of Haarlemmermeer’s response to the crisis (OOV 2010, pp. 17-19 & 125-126; IOOV 2009, pp. 4-10 & 17-18).

Firstly, findings by both the IOOV and OVV will be discussed. Afterwards, the specific findings by each investigative forum are presented in their own sections.

IOOV & OVV

Both the OVV and the IOOV note that the use of the C2000 network during the crisis was flawed. However, whereas the IOOV did not make any concrete recommendations as to how the issues with C2000 were to be resolved, the OVV did. According to the OVV, the problems with C2000 stemmed from the lack of agreements between the responders from

(31)

31 different safety regions, as well as a general lack of discipline on the ether rather than technical issues with the radio mast. The OVV therefore made the strategic recommendation to the Minister of Interior & Kingdom Affairs to standardize the communication structures for supra-regional use of C2000 as well as the direction thereof. The OVV also made the operational recommendation to the Minister of Interior & Kingdom Affairs to improve the efficiency of and discipline on the C2000 channels (IOOV 2009, pp. 49-50, 54, 59, & 65; OVV 2010, pp. 111-113, 117-118, 121, & 125).

Moreover, both the IOOV and OVV forwarded recommendations regarding the need for supra-regional policy for the Schiphol area and the four other safety regions that border it. The Kennemerland Safety Region and Amsterdam Airport Schiphol were given the recommendation by both the IOOV and OVV that they make collaborative policies and agreements regarding incidents that occur on the fringes between plans, organizations, and safety regions. The IOOV suggested that this be done through a synthesis between crisis management plan for incidents at Schiphol and those for general aviation incidents. The OVV on the other hand, stressed the need for binding operationalizable agreements between the Kennemerland Safety Region and Schiphol on information sharing (IOOV 2009, pp. 24-25 & 27-28; OVV 2010, pp. 68, 122, & 125).

One of the more striking aspects of the Poldercrash case is the OVV’s interactions with the various branches of the emergency response. The IOOV found that the Commando Plaats Incident, or CoPI, experienced a great deal of irritation with inspectors from the OVV. Approximately an hour and a half into the response, two investigators from the OVV gained access to the disaster site and attempted to begin investigating. As they had not notified the CoPI of their presence, nor their intention to access the site and investigate, these two OVV employees were removed from the site by members of the Royal Marechaussee shortly afterwards on orders from the mayor of Haarlemmermeer and the CoPI leader. The IOVV did not find any indication that the OVV’s presence affected the response, although it did lead to significant amounts of unnecessary irritation. The OVV was later granted access, and the IOOV notes that the Commandant of the Royal Marechaussee stated that the Strategic Team should have known that they had no legal grounds to stop the OVV inspectors. As such, the IOVV recommended that national policy be overhauled so that it become standard national policy that third parties, such as the OVV, report to the CoPI leader when entering an active crisis zone to avoid the excitation and annoyances observed in the Poldercrash (IOOV 2009, pp. 32-34 & 40).

(32)

32 The OVV responded to the IOOV’s findings in their report on the Poldercrash, providing a copy of their response to the Minister of Interior & Kingdom Affairs as an appendix. Based on findings from an internal investigation, the OVV stated that their investigators had in fact reported to the CoPI upon arrival. However, the OVV did respond to the recommendation, by adjusting investigative guidelines and asking the Minister of Interior & Kingdom Affairs to improve the knowledge of the OVV’s tasks, role, and legal position within the Kennemerland Safety Region. Notably, the OVV stated that they had not been enabled to respond to the IOOV report, unlike other partners (OVV 2010, pp. 137-144).

See Table 4 and Table 5 for an overview and indicator scoring of the findings, lessons to be learned, and recommendations made by the OVV and IOOV for the Poldercrash at Schiphol.

OVV

Whilst the OVV does note that whilst emergency response and medical assistance can generally be well organized in the Netherlands, when incidents occur on larger scales however, the actual responses to the crisis may not occur in the manner intended according to established procedure, as was the case with the Poldercrash. The OVV found that the choices made on which crisis management plans are to be utilized, the ascertaining of the location of the crisis, the correct utilization of emergency services, and the proper use of available resources are determining factors for the course of the emergency response. The OVV found several instances of procedures being incorrectly executed by the emergency responders. Perhaps most strikingly, the municipality of Haarlemmermeer had failed to register victims using registration cards as prescribed in Dutch policy (OVV 2010, pp. 87-79, 92, 96-98, 117-118, & 120-121).

Moreover, Turkish Airlines had failed to register the nationalities of the passengers on the flight, as required by the Dutch Vreemdelingenwet of 2000. Turkish Airlines did indicate however, that such a list had never been requested by Dutch authorities, such as the Royal Marechaussee up until the Poldercrash (OVV 2010, pp. 98-99).

