The effectiveness of Dutch crisis response networks: A study into factors influencing the effectiveness of operational networks within the Dutch crisis response system

(1)

Lidy-Anne Verkade - s1299794

A study into factors influencing the effectiveness of operational networks within the Dutch crisis response system L e i d e n U n i v e r s i t y – M S c C r i s i s a n d S e c u r i t y M a n a g e m e n t – 0 6 2 0 1 9

The effectiveness of Dutch

crisis response networks

(2)

1

Acknowledgements

I would like to express my deepest appreciation to everyone who has, in any way whatsoever, supported the completion of this dissertation and with that my graduation

from the Master’s Program of Crisis and Security Management.

Supervisor: Dr. J. Matthys Second reader: Dr. W.G. Broekema

(3)

2

4 Analysis ... 32 4.1 Case 1 ... 32 4.1.1 Effectiveness of the crisis response network ... 32 4.1.2 Factors influencing the effectiveness of networks ... 36 4.2 Case 2 ... 44 4.2.1 Effectiveness of the crisis response network ... 44 4.2.2 Factors influencing the effectiveness of networks ... 48 4.3 Comparative analysis ... 56 5 Conclusion ... 62 5.1 Results ... 62 5.2 Reflection ... 64 5.2.1 Reflection on the answer ... 64 5.2.2 Limitations encountered ... 65 5.2.3 Future research ... 65 5.2.4 Recommendations ... 66 Bibliography ... 67 Appendices ... Error! Bookmark not defined. A Report Case 1 ... Error! Bookmark not defined. B Report Case 2 ... Error! Bookmark not defined.

(4)

3

Abbreviations

Abbr. Dutch English

BHV Bedrijfshulpverlener Company First Aid Provider CaCo Calamiteiten Coördinator Calamities Coordinator CoPI Commando Plaats Incident Incident Location Command GBT Gemeentelijk Beleidsteam Municipal Policy Team GHOR Geneeskundige Hulpverleningsorganisatie Medical emergency services

IM Informatiemanager Information Manager

KMar Koninklijke Marechaussee Dutch Royal Military Police

MKO Motorkapoverleg Informal consultation network between first responders

NCTV Nationaal Coördinator Terrorisme en Veiligheid

National Coordinator of Terrorism and Safety

OvD Officier van Dienst Officer in Command

OvD-B Officier van Dienst – Brandweer Officer in Command of Fire Department OvD-BZ Officier van Dienst – Bevolkingszorg Officer in Command of Social Care OvD-G Officier van Dienst – Geneeskundige dienst Officer in Command of Medical Service OvD-P Officier van Dienst – Politie Officer in Command of Police

PSAP Meldkamer Public Safety Answering Point

RBT Regionaal Beleidsteam Regional Policy Team ROT Regionaal Operationeel Team Regional Operational Team

(5)

4

1 Introduction

Responsibility for responding to crisis is not assigned to a single organization. During, or in the aftermath of an incident many different organizations have a role to play. Health organizations provide medical services to the wounded; police or military organizations are responsible for intervening and stopping active shooters or preventing further attacks from materializing; and fire brigades rescue people trapped in buildings or vehicles after explosions or fires. Crisis response is thus a shared responsibility, and one that requires these various kinds of organizations to cooperate. In the Netherlands, the necessity of careful crisis response preparations and cooperating mechanisms became painfully clear in 2001 when a fire in a cafe in Volendam killed 14 people and wounded more than 200. Police units that reached the incident location first had rushed to the scene, leaving their cars on the roads leading to the café, and by that blocking them for other emergency services. Clear communication would have omitted the delay of ambulances arriving when they were urgently needed.

1.1 Networks and effectiveness

This study focuses on the design and the effectiveness of the operational level of the Dutch crisis management system. A structure of organizations has been designed to deal with crises that require more than one branch of crisis actors. The way in which this system functions can be seen as a network: a situation where several actors have formal or informal ties (Whelan, 2016). One could wonder how successful cooperation between organizations that vary widely in responsibilities, tasks, working culture and drills can possibly be, even though efficient and effective collaboration is an absolute requisite when dealing with multidisciplinary incidents. These incidents (often) create various urgent and drastic necessities (medical help, armed intervention, possible disarmament of explosives – to name a few) that all need to be addressed simultaneously and very rapidly. This thesis will focus on the effectiveness of this network of crisis actors involved in the first response to such incidents. The research question it aims to answer, therefore, is the following:

What factors influence the effectiveness of Dutch operational crisis response networks?

(6)

5

1.2 Academic and societal relevance

By answering this question, this study adds to the academic field of studies on effective (crisis) networks in various ways. (1) It provides a comprehensive framework on indicators of network effectiveness; (2) it conceptualizes and operationalizes effectiveness so that it can be measured; (3) it adds to empirical literature by incorporating observational data from (simulated) crisis situations; (4) and by comparing the framework of factors amounting to organizational effectiveness with the measured effectiveness of the Dutch operational crisis network, it tests whether theories about factors impacting organizational effectiveness are applicable to operational crisis networks, too.

On a more practical note, this study aims to provide not only academic, but also societal relevance. Conducting academic research on the effectiveness of the Dutch crisis management system, it strives to positively contribute to the improvement, or optimal functioning of the country’s crisis response. In the conclusion, several recommendations are drawn for the multidisciplinary crisis response in order to improve the network’s effectiveness. Furthermore, the framework comprising factors influencing the effectiveness of a network might not be limited to the improvement of crisis response networks, but may prove valuable for any attempts made to improve the effectiveness of any kind of organizational security network.

1.3 Reading guide

The study at hand follows the following structure: - Chapter 2: Theoretical framework

This chapter will outline the key concepts around which this thesis revolves. It will attempt to answer questions such as: What are organizational networks? What is (network) effectiveness? What factors contribute to effectiveness in organizational networks?

- Chapter 3: Methodology

This chapter elaborates on the research design in which the thesis is grounded, the means of data collection and analysis that have been utilized in this study, the reasons why these methods have been used and what effects this has on the reliability and validity of the conclusions of this thesis.

(7)

6

A systematic and coherent portrayal and subsequent analysis of the data gathered, based on the study of documentation and interviews and observations made.

- Conclusion

This chapter provides an answer to the research question, adds a reflection to the conducted study, and delivers several recommendations that would improve the effectiveness of the Dutch crisis response network (and perhaps (operational) security networks in general).

