1
Faculty of Governance and Global Affairs
“Failures in Agility when Responding to Forest Fires – a
study into the 2017 Pedrógão Grande fires”
Cristina da Costa Ferreira
S2004453
January 13th, 2019
Master’s Thesis Crisis and Security Management
Supervisor: Dr. Gabriele Landucci
Second reader: Dr. W. G. Broekema
2 Index: CHAPTER 1 – INTRODUCTION ... 5 1. Problem outline... 5 2. Research Question ... 6 3. Academic Relevance ... 7 4. Societal Relevance ... 8 5. Reading Guide ... 8
CHAPTER 2 – THEORETICAL FRAMEWORK ... 10
1. Conceptualization of Atypical Accident Scenarios ... 10
2. Agility ... 12
2. 1. Agility: conceptual confusion and lack of empirical evidence ... 12
2. 2. Command and Control Agility ... 14
2..2. 1. C2 Agility Conceptual Framework - NATO NATO SAS-085 (2014) ... 15
2 . 3 Unified Theoretical Model of Organizational Agility - Holsapple & Li (2008) ... 16
2. 4 Organisational Behaviour at the emergency stage - Quarantelli (1988) ... 19
2.5 Efficiency and Effectiveness in Agile Crisis Response Organizations: van Veelen et al. 2006. ... 24
2.5. A review of the literature ... 26
CHAPTER 3 – METHODOLOGY ... 28
1. Research design ... 28
3
Identifying the factors that contributed to a lack of Agility ... 32
2. Case study ... 34
Single case study ... 34
Choice of case selection ... 35
3. Data collection and Analysis ... 35
Origin of data ... 35
Data Analysis ... 38
4. Operationalization & Conceptualization ... 38
5) Reliability and Validity ... 46
CHAPTER 4 – ANALYSIS ... 48
1 A (very short) summary of the legal framework & command institutions ... 48
2. Analysing Agility ... 50
Phase 1 - (17th June, 14h35 to 17th June, 19h00) ... 51
Phase 2 (17th June, 19h01, to 18th June 19h00) ... 57
Phase 3 (18th June, 19h01 to 21st of June 23h59) ... 59
3. Factors that inhibit agility ... 61
A) Intra-organizational Information Flow... 62
d) Information Flow from the Public to the Different Organizations ... 68
F) Personnel Burn-out ... 69
L) Impact of Disaster Magnitude ... 70
4
Answering the Research Question. ... 75
Bibliography ... 79
Acronyms ... 89
List of Figures ... 93
LIST OF TABLES ... 94
Annex 1: The components of agility by phase ... 96
Phase 1 ... 96
Phase 2 ... 97
Phase 3 ... 98
Annex 2 – Phases of the SGO ... 100
5 CHAPTER 1 – INTRODUCTION
1. PROBLEM OUTLINE
In 2017, between the 17th and the 24th of June, Portugal suffered a series of extreme forest fires. This Pedrógão Grande complex of fires, claimed the lives of 67 people, most of whom died trying to escape the scene by car. These fires required assistance at the European level in order to be controlled. In addition to the largest loss of life to wildfire in Portuguese history, 200 people were also injured, over 500 homes destroyed, and a total area of 45,328 hectares burnt (Centro de Estudos sobre Incêndios Florestais da Universidade de Coimbra (CEIF), 2017, pp. 5-7).
These events arose from five different fires with distinct origins that were ignited about the same time, in the same region, and eventually merged into a contiguous area. The fire that be-gan in Escalos Fundeiros e Regadas within the concelho (municipality) of Pedrógão Grande, was the most devastating, burning half the total area. According to an important independent report commissioned by parliament, it was later reported that the National Authority for Civil Protection (ANCP) knew that a fire in this location could be highly impactful. However, lack of alertness along the command chain, as well as a general lack of resources, meant that ini-tially not enough attention was given to Escalos Fundeiros (CEIF, 2017, 5-7, 15). It should be noted that on the very same day of the Pedrógão Grande Fires, there were over 100 addi-tional ignitions, leading to a dispersion of resources and personnel and complicating sense and decision-making.