Based on their analysis of the Poldercrash at Schiphol the OVV identified the need for the overhaul of national emergency response policies. The Minister of Interior & Kingdom Affairs and the Minister of Health, Welfare and Sport were therefore requested to create

(33)

33 uniform national agreements on the medical response to emergencies. This policy should address the need for a simplified method for registering large numbers of victims, the uniform upscaling of hospitals during supra-regional medical response, and the national coordination of mobile medical teams, MMTs. Such policy should also allow for MMTs and ambulances to respond supra-regionally and for hospitals to free up capacity. The Ministers were also requested to define which aspects of the crisis response, such as alarming, are to be directed and executed on a regional or national level (OVV 2010, pp. 125-126).

In addition to the operational recommendations made on the need for discipline on the C2000 channels, the OVV asked the Minister of Interior & Kingdom Affairs to consider the possibility that victim registration occur elsewhere than at the site of the incident. The Poldercrash illustrated that this requirement did not aid municipalities in fulfilling their legal responsibility to register victims (OVV 2010, pp. 121 & 125).

Strikingly, the OVV did not put forward recommendations or solutions towards the municipality of Haarlemmermeer’s inability to input data from the CRIB processes into a system shared by other respondents. Nor did the OVV suggest that the municipality switch to a more commonly used system, as opposed to the municipality’s choice for CRIB4ALL (OVV 2010, p. 97-98).

Additionally, the OVV found that the Kennemerland Safety Region’s C2000 connection scheme for the medical response was not followed, nor did these plans consider that the first ambulance to arrive might not be from the same region. As a result, the first ambulance to arrive, one from the Amsterdam-Amstelland Safety Region, was unable to coordinate and communicate to the Kennemerland Safety Region (OVV 2010, pp. 107-111).

The OVV also identified the need for several new skills and resources to be acquired in order to improve the emergency response. Despite a number of previous agreements, the Kennemerland alarm center failed to request assistance from LMAZ, nor had it implemented the Octopus system to automatically alert the necessary ambulance responders. The Kennemerland alarm center was also found to be lacking in the means to display overviews or maps. The Minister of Interior & Kingdom Affairs received the recommendation on the need for LMAZ to be utilized during large-scale crises, whereas the Kennemerland Safety Region

(34)

34 was to secure the requisite items and the implementation of Octopus for the alarm center (OVV 2010, pp. 72-79 & 109).

The OVV also identified the need for the Kennemerland Safety Region to translate its plans, manuals, and protocols for the alarm center into checklists or other types of easy to reference materials. Additionally, the need for additional training of centralists regarding the processing of information and alerting of emergency services such as ambulances was identified (OVV 2010, pp. 83-84).

Moreover, centralists at the Kennemerland alarm center were not separated from those who were occupied with their everyday duties in contrast to standing procedure. Dutch policy requires that centralists handling an incident move to a multi-disciplinary coordination room. This was however, not found to be the case in the Poldercrash (OVV 2010, pp. 79-81).

The Minister of Interior & Kingdom Affairs also received the recommendation that the Leidraad vliegtuigongevallenbestrijding be adjusted so that a uniform national standard for crisis plans at Dutch airports and their surrounding environments is facilitated. The OVV recommended that Schiphol serve as a pilot location for this (OVV 2010, pp. 66-68 & 125).

See Table 4 for an overview and indicator scoring of the findings, lessons to be learned, and recommendations made by the OVV for the Poldercrash at Schiphol.

(35)

35 Table 4: Poldercrash OVV Investigation Indicator Scoring

Variable: Understandings Gained (Schillemans & Smulders 2015)

Indicator Source(s) Implementation Of New Procedures As Old Procedures Were Found To Be Insufficient

Need for binding procedural agreements on procedure with Schiphol on information sharing (OVV 2010, pp. 68, 122, & 125).

Old Procedures Were Found To Be Sufficient, But Not

Implemented

Failure by the municipality of Haarlemmermeer to register victims using registration cards as prescribed in Dutch policy (OVV 2010, pp. 87-89, 92, 96-98, & 120-121).

Centralists at the Kennemerland alarm center were not separated from those occupied with everyday tasks as required by law (OVV 2010, pp. 79-81).

Failure by Turkish Airlines to register the nationalities of the passengers on the flight as required by the Dutch Vreemdelingenwet (OVV 2010, pp. 98-99).

The Kennemerland Safety Region’s connection scheme was not followed, nor did the standing procedures consider that the first ambulance to arrive may be one from outside of the region (OVV 2010 pp. 107-111).

Variable: Skills Learned (Schillemans & Smulders 2015)

Indicator Source(s)

New Skills/Resources Recommended

Recommendation made to the Minister of Interior & Kingdom Affairs on the need for the use of LMAZ (OVV 2010, pp. 72-79 & 125).