(8)

7

2 Theoretical framework

2.1 Organizational networks

The 1980s brought about a wave of reforms in the field of public management, related to public governance in particular. Globalization at the one hand, increased regionalization at the other; the call for smaller governments and other developments gave rise to a doctrine of bureaucratic reforms intended to shape the public sector in such a way as to better correspond with advancements in society (Hood, 1991). The fiscal crisis of the 1970s led to the conviction of many that governments should work harder and cost less. This reinvention of government was based on the thought of economic rationalism: many of the reforms of this doctrine named New Public Management1 centered around the idea that the public sector should take over well-functioning (management) concepts from the private sector (Denhardt & Catlaw, 2014). As such, it constituted a “framework for reorganizing management procedures in the public sector with the aim of greater effectiveness and efficiency” (Kalimullah et al., 2012: 9). It soon appeared, however, that the NPM doctrine was not as universally applicable as was hoped, and moreover that it had problematic effects, one of them being that rather than increasing overall organizational effectiveness, it engendered “management attitudes obsessed with intermediate organizational objectives” (Dunleavy, 2006: 472).

In the early 2000s, a new doctrine emerged: that of New Public Governance (NPG). Where NPM attempted to run the government as if it were a business through market-type mechanisms (such as contracts, incentive structures), New Public Governance centered around a network idea. Rather than the output and intra-organizational focus of New Public Management, NPG devotes itself to ‘relational organization’ (Osborne, 2006: 384). One of its main assertions is that of a plural state, where “multiple interdependent actors contribute to the delivery of public services” (idem).

Collaboration between various kinds of actors occurs between for-profits, not-for-profits, mixes of these two, and (as promoted by NPG) within governmental spheres; between public sector organizations. These structures of cooperation, where individual organizations that each have their own tasks and responsibilities have relations and cooperate with other individual organizations are often described using

1_{Only in the 1990s were the reforms bundled by academics and recognized as a doctrine that was}

(9)

8

the term network. Merely the existence of cooperation in some way between two or more organizations, however, does not necessarily indicate the existence of a network. In order to constitute an organizational network, (1) the involved organizations need to be autonomous and function independently, while (2) entertaining relations in either formal or informal manners, so that they (3) can (continue to) achieve their individual responsibilities, while simultaneously (4) achieving synergy by cooperating with the other involved organizations towards a shared aim.

O’Toole described the importance of networks from a public administration perspective, stating that “Policies dealing with ambitious or complex issues are likely to require networked structures for execution” (1997: 46). Wicked problems (problems that have no stopping rule; and are all essentially unique; solutions of which are not either true or false; solutions are a ‘one shot operation’, and cannot be tested (Rittel & Webber, 1973)) can only be addressed through a cooperation of various organizations with each their own expertise and resources.

Whelan (2016), too underlines the significance of this type of cooperation between organizations, focusing specifically on the field of national security in his comprehensive study of the phenomenon in relation to operational coordination. Stating that network forms of governance “involve repetitive exchanges between a set of autonomous but interdependent organizations in order to achieve individual and shared objectives” (2012: 4), he asserts that governments are (and should be) more and more introducing network structures to their crisis management systems. His arguments resembled Ric Smith’s findings (2008), who researched the most effective way of designing the Australian crisis management system. Where the elected government had indicated a preference towards a Homeland Security organization that resembled the one in the United States2, his study showed the benefits of adhering to the status quo in terms of keeping various independent organizations, but creating a network through which they could cooperate when necessary.

2_{One department in which all crisis related disciplines are brought together with the idea of that}

bundling increasing the cooperation between them and thus increasing effectiveness of crisis management

(10)

9

2.2 Organizational effectiveness

“Organizational effectiveness is both a powerful and a problematic concept,” says Forbes (1998: 183). Where it enables a critical reflection of the way in which organizations function, the term in itself means different things to different people and attempts to measure it have been undertaken in many different ways. Seeing as different organizations have different goals and therefore do not all perceive ‘functioning well’ in the same way, attempts to study the effectiveness of organizations have often focused on specific types of organizations (e.g. by distinguishing profit from nonprofit organizations (i.a. Balser & McClusky, 2005)). As for many private organizations (businesses) goals center around making profit, their effectiveness is often related to concepts such as market growth, profitability and customer satisfaction (e.g. Steers, 1975). For national security organizations, the ultimate aim is not to make profit, but to live up to what is demanded of them by the government through laws and regulations and to meet the expectations of society concerning the (relative, or optimal) safety and security of and in their country. What defines effectiveness is that an actor, organization, or a material product has a goal, or is installed in order to meet a certain need, or should have a certain effect; and that this goal is indeed reached, or the need has been met, or the desired effect has been achieved.

Research projects studying the effectiveness of networks have often followed this line of reasoning: the effectiveness of a network of protected nature areas was judged according to i.a. the occurrence of several animal species and the quality of forest conservation (Martínez et al, 2006); and the effectiveness of global health networks according to the effect they had had on policy changes or political agenda setting related to i.a. tobacco use and alcohol harm (Shiffman et al, 2016).

In an in-depth study on networks and national security, Whelan (2016) states that effectiveness in that context mostly relates to two main concepts: that of information sharing, and the extent to which the information sharing is efficient (2012: 39). (National) security networks, he states, are installed to that objective, and effectiveness therefore means whether or not they achieve that. Zheng et al (2010), however, while underlining the importance of knowledge management and knowledge integration, state that information sharing is a factor that supports an organization in achieving its goals, and is therefore a factor contributing to

(11)

10

effectiveness rather than a synonym of the state of effectiveness of an organization. In a study on coordination by officers in command in the Netherlands, Wolbers et al. focus on the effectiveness of crisis networks in terms of whether the total of measures making up the crisis response represent a coherent and unified set of actions, i.e. on the integration of the organizations and the subsequent unity of measurements taken by the network (2018: 1521). Crisis response actions should, however, not only be coherent and unified, they should also be appropriate for the situation at hand and well-executed. That is what operational networks aim for in their crisis response. The continuation of this study will therefore adhere to Zheng et al. (2010)’s conceptualization of effectiveness: the extent to which the network achieves or doesn’t achieve its goal(s).