The high number of ignitions were in part a result of the extreme weather conditions regis-tered that month, which included a severe heatwave mid-June, increasing the risk of fire throughout Portugal. The firefighting on the 17th June was greatly impacted by the interaction of the flames with a thunderstorm that occurred around 19h00, producing horizontal convec-tive currents. These in turn caused abnormally high speed fire plume flow and impacted wind velocity, thereby giving rise to extraordinarily severe fires. This escalated the situation and made it much harsher to combat (CEIF, 2017, pp. 56-57).
6
In spite of these adverse conditions, with all eyes turned on the record-breaking Pedrógão fires, the public and the media demanded transparency about what happened during the oper-ational response. As a result, the government undertook urgent measures to understand the course of events, producing a number of reports that are accessible to the public. While rec-ognizing the difficulty of the situation, the documents note slow reaction times, lack of coor-dination among operatives, the frequent collapse of the communication system (SIRESP), among other problems (CEIF, 2017; CTI, 2017; CEIPC, 2017). In the context of Portuguese emergency response, these documents are detailed in an unprecedented way, and for the first time allow the analysis of the emergency response during wildfires. As such, my research question is:
• “What are the factors that interfere with agility in emergency response to wildfires?”
2. RESEARCH QUESTION
My research question concerns the subject of agility. The thesis has two expectations: first, that while agility is, in principal, possible during the context of wildfire emergency response, this was not the case in Pedrógão Grande; second, that agility within emergency response in wildfires can only be achieved when organizational structures are constructed taking into ac-count and allowing for agility. With this in mind, I will first study which aspects of agility where achieved in the emergency response in the Pedrógão Grande fires, and then assess which factors hindered its manifestation.
To answer the research question, the thesis will use a simplification of the concept of agility as used by Holsapple & Li: “the ability to converge alertness to change (recognising opportu-nities/challenge) – both internal and external - with the capability to use resources in a re-sponding (proactive/reactive) manner”. The reasoning behind this will be expanded upon in the third section of this thesis, which discusses the definitions of agility presented in the theo-retical framework.
This thesis defends that agility needs to be explored further within the paradigms of crisis management. This is in keeping with the literature (Holsapple & Li, 2008; Woltjer et al.,
7
2015; Lai, 2017; Wikberg et al., 2016; Wikberg et al., 2017), which states that more studies providing empirical data, as well as theory, are needed linking the two.
The first part of the thesis concerns itself with which aspects of agility were achieved in each of the phases of crisis response at Pedrógão Grande, by adapting Holsapple & Li’s work-de-sign perspective on organizational agility to emergency response during wildfires (Holsapple & Li, 2008). In order to do so, some inspiration was drawn from the field of command and control (C2) agility, in particular as developed by NATO SAS-085. It will be used to comple-ment Holsapple & Li because it attributes a sense of urgency to choosing a correct agility ap-proach that translates importantly to wildfire emergency response, but which was not as in-trinsic to Holsapple & Li.
The research question will be answered within the framework described by Quarantelli (1988), who discerns that within emergency organizations, management issues usually can be categorized in to three areas: the communication process and information flow; the exercise of authority and decision making; and the development of co-ordination and loosening of the command structures.
3. ACADEMIC RELEVANCE
Academic relevance in this thesis is two-fold. On the one hand, as aforementioned, although
the concept of agility is used within crisis management, there is very little theory as well as empirical studies, to connect the two aspects. Agility is often confused, or used interchangea-bly with other concepts such as adaptability, flexibility, etc. My thesis addresses this gap in the literature by drawing a theoretical framework relating the concepts, and analysing agility in the response to Pedrógão Grande.