Need for the Kennemerland Safety Region to reformulate the plans, manuals, and protocols for the alarm center into checklists (OVV 2010, pp. 83-84).

Need for additional training of centralists regarding the processing of information and alerting emergency services (OVV 2010, pp. 83-84).

The Kennemerland alarm center was found to lack facilities for displaying overviews or maps, as well as a complete integration of the Octopus system (OVV 2010, pp. 76-77).

No New Skills/Resources

Recommended

No recommendations or solutions suggested as to solving to the municipality of Haarlemmermeer’s inability to input data from the

(36)

36 CRIB processes into a system shared by other respondents (OVV 2010, p. 97).

Variable: Distinguish Between Strategic & Operational Lessons Learned (Schillemans & Smulders 2015)

Indicator Source(s)

Both Strategic & Operational Recommendations

Made

Strategic

Recommendations made to the Kennemerland Safety Region on the creation of a supra-regional crisis policy for the Schiphol area involving both Schiphol and the four safety regions that border it. (OVV 2010, pp. 68, 122, & 125).

Recommendations made to the Minister of Interior & Kingdom Affairs to standardize the communication structures for supra-regional use of C2000 as well as the direction thereof (OVV 2010, pp. 111-113, 117-118, & 121).

Recommendation to Minister of Interior & Kingdom Affairs to adjust the Leidraad vliegtuigongevallenbestrijding so that a uniform national standard for crisis plans at airports is facilitated. Schiphol was recommended as a pilot location (OVV 2010, pp. 66-68 & 125).

Recommendation made to the Minister of Interior & Kingdom Affairs to create uniform national agreements on medical response to emergencies. Specifically, regarding a simplified method for registering large numbers of victims, the uniform scaling of hospitals during supra-regional medical response, the national coordination of MMTs, and the implementation of the LMAZ (OVV 2010, p. 125).

Recommendations made to the Minister of Interior & Kingdom Affairs and Minister of Health, Welfare, and Sport to organize medical response to crisis on a uniform national level so that MMTs and ambulance can respond supra-regionally and hospital capacity made available. The Ministers were also requested to define which aspects of the crisis response, such as alarming, are to be directed and executed on a regional or national level (OVV 2010, p. 126).

Operational

Recommendations made to the Minister of Interior & Kingdom Affairs to improve the efficiency and discipline on the C200 channels (OVV 2010, pp. 99-107, & 121).

Recommendation to the Minster of Interior & Kingdom Affairs to reconsider the need for victim registration to occur at the site of the incident in order to allow municipalities to fulfill their legal responsibilities (OVV 2010, p. 121-125).

(37)

37 Either Strategic or Operational Recommendations Made Recommendations Made But The Level At Which This Is To Be Implemented Is Unclear

(38)

38 IOOV

The IOOV report on the Poldercrash highlighted a number of shortcomings in the emergency response efforts by the Kennemerland Safety Region, the municipality of Haarlemmermeer, and the Geneeskundige Hulpverlening bij Ongevallen en Rampen, or GHOR.

Unlike the OVV, the IOOV presented their recommendations and findings in an overview in their summary chapter. This overview also contained information on whether the Kennemerland Safety Region, municipality of Haarlemmermeer, or GHOR were to implement which lessons, as well as with which external partners this was to be done (IOOV 2009, pp. 9-10).

The IOOV found a number of insufficiencies in existing procedures, which were to be solved through the creation of new procedures. Firstly, the coupling of VOS and GRIP levels was found to require a superfluous manual step by centralists in the alarm center, despite VOS and GRIP levels being linked in protocols. Secondly, the GHOR were found to be lacking a system for registering large numbers of victims. The coordination of the registration process of victims and their relatives was also found to be lacking. Moreover, a number of shortcomings with information sharing and management as a result of organizational structures, as well as the failure to fully implement a net-centric response structure (IOOV 2008, pp. 25-28, 42-45, 52-54, 61, 66-67, 73, 77-78, 85-86, & 93-96).

Moreover, the IOOV found that some procedures were not executed to a satisfactory extent. The Kennemerland alarm center’s multi-disciplinary process coordinator, MPC, was primarily occupied with the coordination of police, fire brigade, and ambulance disciplines and was therefore unable to fulfill their role as an information manager. Additionally, the MIMMS system for triage was not implemented properly by the GHOR, as too much focus was placed on victims’ ability to walk rather than potential trauma (IOOV 2009, pp. 59-61, 64-67, & 93-96).

The IOOV also identified the need for several new skills and resources to be acquired in order to improve the emergency response. Firstly, data entered into GMS by Kennemerland alarm center centralists was not presented in such a manner that centralists would not be able

Accidental Partners: A Comparative Analysis of Learning as a Form of Accountability in Dutch Accident Investigations