2.3 Factors influencing the effectiveness of networks

As the effectiveness of networks has been defined, the question arises what factors influence this. What elements of networked cooperation support or hinder networks’ endeavors to attain their goals? The first attempt to explore factors potentially influencing the effectiveness of (primarily public sector) networks was made by Provan & Milward in 1995. Their study distinguished several structural characteristics of the network itself, and various contextual factors that they understood to conjointly shape network effectiveness. Several other scholars have come up with adapted or new frameworks, focusing on various kinds of networks and addressing different categories of factors and levels of effectiveness.

2.3.1 Whelan’s framework

An in-depth analysis on factors influencing national security networks’ effectiveness was made by Whelan (2016). Whelan designed this framework to apply it to the Australian national security system. We believe that it is applicable for operational crisis networks, too, for a number of reasons: Whelan’s factors are not specifically centered around a specific type of network but can be understood to be general factors for any given organizational network; both types of networks have a similar aim (preventing or addressing crises, establishing security); and the organizations involved in the national security network are often the same as those part of the

(12)

11

operational crisis network. Whelan presented a framework composing five pillars: (1) network structure, (2) network culture, (3) policy, (4) technology, and (5) relations. (1) Network structure is made up of two different units of analysis: the way in which the network is designed, and how it develops when confronted to changes in tasks or circumstances or situations. The design of networks can be divided into three basic categories: one where information flows through a single managing actor, one where each actor has individual relations with every other actor and information flows from everyone to everyone, and one where every actor is linked to only several other actors and information flows occasionally have to pass these actors before reaching the targeted receiving actor. In a graph, these designs would look like this:

Design 1: all actors are connected to a managing actor through which all information flows

Design 2: all actors have direct relations with all other actors, information flows directly from one actor to all other actors

Design 3: no actor is connected to all actors, in order to share information with all actors involved, actors function as intermediaries

The third design as a form of network is contested. Provan and Kenis (2008) in a study on modes of network governance distinguish modes of governance along two

Managing actor actor 1 actor 2 actor 3 actor 4 actor 1 actor 2 actor 3 actor 1 actor 2 actor 3 actor 4 actor 5

(13)

12

dimensions: the extent to which network governance is brokered (ranging from shared governance to one lead organization); and whether the network is “participant governed or externally governed” (idem: 234). They thus do not mention the third design as a type of network.

Which of these designs is preferable depends on the situation. Whelan (2012: 33) states that when the network is needed to function as an authority on its own, and/or when trust between the various actors is low, the first design may be preferable. A managing actor can broker between several parties that would not successfully cooperate as a result of a lack of mutual trust. However, when a network does not comprise many different actors and all actors agree on the goal and what is necessary to achieve it, the second design may be efficient, as it allows for quick information sharing and actors are able to connect directly with the actor they need to cooperate with.

The development of a network in terms of how dynamic it is looks at whether a network can respond to particular situations or changes. One way of analyzing this is by studying the actors involved in the network in various situations. It may not be efficient to include all actors in situations where only half of them have a role to play, or in exceptional situations additional actors may need to be added to the network. (2) Network culture, too, is made up of two components: the culture of the network as a whole, and the individual cultures of the separate organizations coming together within the network. Although culture in itself is a widely disputed concept and network culture not less so, Whelan defines it as “the shared beliefs, values and attitudes that can be identified at the network-level” (2012: 34). The various disciplines involved in crisis management like police and military forces are generally accepted to have their own organizational cultures, (radically) different from other types of organizations (i.a. Crank, 2014). A shared culture amongst different organizations or actors working together increases effectiveness as the actors understand each other quicker and are more willing to cooperate as they all strive for the same (or at least, very similar) goals.

The second component revolves around the management of differences in cultures that the individual organizations portray and come together in the network. This may be a factor of concern for any network, but especially so in crisis management networks. In case no shared culture is present amongst the organizations, or if certain

(14)

13

differences exist, these differences should be mitigated in order for the network to function optimally.

(3) Network policy, like the former pillars, consists out of a dual analysis. First, it centers around the policies that are in place that guide the network’s functioning. It for example focuses on how information sharing, and the individual organizations’ tasks and responsibilities are defined in policy documents. If procedures and aims are clear for both the individual organizations and the multidisciplinary as a whole, the actors involved in the network can act more effectively towards those set goals. Secondly, it analyzes the flexibility of these policies. In certain occasions, information sharing may need to be addressed in different ways (for example, when information needs to retrieved from unusual sources), or tasks and responsibilities may be rearranged (for example, when the capacity of an organization is limited and other organizations need to take over some of its responsibilities). If a network’s policies are very rigid, this may have a negative impact on its effectiveness.

(4) Network technologies focuses on both the design and the usage of information and communication technologies. Are functional technologies in place that support effective sharing of information? Are they designed in such a way that allows for all needed data to be shared, that is user-friendly, and at the same time deals with the protection of data (often a priority in the field of national security)? On the usage part, it focuses on whether or not these technologies are indeed used by the various organizations: great technology is futile if no actors use it.

(5) Network relationships centers around relationships between the individual persons in the network. The analysis focuses on the concept of trust: do the people involved in the network trust each other to function well as representative of their organization and as part of the network? Where trust has a positive influence as it increases cooperation and willingness to share information, distrust undermines the effectiveness of organizational networks as it does the exact opposite (Whelan, 2012: 37).

2.3.2 Additional factors

Whelan’s framework seems, at first sight, to present an exhaustive framework of factors influencing network effectiveness. Although it indeed provides a valuable base for analysis, an in-depth study of (public sector) network effectiveness literature

(15)

14

provided several other determinants affecting the effectiveness of organizational networks. Often, factors were titled or categorized differently and therefore included in Whelan’s framework in another form. Some, however, are remarkably different from the five factors mentioned above. They are treated in the following and added to Whelan’s factors in order to allow for a comprehensive analysis of factors influencing network effectiveness.

(6) Knowledge sharing: Where Whelan views information sharing as constituting the effectiveness of a network (as discussed in 2.2), many other academics believe it to be rather a factor attributing to it (i.a. Zheng et al., 2010; Weber & Khademian, 2008). Whelan refers to information sharing in the fourth factor of his framework, ‘technology’, but focuses on the availability of communication and information technology and their functions and user-friendliness and whether they are in fact used. Weber & Khademian refer to another aspect of information, related to “the transfer, receipt, and integration of knowledge across the network” (2008: 334), stating knowledge sharing is one of the main advantages of networks. Instead of focusing on the utilization of communication technologies, this sixth distinguished factor zooms in on the process of information sharing between the different organizations making up the network. Correct and complete information should be shared in a timely manner in order for the network to be able to take adequate action.