The second reason my research is academically relevant is because there are no studies re-garding operational response in Portugal. This can be confirmed by resourcing the Scopus and Google Scholar databases, showing that response activity during the critical emer-gency period has not been previously studied in Portugal. In a comprehensive paper
Fer-8
nandes et al. (2014) pointed out this research gap. This is probably because prior to 2017, re-ports proving a detailed minute-by-minute description of the emergency response were not available. In this respect, the reports commissioned for Pedrógão Grande are noteworthy.
4. SOCIETAL RELEVANCE
In Portugal, desertification is one of the main underlying causes of the large scale of fires in sparsely populated rural areas. This is because abandoned agricultural fields accumulate un-harvested biomass that fuel the fires and threaten the wildland-urban interface. The remaining elderly population are unable to work or clear the land, much less contribute to the manage-ment of fuel breaks – something that will only get worse in the future as a result of Portugal’s ageing population.
Thus, fires such as the one in Pedrógão Grande are a serious setback for the rural develop-ment agenda. Furthermore, the costs of suppression, as well as the costs for society, have been huge. On average, since the start of the millennium, three fold more is spent a year sup-pressing fires, than preventing them – 78 million Euros, compared with 25 million Euros. This can also be contrasted with the total cost to society which has varied significantly over the same time period, between a minimum of 127 million Euros (2008) and a maximum of 1.3 billion Euros (2003) (CTI, 2017, 36-38).
Fernandes et., al suggest that one significant improvement to minimize area burnt would be to incorporate operational improvements that lead to faster containment (Fernandes et. al., 2016, pp. 1). In short, societal relevance of my thesis comes from how studying operational re-sponse can hopefully impact practical rere-sponse, therefore minimizing the damage caused by fires in Portugal.
5. READING GUIDE
The thesis will be divided as follows: (1) The previous section provided a brief introduction on the topic I will research, as well as on its social and academic relevance. (2) The theoreti-cal framework provides an overview of the academic literature on agility, focusing on its
9
multi-disciplinary nature. (3) The methodology includes the research design, a justification of the cases selected and the relevance of the theories used, as well as the operationalization of the key concepts. (4) The analysis is divided in two parts. The first discerns whether agility was reached in the response to the Pedrógão Grande fires, while the second part studies the factors that interfere with agility. (5) Finally, the thesis ends with a conclusion reflecting on the findings, discussing the limitations of the research, and, of course, answering the research question - What are the factors that interfere with agility in emergency response to wildfires?
10 CHAPTER 2 – THEORETICAL FRAMEWORK
1. CONCEPTUALIZATION OF ATYPICAL ACCIDENT SCENARIOS
The literature distinguishes between “normal accidents” and “atypical accidents”. The latter are described by Paltrinieri & Khan as events that were disregarded during the hazard identi-fication processes “because they deviated from normal expectations of unwanted events or worst case reference scenarios” (Paltrinieri & Khan, 2016, pp. 13). Although, ideally, reason-able doubt would increase as safety professionals began to recognize early warnings, in the case of an atypical accident available information is disregarded (Paltrinieri & Khan, 2016, pp. 14).
The authors further distinguish atypical incidents between Black Swans and Dragon Kings, whereas Perfect Storms are addressed in brief. The first is an analogy to the well known inci-dent of how, until the 17th of century, all swans where presumed to be white, until the sight-ing of the first black one. Consequently, in crisis management, a black swan is a rare event that has never been encountered before and has three characteristics: It is rare as it was until then, never considered a possibility; it has extreme consequences; it can only be explained af-ter it has occurred and cannot be anticipated. As a result, Black Swans shataf-ter predictability and current assumptions. On the other hand, dragon kings are extreme events which can be identified as outliers in the power law distribution. Dragon kings occur normally as a collec-tion of several small fractures that together lead to an extraordinary amplificacollec-tion of the event, even though these failures could have been potentially identified, and the accident pre-dicted (Paltrinieri et al., 2016, pp. 16-18).
Finally, perfect storms are briefly described as events where mostly aleatory uncertainties act in conjunction with rare but known events (Paltrinieri et al., 2016, pp. 17, 19). According to Maes & Dunn, this is seen as “an extremely bad roll of dice” where even if the underlying events are understood a priori, their very rare conjunction is very unlikely and subject to con-siderable uncertainty (Maes & Dunn, 2017, pp. 19).The Fukushima Daiichi incident is given as an example. Although the nuclear plant was designed to sustain earthquake damage, the fast succession of an earthquake with a tsunami that caused flooding, energy failure, and
11
damage to the back-up energy system, was not considered. This combination caused the reac-tors to overheat, and ultimately, the release of high levels of radiation (Maes & Dann, 2017, pp. 26).
Atypical accidents were described above in order to qualify the Pedrógão Grande incident as atypical and, therefore assert that it unsettled pre-conceived notions as to how to engage in an effective response. This increases the significance of analysing the event in the context of agility, as it tests whether institutions were able to to cope with changes in circumstances. At the same time, it is relevant to discuss Pedrógão as an atypical accident because of the con-siderable damage caused, which impacted society at different levels.
While there is agreement in the literature that there is some leeway as to whether incidents can be categorised as Black Swans, Dragon Kings or Perfect Storms (Maes & Dunn, 2017, pp. 21; Paltrinieri & Khan, 2016, pp. 17), in this thesis the Pedrógão Grande incident is seen as a perfect storm. It combines uncertainties with rare but known events to cause a record-breaking amount of casualties. Thus, while there was awareness that Pedrógão was excep-tionally sensitive because of the accumulated biomass caused by human negligence, as well as its geographical location which would facilitate fire propagation, it is also true that no matter the response, the events would not have become so uncontrolled had it not been for the unlikely conjugation of climatic conditions.
Thus, while the interaction of the flames with severe atmospheric instability, causing the col-lapse of the convective column and forming a pyrocummulus cloud is a known possibility, this was not foreseeable in the case of Pedrógão as these cases are very rare and for the most part the technology does not exist in Portugal or in Europe to predict such scenarios (CTI, 2017, pp. 89). The formation of pyrocummulus clouds can only occur when the weather is extremely dry, but the atmosphere is humid, and cause erratic weather and a column of smoke that hampers visibility for firefighters (CTI, 2017, pp. 81). In the case of Pedrógão Grande, this cloud reached almost 13 km at its peak, and was associated with the downburst that occurred at the same moment, which meant that high intensity winds moved vertically in the direction of the ground, spreading the fire in all directions of the radius of the impact (Agência Lusa, 2017).
12
According to IPMA, the official national authority on weather and meteorology, the incident was caused by “the conjugation of the aforementioned factors with local effects related to at-mospheric convection phenomena, connected to the low humidity”, leading to a complex and exceptional situation (Agência Lusa, 2017). Furthermore, it is impossible, from a technologi-cal standpoint, to accurately predict exactly what, where, when and with what intensity ab-normal atmospherically events would occur (IPMA, 2017, 37-40). The full extent of the me-teorological conditions were only registered a posteriori.
2. AGILITY
Crisis management situations are for the most part extraordinary events that modify the pace and the demands of a normal situation. As such, this rupture with normal expectations means that circumstances are often complex and dynamic, and require individuals to make optimal decisions under constraints of high risk, uncertainty, high work-load and pressure – which in turn, make the situations themselves cognitively demanding. Agility, therefore, is normally seen as the ability to cope with changing circumstances. It is therefore a valuable asset to or-ganizations and inter-organizational structures during crisis management.
Agility appears in the literature across disciplines, most of which are to some extent related to each other (Woltjer et al., 2017, pp.1). These include military command and control, organi-zation theory, and crisis management, all of which will be expanded on below, after a brief discussion on the conceptualization of agility.