(7) Network age: Kenis & Provan (2009) argue that network age plays an important role in network effectiveness, as it can take up to five years before a network realizes important progress as opposed to the situation beforehand. In the continuation of this study, ‘age’ will be treated as ‘experience’ since the members that make up the operational network this study has taken as object of analysis fluctuate.3 Where ‘age’ focuses on the total of the network, ‘experience’ relates to the individual members that make up the network. Experience in leading crisis situations may prove a strong benefit, as the crisis leaders may have learned from earlier experiences (i.a. Carmeli & Schaubroeck, 2008), whereas lack of experience may have a severe negative influence (i.a. Taneja et al., 2014). Experience with other members, too, may have an impact: knowing how the others operate as well as the establishment of a trust

3_{Network age can therefore not be determined, as usually the networks formed during crisis situations}

are made up by officers in command who work in shifts; the composition of the network thus changes every time.

(16)

15

relationship may positively influence how a network functions (Raab et al., 2013; Turrini et al., 2010).

(8) Interaction: Turrini et al. offer a contested potential influencing factor: networking between the organizations, defined as “time spent in interactions among network constituencies” (2010: 536). In public sector networks, managers regularly face situations where they interact with, or steer not-direct subordinates, which has a positive effect on the effectiveness of resources used (O’Toole & Meier, 2003 in Turrini et al., 2010). It is contested as some scholars found managers of networks undertaking in-depth networking were as a result occupied with “facilitating interactions, brokering disputes, arranging meetings” (idem: 544). In the case of multidisciplinary cooperation, however, these perceived as negative pursuits are an essential part of the network’s responsibilities.

(9) Leadership is a frequently mentioned influencing factor in network effectiveness literature, several authors even claim it to be a ‘key factor’ (i.a. Conrad et al., 2003; Weiss et al., 2002). A strong leader can steer and guide the network in the perhaps chaotic crisis situations and subsequent cooperating process. He can motivate and inspire the members of the network. He can support in decision-making by taking key decisions or aiding the process from his perhaps impartial or distant outlook; and ‘activate’ the others: “selecting the appropriate actors and resources for the network” (Turrini et al., 2010).

Several factors found in literature have been left out as they do not correspond with the scope of the study at hand. Raab et al. (2013), for example, discuss system stability. This may certainly influence the outcome of the complete network process, but it considers an external factor affecting the network rather than an internal determinant. Resource munificence (i.a. Turrini et al., 2010), too, has not been incorporated in the continuation of this research: rather than focusing on the network as a whole, it analyzes the functioning of, or the staff available to the individual organizations making up the network. Here, too, a possible influence on the network outcome cannot be denied – but it does not align with the focus of the current study.

(17)

16

3 Methodology

3.1 Research design

The study at hand has a dual aim: it attempts to offer a framework of factors influencing the effectiveness of organizational networks; while simultaneously applying that to the Dutch crisis management system in order to study its effectiveness at the operational level. The former provides mostly explanatory research because it tests existing frameworks and academic claims regarding factors influencing network effectiveness, then applies these specifically to the field of operational crisis networks and finally attempts to come up with a comprehensive framework on the factors that have indeed been proved to affect effectiveness. The latter, application of this framework to the Dutch crisis management system, can be viewed as applied research, and more specifically as evaluation research (a type of research often used to evaluate the effectiveness of i.a. governmental programs (Neuman, 2014: 28)).

It is, then, of qualitative nature: it provides an in-depth analysis of several theoretical concepts. Rather than measuring the exact amount of influence a given factor has on the effectiveness on a network, it intends to gain a deeper understanding of the range of factors that could possibly apply. Furthermore, data has been collected through usage of qualitative methods.

The research has followed a nonlinear path: a concept framework of factors influencing effectiveness was built first upon already existing literature in the field of (public sector or national security) network effectiveness, but rather than viewing it as set in stone, it was a flexible framework that was reanalyzed after every step of the continuation of the study, in order to allow the final framework to incorporate all data retrieved during the course of the research.

When pictured as a graphic, the causal mechanism around which this study centers would look as follows:

(18)

17

The independent variable consists out of the various factors found in the literature that are deemed to have an effect on the dependent variable, which is established as the effectiveness of networks.

As for the scope of this research: The individual functioning of the autonomous organizations is an important determining factor of the total outcome of the organizations interacting and delivering a common result. One could state that ‘a chain is only as strong as its weakest link’: the network is dependent of every organization both to carry out their own responsibilities in an optimal way and of them to function as an integrated part of the network. This study, however, will not focus on the effectiveness of the crisis management system in the Netherlands as such, but it takes as subject of analysis the way in which it is structured: as an organizational network. It will limit itself to study network effectiveness at, indeed, network-level and thus leave outside of its scope the individual actors per se.

3.2 Case study

In order to examine in-depth the functioning of the Dutch crisis response system, case-study research has been conducted, as we “seek[…] to construct representations based on in-depth, detailed knowledge of cases” (Ragin, 1994 in Neuman, 2014: 42). Case studies allows us to compare theoretical findings with lived experiences and

(measured) effectiveness Structure Culture Policies Technology Relations Knowledge sharing Experience Interaction Leadership

(19)

18

therefore see if the academic theories using abstract concepts line up with reality. Furthermore, studying a few cases into detail allows for a deeper comprehension of the networks’ functioning as through examining specific cases “intricate details of social processes and cause-effect relations become more visible” (Neuman, 2014: 42). Below, first the general context in which the cases studied are grounded will be described. Then, the selection of specific cases will be argued for.

3.2.1 Context: the Dutch crisis response network

This study will focus on a network within the Dutch crisis response system, meeting all four criteria for an organizational network as discussed in 2.1. A crisis situation demands input from various disciplines, amongst which fire brigades, the police, medical services, social care, and in certain cases the military. During ‘mono’ incidents (incidents where only one of these disciplines is essential) these organizations are autonomous and operate independently. However, during crises or terrorist incidents, the organizations involved actively coordinate their (first) response.