2. 1. AGILITY: CONCEPTUAL CONFUSION AND LACK OF EMPIRICAL EVIDENCE
There is a significant overlap in the literature as to whether the concept described above should be referred to as agility or as adaptability/adaptivity. This is acknowledged in Harrald, 2006; Holsapple & Li, 2008; Lai, 2018; Wikberg et al., 2016; Wikberg et al., 2017; Woljer et
al., 2015, who state that the terms are often used interchangeably. To Lai, while the literature
13
agility, and lessons learnt, these are scattered across various disciplines (Lai, 2018, pp. 460). As such, there is a limited theoretical work on what it takes to develop this skill, especially as findings are clouded by definitional confusion regarding the concept itself. Consequently, it is unclear as to how a taskforce can develop this skill, and how it can be successfully applied to organizations, particularly during extreme events. This means that organizational agility must be explored from a capacity approach, so as to study how organizations find solutions to pressing problems (Lai, 2018, pp. 460-462).
Overall, it appears to me that the lack of conceptual cohesion means that authors disagree as to what are its key attributes. A literature search shows that the concept is often associated with other terminology, such as “agile interoperability” (Barthe-Delanoe et al., 2013), “agile coordination”(van Veelen et al., 2006) and “agile decision-making” (Devine et al., 2015), without it being clear how this all relates together into agility in itself. Furthermore, there are occasions where such a connection is implicit, rather than explicit (Quarantelli, 1988). Thus, while Lai characterizes agility as “to cope or ingeniously adapt to a set of circum-stances or devising resourceful solution to intractable problems” (Lai, 2018, pp. 460), Wik-berg et al., instead adopt the term adaptability/adaptivity, and use the definition given in NATO SAS-085, a document developed by NATO to research command and control agility, to describe agility (Wikberg et al., 2016, pp.2; Wikberg et al., 2017, pp. 205), therefore es-sentially equating the two terms, and stating that “the broad definition covers aspects such as changing processes, culture, equipment, tactics, leadership, etc” (Wikberg et al., 2017, pp. 204).
Interestingly, within the framework of C2 Agility, adaptability is discussed as one of six in-terdependent enablers of agility, alongside responsiveness, versatility, flexibility, resilience and innovativeness, without which agility could not be manifested. Adaptability is hence characterized as “the ability to change the organization and/or work process” (NATO SAS-085, pp. 199). In short, such a definition falls short of C2 theorists’ understanding of agility as the capability to effect, cope with and/or exploit changes in circumstances, as it merely in-dicates that change is possible, but not whether it is adequately applied at a certain point in time.
14
In summary, this sub-section within the theoretical framework has supported my earlier claim that literature linking agility to crisis management is incomplete. This thesis, nonetheless, chooses to focus on agility, rather than adaptability or any other related terminology, as its main focus. The following subsections develop the idea of how to assess agility within differ-ent schools of thought, specifically command and control theory, organization theory and cri-sis/disaster management. These will be followed by a brief discussion comparing these per-spectives.
2. 2. COMMAND AND CONTROL AGILITY
While C2 Agility is not the focus of this thesis, it will be expanded on as C2 researchers have conducted extensive research on agility. This can be justified in that C2 has traditionally been a predominantly military field, and has been amply funded by its interested parties, with much of the literature on C2 Agility having been published by the USA Department of De-fence, such as Alberts et al., (2007), Alberts & Hayes, (2006), Soykan & Alberts, (2015), among others. Thus, I defend an overview on C2 Agility as one of the main developers of agility theory, with its conceptualizations being used outside of the military, such as in (Wik-berg et al., 2017; Wik(Wik-berg et al., 2016), as well within the field of economics.
C2 Agility theory grew from the concept of command and control (C2), which originated as a military practice in the 19th century, designating the hierarchy between a commander, who would recognize and dictate what needed to be done, and the strict obedience of the troops (Alberts et al., 2015, p. 23). In contrast, for the last few decades, C2 scholars have increas-ingly supported that hierarchy in command and control is insufficient to achieve success in the 21st century, due to the increased complexity of this new world resulting from technolog-ical advances. Network capacities mean that the speed at which information can be sent/re-ceived and subsequently processed has become a competitive advantage, and there is in-creased exposure to public opinion. This gave birth to C2 Agility, the idea that in order to maximize efficiency, C2 should not be rigid, but agile. This does not mean that traditional hi-erarchy is never the appropriate approach, but rather that it has become applicable only to certain, increasingly rare, situations (Alberts, 2007, pp. 14-17).
15
For C2 Agility theorists, mission success in today’s conjuncture requires a variety of entities, from different areas of expertise, with interrelated or even competitive stakes who organize themselves with a dynamic chain of command according to what might be most effective in a certain situation (Alberts & Hayes, 2006; Cogan & Waddell, 2009; Essens, Spaans &
Treurniet, 2007; Soykan & Alberts, 2015; Alberts et al., 2015). C2 Agility is therefore de-fined as the selection of an approach to command and control that is appropriate to the mis-sion, the force and its capabilities, and the operational environment; and making appropriate adjustments when these factors change (Alberts et al., 2015, p. 15). Agility, on the other hand, is defined as “the capability to successfully effect, cope and/or exploit changes in the circumstances” (NATO SAS-085, 2014, pp. 199). Therefore, these two ideas differ in that there needs to be agility as a whole for command and control agility to manifest; however agility is not dependent on the configuration of the command ladder within a colligation or organisation, but rather with that of the mission itself.
2..2. 1. C2 AGILITY CONCEPTUAL FRAMEWORK - NATO NATO SAS-085 (2014)
C2 Agility theory has been extensively developed by the NATO military alliance in the NATO SAS-085 Final Report. This document is the state-of-the-art within C2 Agility The-ory, and is based on the works of C2 authors over the last two decades, particularly from the work of David S. Alberts (Alberts & Hayes, 2003; Alberts & Hayes, 2006; Alberts, 2007, among others). Adopting the contributes of its predecessors, NATO SAS-085 identified that Agility is composed of six enablers, which are versatility, responsiveness, robustness, flexi-bility, resilience, adaptability and innovation. It distinguished five increasingly networked C2 Approaches: Conflicted C2, De-Conflicted C2, Coordinated C2, Collaborative C2 and Edge. From there, NATO SAS-085 developed the C2 Agility Conceptual Model, which aimed to evaluate a number of testable hypotheses as well as the importance of C2 Agility by using empirical evidence based on retroactive case-studies, as well as simulation-based experi-ments. It concluded that C2 Agility was desirable as it improved the likelihood of mission success (NATO SAS-085, 2014, pp. 22). As C2 Agility aims to respond to the 21st century in a way that traditional C2 simply cannot, the NATO SAS-085 working paper focuses heavily on complex endeavours, i.e. non-traditional military threats that involve a number of different
16
entities or a collective. Ten highly heterogeneous case studies were therefore studied, ranging from peace-keeping to cyberwarfare. (NATO SAS-085, 2014, pp. 153). In short, the case-study methodology was created taking into account that, in order for it to be useful, it should be applicable to most situations that involve C2 Agility. This in turn would allow for the dif-ferent case studies to be comparable to each other.
As a result, the authors of NATO SAS-085 developed both an evidence table to be filled by generic indicators (i.e. yes, no, high, low...), that divided the case studies into temporal phases, to be decided as appropriate for each, as well as a case-study template, where ques-tions would be asked to discuss the specifics of each case (NATO SAS-085, 2014, p. 154-156).