As soon as a Public Safety Answering Point (PSAP, in Dutch: Meldkamer) receives notification of an incident, they assess the situation. Based on this assessment, they determine the scale of the crisis, and which crisis response organizations are needed to address it. For any serious crisis and every (suspected) terrorist incident, at least these organizations will be alerted to head to the scene: medical services, fire department, police. Depending on the gravity and scope of the calamity, various other organizations, actors or institutions will be notified (e.g. NCTV, municipality, Veiligheidsregio (‘Safety Region’), LOCC, Rijkswaterstaat, waterschap).

Actors of the three emergency services (police, fire department, ambulance) arriving to the scene first immediately set up a so-called ‘Motorkapoverleg’ and with that, establish a network setting of interorganizational cooperation. This cooperation is formalized in a CoPI (Commando Plaats Incident – or ‘Incident location command’) when the incident is deemed to require upscaling in accordance with the GRIP-procedure.

(20)

19

GRIP-procedure

When a calamity (of all kinds) is of such nature that it requires a multidisciplinary approach from a variety of emergency services, the otherwise autonomous emergency services merge into a single organization/network in order to coordinate their (operational) addressing of the situation. For that process to occur smoothly, the GRIP-procedure was designed. It knows 5 levels of upscaling. A higher level may indicate a more serious situation and/or a larger geographical reach of the incident.

GRIP level Scope Forums involved

GRIP 1 Source area CoPI

GRIP 2 Source area + impact area CoPI + OT

GRIP 3 Administrative involvement / decision-making CoPI + OT + GBT

GRIP 4 Crossing municipal borders CoPI + ROT + RBT

GRIP 5 Crossing regional* borders CoPI + ROT + RBT

OT: Operationeel Team (Operational Team)

GBT: Gemeentelijk Beleidsteam (Municipal Policy Team) ROT: Regionaal Operationeel Team (Regional Operational Team) RBT: Regionaal Beleidsteam (Regional Policy Team)

*‘regional’ here does not indicate the provincial regions but the Veiligheidsregio’s (Safety regions)

(Cools et al., 2017)

Both the Motorkapoverleg and the CoPI are made up by officers in command (‘Officieren van Dienst’: OvD) of the emergency services, function as the forum present at (or in close vicinity of) the incident location and in charge of operational management. Additionally, in a CoPI several other positions are occupied: CoPI leader, information manager (IM), plotter, and the officer in command of social care (OvD-BZ). The Motorkapoverleg or CoPI (either one or the other: the first if no upscaling is deemed necessary, the second if upscaling occurs and follows the GRIP procedure) remains in place for as long as the crisis continues, regardless of further upscaling of the calamity (upscaling will lead to involvement of more and/of other actors, often at higher levels, that will not replace but add to the already ongoing crisis management).

(21)

20

These two forums of multidisciplinary organizations cooperating in response to crises, functioning as a network made of various otherwise autonomously acting organizations, are the subject of study in this thesis.

3.2.2 Choice of cases

The selected cases were studied in a cross-sectional way. Although not all data was gathered at one point in time, the time period was a rather limited one and deemed irrelevant for the outcome of this study as no major changes in the functioning of Motorkapoverleggen (MKO) / CoPI’s were instructed during the timeframe. The focus of this research has not been to study changes over time, but rather to get an impression of the current functioning of MKO/CoPI’s.

Multiple cases were selected as object of study: a single observation would not give a complete understanding of whether or not the factors deemed to influence network effectiveness are valid for operational-level crisis response networks as different types of crisis demand different methods of (operational) action. Officers in command of emergency services are frequently trained in being part of a MKO/CoPI during crisis simulations. Situations that resemble reality as closely as possible are designed to allow actors partaking in MKO/CoPI’s to prepare for real-life operations. These training moments therefore present a suitable opportunity for studying (factors influencing) the effectiveness of these networks: (1) as the simulations closely resemble real situations, so does the manner in which the MKO or CoPI acts; (2) studying simulations omits ethical dilemmas that making observations in a real situations could bring along (such as (unwillingly) hindering operations out of a need for clear observations, or the unwanted effect that an observer may have on the officers in command when they should be able to focus on the crisis situation completely); (3) where after real crisis situations (especially multidisciplinary) evaluation may take some time, and may not be available to the researcher, evaluation opportunities are specifically arranged after each simulation in order to discuss the training and thereby learning from it. Furthermore, this grants the researcher the opportunity to ask for clarification of observed actions or matters, for meaning or motivations behind decisions made immediately after a simulation is done and with that gains a more in-depth understanding and evaluation of the cases. Cases were selected based on several criteria:

(22)

21

- MKO with OvD’s of at least the fire brigade (OvD-B), police (OvD-P) and medical services (OvD-G) or complete CoPI formation;

- Researcher was allowed access to site and could make live observations of MKO/CoPI’s operations;

- Researcher was allowed to participate in the evaluation session (as a spectator, not necessarily having an active role);

- Researcher had opportunity to speak with OvD’s and relevant actors involved. In order to maximize the observations of deviating independent variables (allowing a more inclusive and accurate measurement), the cases were furthermore chosen based on:

- A combination of MKO and CoPI observations (including both informal and formal forums of multidisciplinary cooperation);

- Trainings organized by different Veiligheidsregio’s as they may have different standard operating procedures.

This lead to the selection of the following cases:

- On 28 March 2019, Veiligheidsregio Kennemerland organized a CoPI training day. Three crises were simulated during an entire training day. A complete CoPI took part in the training (OvD-P, OvD-G, OvD-B, OvD-BZ, Spokesperson, Plotter, Information Manager, CoPI leader), as well as several other OvD’s functioning as PSAP, on-site units, HOvD, etc.

- On 11 April 2019, the emergency services of Dutch provinces Drenthe, Friesland and Groningen organized a Motorkapoverleg training day. A total of four cases was presented during the day, simulated in the open air on an industrial terrain and in the city center. The Motorkapoverleggen all incorporated an OvD-P, OvD-G and OvD-B.