2 . 3 UNIFIED THEORETICAL MODEL OF ORGANIZATIONAL AGILITY - HOLSAPPLE & LI (2008)
Holsapple & Li (2008) propose a unified theoretical model of organisational agility that in-vestigates the attributes of knowledge intensive work-design system (Holsapple & Li, 2008, pp.2). The authors defend that since a work-design perspective refers to the way in which work routines are arranged and renewed, agility is shaped by work-design. Furthermore, since work-design, knowingly or not, occurs in all types of organisations a unified theoretical model based on it should be applicable to all organisations (Holsapple & Li, 2008, pp. 3). That being said, their work is clearly oriented in the direction of business and commerce, with several allusions to “the extended enterprise chain”, “the marketplace”, and “the busi-ness environment” (Holsapple & Li, 2008, pp. 7), among others.
Although it is recognized that the literature on agility is very limited and that the factors that contribute to agility are disputed among scholars, this theory aims to answer the question of how agility can be reached and maintained by organisations (Holsapple & Li, 2008, pp. 2).The authors argue that in order to be agile, a work-design system aims to “sustain superior performance by dynamically arranging work so as to ride atop the waves of change and main-tain balance in weathering the inevitable, unsettling, novel storms that can strike any organi-sation – often with little warning” (Holsapple and Li, 2008, pp. 4) After a review on the
con-17
cepts of agility, specifically within the contexts of entrepreneurship and strategic manage-ment research, the authors introduce what they see as a general purpose definition of agility. Thus, agility is defined as “the result of integrating alertness to changes (recognising opportu-nities/challenges) – both internal and environmental – with a capability to use resources in re-sponding (proactive/reactive) to such changes, all in a timely, flexible, affordable, relevant manner” (Holsapple & Li, 2014, pp. 6).
Two dimensions are presented in this definition: response-capability and alertness. Alertness is seen as an opportunity-seeking capability with both internal and external vantage points. Within alertness lie two specific capabilities, strategic foresight, the ability to anticipate dis-continuities in the business environment and the marketplace, as well as impending disruptive moves by competitors, and systemic insight, the capacity to see the interconnections between an organisation's capabilities. Response-capability is the capability to select and enable ac-tions. It is formed by value evaluation, coordination and integration, organisational learning and reconfiguration (Holsapple & Li, 2014, pp. 7).
18
In addition, Holsapple & Li define agility effectiveness in terms of four elements: timeliness, flexibility, relevance and affordability. Timeliness is different from speed, and means action at the appropriate time. Flexibility refers to the range of ways available to achieve success. Relevance accounts for the fact that just for a change to be worthwhile, it has to be relevant, and not just change for change’s sake. Affordability recognises that there are costs associated to change; if these outweigh the relevance then agility becomes impractical. All four of the measures of effectiveness relate to agility, but on a more detailed level, the measures apply directly to each component of alertness and response-capability. For instance, timeliness of coordination refers to the appropriate timing in the management of the dependencies, while flexible coordination refers to the ranges of ways available to manage the dependencies (Holsapple & Li, 2008, 7-10) (Fig 2.).
FIGURE 2 A TAXONOMY OF AGILITY (HOLSAPPLE & LI, 2008, PP. 9)
Finally, a work-design model is produced, portraying three design levels: operational design, which is concerned with the way work episodes are initiated, performed and terminated in re-lation to exterior action; episodic design, with structuring and governing operational work-design; and, finally, strategic design when the execution does not demand rigid adherence to what has been planned at an operational level, but is subject to modification regarding the particular episode. As all three levels relate back to each other, the theory postulates that an organisation's agility is influenced by work-design at each of the three levels (Holsapple & Li, 2008, pp. 10-12).