3.3 Data collection and analysis

In order to study the presence of the factors influencing the effectiveness of networks within the crisis management networks, data was collected in various ways. The MKO/CoPI’s were studied as a network alongside the concept framework of influencing factors as set out in 2.3. The triangulation method was applied in the sense that various methods were used to study the same subjects, in order to improve accuracy of the results (Neuman, 2014: 166). Documentation about the Dutch crisis

(23)

22

management system, more specifically the MKO/CoPI, was analyzed in order to, amongst others, analyze the design of the network structure and whether there is any room for flexibility should the situation so require, and the policies in place (regarding i.a. responsibilities and tasks, information sharing processes). Then, the cases as set out in 3.2.2 were observed. Reports of the observations were written (Appendix A), based on which analyses of all potentially influencing factors were made in accordance with the indicators presented in the framework below (3.4.1). The analyses of the cases were compared in a comparative analysis and portrayed in a comprehensive framework (4.3).

3.4 Operationalization of variables

In order to be able to measure qualitative data, the key concepts used and their respective dimensions need to be operationalized. Based on the factors influencing effectiveness as discussed in the theoretical framework, a conceptual model has been designed which was used first during the observations made, and second as a basis for the analysis and conclusions drawn. The concept of effectiveness has been differentiated into several quantifiable units. A corresponding measurement scheme portrayed below enabled assessment of the networks’ effectiveness.

3.4.1 Operationalizing factors influencing effectiveness

The factors distilled from literature that have been deemed to influence organizational network effectiveness have been operationalized into assessable items as follows: (1) Network structure studies the design and the development of the governance of the network. The design deemed most efficient for crisis response networks is the second model as distinguished in 2.3.1: ‘everyone to everyone’. This model will therefore be included in the framework below in order to analyze whether that indeed influences effectiveness. Secondly, the flexibility of the design is analyzed. If the same organizations take part in every situation and responsibilities and management within the network are always the same, the structure is considered inflexible; if (representatives of) organizations can be added to or left out of the network, and responsibilities and/or management of the network depend on the situation at hand, the structure is considered flexible.

(2) Network culture has two components that can be assessed separately. First, whether a shared network culture, which is understood as shared beliefs, values and

(24)

23

attitudes, can be identified or not. Second, whether cultural differences can be distinguished between organizations within the network and, if so, if this is problematic and whether these differences are mitigated.

(3) Policies are assessed based on the actual policies that are in place on a number of issues: the tasks and responsibilities of each organization within the network; and the information sharing within the network. Then, the flexibility of these policies is analyzed to see if they can vary depending on (the gravity or nature of) a calamity. (4) Technology is assessed based on the design of technologies in place and, subsequently, the usage of those technologies. Considering design, first is studied whether information and communication technologies are in place at network-level. If the answer is positive, the technologies’ user-friendliness, required functions and well-functioning are assessed. Then, analysis will zoom in on the usage of the technologies, more specifically on whether they are used by all (required) actors, and whether the (required) actors use the technologies well or effectively.

(5) Relationships between the organizations making up the network is assessed in two ways: first, whether the network members trust each other with regards to their functioning within their own respective organizations; and second, whether the members trust each other to function well as a part of the network (i.e. in terms of cooperation).

(6) Knowledge sharing assesses information shared in terms of quality (clarity and completeness) and speediness of sharing (analyzing the time between one of the network members received a piece of information until the moment they share it with the rest of the MKO/CoPI members).

(7) Experience: Officers in command making up MKO/CoPI’s may vary every time, depending on the region where an incident takes place and which officer in command is on duty at that particular time. Age of the network itself, then, may not be as relevant as the level of experience of the officers in command is. This factor will therefore asses (how often) network members have been part of an MKO/CoPI before, and on a more personal level: whether the members representing the various organizations worked together in a MKO/CoPI before.

(8) Interaction is assessed two-fold: focusing on the networking done during status quo situations in order to prepare for cooperation during crises; and the actual networking during a crisis – here focused not on interaction between MKO/CoPI-members, but on the communication towards, and perhaps coordination of “not direct

(25)

24

line subordinates” (Turrini et al., 2010: 543) by the officers in command (in other words: when e.g. an OvD-P is in charge, does he communicate with, or command fire fighting personnel?).

(9) Leadership is analyzed based on a few components: whether a leader is present in the network, and if so: whether he motivates and inspires the other members; whether he influences the decision-making process; and whether he activates other members. This leads to the following model:

(26)

25

FACTOR – CONCEPT SUBCONCEPT VARIANTS/CATEGORIES OF SUBCONCEPT INDICATORS

(1) Structure Design Everyone to everyone - All actors have direct communication with all other actors; - No single actor through which all information flows

Not everyone to everyone (different model) - Deviation from above indicators;

- Actors do not communicate directly to each other; - Single managing actor through whom all info flows; - Information flows occur through a ‘wheel design’

Development Inflexible No deviation from structure design

Flexible Deviation from structure design

(2) Culture Shared culture Identified Actors present:

- Shared beliefs (i.a. goals, strategies) - Shared values (i.a. best outcomes, priorities) - Shared attitudes (i.a. in addressing the situation)

Not identified No indication of above indicators

Mitigation of differences Cultural differences mitigated Actors attempt to overcome cultural differences: explain when things are not understood, try to get everybody ‘on the same page’, discuss differences in opinion in a positive manner

Cultural differences not mitigated No attention paid to possible different cultures or differences of opinion as a result of organizational culture variations

(3) Policies Policies in place Tasks and responsibilities Documentation indicating the tasks and responsibilities of the networks (and their members) is (not) available

Information sharing between organizations Documentation indicating the manner in which information ought to be shared between the different members in the networks is (not) available

Flexibility of policies Inflexible No deviation from aforementioned policies

Flexible Deviation from aforementioned policies

(4) Technology Design Are technologies in place? If yes: Technologies are (not) available for usage by MKO/CoPI

User-friendliness The technology is (not) easy to understood and use

Offer all functions required All options that the members need are foreseen in the technology (i.a. communication, visualization)

(Usually) function well The technology functions, members do not experience problems using it

(27)

26

them

Used effectively The actors (do not) use the technology:

- In the correct way (so that technology functions well); - For the right purpose

- At the right time

(5) Relationships Individual Trust in independent well-functioning - Actors accept/agree with responsibilities given to others; - Actors evaluate each other positively

Distrust in independent well-functioning - Actors question responsibilities given to others; - Actors evaluate each other negatively

Within network Trust in well-functioning within network - Actors accept (operational) proposals made by others; - Actors accept decisions taken by others

Distrust in well-functioning within network - Actors questions (operational) proposals made by others; - Actors questions decisions taken by others

(6) Info sharing Quality Clarity Information is communicated clearly; all members understand what is being shared

Completeness All relevant information is shared, no important information is left out

Speediness Rapid sharing Relevant information is shared as soon as possible with relevant actors

Slow sharing Relevant information is not shared in a timely manner

(7) Experience As OvD Experienced A network member is regarded as experienced when (s)he has had the position they fulfill now for at least one year

Inexperienced A network member who has operated as such for less than one year

With other OvD’s Previous cooperation Members of network have worked together before in crisis situations or simulations

No previous cooperation Members have not worked together before in crisis situations or simulations

(8) Interaction Before crisis Previous interaction opportunities The network members have met each other before the case, i.a. during networking opportunities, meetings or other trainings

No previous interaction opportunities The network members have not met each other as no networking opportunities, meetings, trainings were organized before

During crisis Commanding other-than-direct subordinates Network members command on-site units that would not normally fall directly under their lead, i.a. OvD-P instructs fire fighters

(28)

27

(9) Leadership Presence A leader is present within the network One of the network members is appointed or acts as the leader of the network

No (clear) leader is present within the network No network member has or takes the lead

Motivation – inspiration The leader motivates/inspires other members - Leader creates a positive working atmosphere;

- Leader actively instigates other members to achieve the best results

The leader does not motivate/inspire Leader does not act as described above

Decision-making The leader supports the decision-making process - Leader clearly structures the meetings;

- Leader asks about issues that have to be decided on; - Leader asks for suggestions / possible strategies; - Leader ensures a final decision is taken

The leader does not support the process Leader does not act as described above

Activation The leader activates members - Leader indicates which actor is responsible or appropriate for tasks; - Leader decides whether other resources are needed for the network and how that will be arranged

(29)

28

3.4.2 Operationalizing effectiveness

Measuring effectiveness is as contested as the definition of the concept itself. Analysis (and therefore measurement) can occur at several levels; and the criteria upon which analysis is based are more normative than factual: “imperatives about the preferred state of a system [...] cannot be completely derived from [...] fact[s]; nor can they be proven empirically” (Simon, 1976 as cited by Kenis & Provan, 2009: 443). Where the effectiveness of public health networks can be measured by assessing the results for patients treated, and for-profit networks by the number of services or products provided, or how well these were received; the ‘security’ that security networks attempt to provide is in itself a widely challenged concept. Definitions or perceptions of security vary, and complete security for all is a state that cannot be reached as new threats develop continuously (i.a. Williams, 2012). Measuring how effective a network has been in providing security, then, would not provide any reliable results. A similar issue arises when looking at crisis response networks in particular: defining how ‘well’ a network responded to a crisis is hardly quantifiable. Counting the number of survivors seems questionable, as does measuring the amount of time it took before a network ‘solved’ the crisis: criteria for judging a crisis to be over are subjective to say the least, and when a crisis lasts for a long period of time, it does not necessarily mean that the crisis response network is not functioning optimally. Instead, as discussed in chapter 2.3, this study considers effectiveness to be the achievement of goals for which the network was created. The effectiveness of the Dutch crisis management network can thus only be analyzed after distinguishing its various goals and measuring whether they are met. An important additional factor when measuring the effectiveness of networks is added by Provan & Kenis (2008): “the attainment of positive network level outcomes that could not normally be achieved by individual organizational participants acting independently”. The goal of a MKO/CoPI may be to ‘limit the number of casualties or material damage as the result of a crisis’, but this can be regarded as the goal for any of the organizations individually, too. To that end, goals of the network of operational emergency services during crisis incidents, that could not be achieved should the independent organizations work individually, should be studied and subsequently evaluated whether these aims were in fact achieved.

(30)

29

Aims of the multidisciplinary cooperation, apart from those inherent to the responsibilities of the organizations making up the MKO/CoPI, are (de Jonge, 2008):

1. Defining the situation and possible developments at the incident location (multidisciplinary);

2. Deciding on a multidisciplinary, coordinated strategy for rapid emergency assistance and/or incident control;

3. Leading the disciplines taking part in the crisis response at the incident location;

4. Providing the operational units at the incident location with information about the incident control;

5. Providing the mayor (and in cases of GRIP 2 or higher: the ROT) with (multidisciplinary) information about developments at the incident location; 6. Taking measures aimed at preventing and/or limiting the number of casualties

at the incident location;

7. Coordinating emergency assistance to victims.

In order to measure the effectiveness of the MKO/CoPI’s, their performance is judged in accordance with these 7 criteria. Each of the criteria is rated as follows: (0) aim not reached; (1) aim partially reached; (2) aim reached. Coding the results for each aim in this way enables a comparison of effectiveness. When there was no mention of a certain aim, or the actions taken in order to address a certain aim were completely irrelevant or wrongfully executed, this is considered as ‘aim not reached’. When a certain aim was treated by the network (or part of it), but no adequate actions were taken, or actions taken were not satisfactory as more or other strategies would have been necessary, this is considered as ‘aim partially reached’. When the network adequately addressed and fulfilled a certain aim, this is considered ‘aim reached’. In certain scenarios below, a more meticulous rating is needed, when a goal is more than merely ‘partially reached’, but simultaneously not fully reached. This is indicated with ‘1-2’. For each case observation, the following coding scheme was used:

MKO/CoPI training dd mm yy Goal Rating 1 0 2 1 3 2 4 0 5 1

(31)

30

6 2

7 1

Example of coding scheme

The lowest possible score would be 0, the highest possible score 14. A CoPI was deemed effective with a minimum score of 11: an average of around 1,5 per goal. In that way, while allowing for a non-perfect score on a few goals which can be deemed inherent to chaotic crisis situations, the average rating had to vary between ‘aim partially reached’ and ‘aim reached’.

3.5 Validity & reliability

Yin (2003) outlines several tests that portray the soundness of case studies, centering around two main concepts: validity and reliability. Validity concerns whether what was measured was indeed what the researcher attempted to measure, and it consists of several components: construct, internal and external validity. Construct validity occurs when the operational measures used correctly reflect a study’s key concepts (Yin, 2003: 34). The key concepts have been conceptualized and various sub concepts were distinguished based on existing literature relating to security or even crisis response networks. For each of these sub concepts a number of indicators has been determined, as included in the framework presented in 3.4.1. Internal validity regards whether the independent variable(s) distinguished are in fact the factors influencing the dependent variable. This issue has been partially addressed by undertaking a literature study in order to find more factors possibly influencing network effectiveness than the five found by Whelan (2016). Four other potential factors were found and added to the concept framework. During the case observations, the researcher was aware of the possible existence of other factors and actively attempted to distinguish any possible determinants of network effectiveness that did not fit in any of the existing categories. By using triangulation (Neuman, 2014: 166) through utilization of various different methods (analysis of documentation, case observations, informal short consultations with experts in the field during the cases observed), the internal validity was optimized. External validity concerns the generalizability of the conclusions. Generalizability in qualitative studies with a limited number of cases is generally limited, but in order to improve the validity, a total of seven scenarios was observed. Although they were no real crisis situations, cases were chosen based on

(32)

31

whether the simulations resembled reality as much as possible. Relations observed during these cases are therefore understood to correspond with relations during real crisis situations. While this study does not foresee in an exact measurement of the amount of influence per factor, which may be seen as a weakness, its aim is not to establish such a quantitative calculable scheme. Instead, it studies which factors are and which are not of influence on a crisis response network’s effectiveness.

Reliability is warranted when a study into the same causal mechanism of independent variables influencing a dependent variable conducted at a different point in time or using different methods would produce the same, or at least highly similar results. As the framework of factors is based upon a large foundation of network (effectiveness) literature, reliability of the factors analyzed was optimized. By observing cases rather than basing results on policy papers only, the actual cooperation dynamics could be analyzed rather than a ‘paper reality’ or guidelines for cooperation representing an ideal state instead of the factual situation. The cases observed were not real crisis situations, but simulated scenarios. The scenarios closely resembled real-life situations, though, which was confirmed by participants.

(33)

32

4 Analysis

4.1 Case 1

The first case was a CoPI training day covering three scenarios, organized by Veiligheidsregio Kennemerland. The scenarios were as follows:

- A ship was stuck a few hundred meters from the Dutch coastline. Cargo was falling off, and soon was revealed that they contained hazardous material. Meanwhile, many civilians were present at the beach and a large event was planned for a few hours later.

- A family that had sought asylum in the Netherlands was scheduled to be deported from their house back to their home country. Groups (both opponents and proponents) had gathered and started riots in the neighborhood.

- A fire had started in a parking garage. The fire was hard to control and rapidly increased. One person was found dead and several were heavily wounded. On top of the parking garage were a few layers of apartments where elderly and needy people were housed, who had to be evacuated.

4.1.1 Effectiveness of the crisis response network

Effectiveness of the CoPI’s was assessed based on seven tasks as set out in 3.4.2. CoPI/MKO goals:

1. Defining the situation and possible developments at the incident location (multidisciplinary);

2. Deciding on a multidisciplinary, coordinated strategy for rapid emergency assistance and/or incident control;

3. Leading the disciplines taking part in the crisis response at the incident location; 4. Providing the operational units at the incident location with information about the

incident control;

5. Providing the mayor (and in cases of GRIP 2 or higher: the ROT) with (multidisciplinary) information about developments at the incident location;

6. Taking measures aimed at preventing and/or limiting the number of casualties at the incident location;

(34)

33

Each crisis simulation was rated after the case was evaluated, based on observations and remarks made by the network members. The first scenario was evaluated as follows: CoPI training 28 03 19 (1) Goal Rating 1 1-2 2 2 3 2 4 n/a 5 n/a 6 2 7 n/a Total 7,5

As the first case was concluded immediately after the first CoPI meeting, observations regarding several goals could not be made (indicated with ‘n/a’). The other goals, however, were addressed efficiently. The situation and possible developments were defined rapidly and correctly, although a short discussion between the OvD-G on the one side and the OvD-B and OvD-P on the other led to a short delay (10:45)4_{. The}

organizations involved in crisis response soon came to a coordinated strategy where they all carried out their own tasks and responsibilities as individual organizations, as well as in support for the others. The OvD-B was concerned with defining the (toxic) contents of the cargo approaching the coastline and cleaning it safely; the OvD-G prepared medical sites and teams that could provide aid in case of casualties, supported in this by the OvD-BZ, who took care of communication to the citizens living in vicinity of the beach; and the OvD-P cleared the beach of people so they would not be harmed by possible toxic substances and so that the fire department could work efficiently. The non-operational members of the team (IM, plotter, CoPI leader) formulated a clear image of the situation and supported the process of decision-making during the CoPI meeting (10:45). Before the CoPI session (in MKO’s) and immediately after, the OvD’s led their disciplines in response to the shared image of the situation and needs of other organizations. Measures taken directly aimed at the prevention of casualties, which was successful: no victims were

4_{Throughout the analysis, time indications like this will refer to the corresponding timestamps indicated in the}

(35)

34

registered. Although not all seven goals could be evaluated, the four that could were all rated positively. During the evaluation of the case, CoPI members and evaluating staff stated they believed the situation was under control, and that cooperation had been successful.

The second case involved two CoPI meetings and therefore allowed for a more complete observation. The ratings were assigned as presented in the table:

CoPI training 28 03 19 (2) Goal Rating 1 1 2 2 3 2 4 2 5 1 6 1 7 2 Total 11

The case at hand was a rather complicated one: riots had started in a neighborhood, creating an unsafe situation for both people living there and emergency services entering the neighborhood to provide aid or support. Although cooperation and coordination went rather smoothly, a piece of vital information missing from the operational image of the situation influenced the ratings drastically. Matter of concern were a few schools in the neighborhood that were about to end; would it be safe for the children (and their parents for that matter) to leave the school grounds and go home? During the CoPI meeting, the network estimated it to be safe: the OvD-P had not received any signals of threat directed at the children and all members agreed that keeping the situation as close to normal as possible would be preferable (13:26). It was only afterwards, during the evaluation, that the OvD-BZ shared with the others that he had received information during the case that riots occurred around, and even on school grounds, too (13:52). The OvD-P immediately stated that had they known that, they would have operated very differently (idem). The safety of the children and their parents was, in hindsight, not guaranteed: a great point of concern. Apart from that, the CoPI functioned rather well. A coordinated strategy was come up with rapidly (for a large part directed by the OvD-P), which was necessary as police escorts were required to join other emergency services entering the neighborhood in

The effectiveness of Dutch crisis response networks: A study into factors influencing the effectiveness of operational networks within the Dutch crisis response system

Lidy-Anne Verkade - s1299794