19
FIGURE 3 A WORK-DESIGN MODEL OF ORGANIZATIONAL AGILITY (HOLSAPPLE & LI, 2008, PP 12)
2. 4 ORGANISATIONAL BEHAVIOUR AT THE EMERGENCY STAGE - QUARANTELLI (1988)
While not explicitly a theory on agility, Quarantelli (1988) provides an interesting framework summarising organisational behaviour at the emergency stage of community disasters. I sug-gest that to some extent agility is implicit; my assumption is based on Quarantelli’s criticism of the belief that disaster planning in itself will prevent a crisis or an emergency. Rather, the author states instead that although there are known tactics that can be applied for good crisis management, “there are always situational factors or other contingencies which require par-ticular adjustments to attain a specific goal if the overall objective is to be achieved” (Quar-antelli, 1988, pp 375). Furthermore, the author considers it an error to assume that central-ised control has to be imposed from the top down in emergency situations, but that loosening rather than tightening the command structure is best for emergency periods (although not necessarily for other phases). As a result, co-ordination and not control needs to be empha-sised in these situations (Quarantelli, 1988, pp. 381).
20
To the author, “good crisis management, to a considerable extent, is the application of tactics which are specifically relevant to the situational contingencies of a given community disas-ter” (Quarantelli, 1988, pp. 375). The author states that a general theme of this article is the need for disaster planners and managers to “operate in the real world”. Thus, while Quaran-telli emphasizes that preplanning is necessary and can maximize overall organizational coor-dination, he also defends that many of the problems that occur during a disaster fall are prob-lems that can only be solved via tactics (Quarantelli, 1988, pp. 365, 383-384)
Quarantelli identifies that in the typical community disaster, management problems usually occur in relation to three issues: the communications process and information flow; the exer-cise of authority and decision making; and the development of coordination and loosening of the command structure. Although community disasters can refer to many types of collective stress situations, in this context they are seen by Quarantelli as referring exclusively to the “consensus type community crises generated by natural or technological agents”. These ex-clude conflict type situations such as wars riots or terrorist attacks, as well as non-community kinds of disasters, such as transportation actions that do not impact the functioning of a com-munity (Quarantelli, 1988, pp. 373).
Problems in the communication process mean that while the processing of information may be present, the information sent out does not meet the requirements of the situation, therefore undermining efforts for preparedness and the efficiency of the operation. The author notes that while, during disasters, there may be damage to communication equipment or system overloads, failures in communication are normally the result of poor, incomplete and/or inef-ficient information flow. The author gives five categories of information failure problems. (Quarantelli, 1988, pp. 375-379):
1) Intra-organizational information flow: Which can occur for various reasons during an emergency, namely, due to the adoption of non-routine tasks that create confusion and opera-tional problems.
2) Information flow between organisations: In times of crisis it is often difficult to initiate and maintain communication lines with other entities with which, sometimes, there were no
pre-21
disaster relationships. Prior panning can help mitigate this issue, but training has to empha-sise that disaster managers must be capable of having to work with unfamiliar officials and groups.
3) Information flow from organisations to the general public: Although messages issued from organisations to the general public (e.g. evacuation notices, etc.) must be clear in order to have the desired effect, they are, according to the author, often handled poorly and rele-gated to second status priority. This can be highly problematic, as it forces the public to infer its own assumptions regarding the extent of the danger and how to respond, potentially leadin to dangerous situations.
4) Information flow from the public to different organisations: While communication with the public creates a new source of information, for instance, requesting aid, a crisis can create an overload of public communication and delays in response in situations that require urgent actions.
5) Information flow within different systems of organisations: It can often be difficult to ini-tiate communication between sets of interrelated, specialised information systems, due to a lack of knowledge of whom is most appropriate to contact in a particular moment. Further-more, while awareness of protocol is important, during a crisis there is an emergent and ad-hoc quality to conducting tasks, and decision-making key-decision making points may shift. An example would be the differences between medical systems and systems that provide se-curity; rather than a one way information flow among the participating organisations, there are multiple chain and communications between the various multi-layered groups. In conclu-sion, the author suggests that it is easier to cope with information flow problems that consist mostly of vertically linked instead of horizontally linked sub-units.
According to Quarantelli, while during crisis, chains of command and lines-of authority sel-dom break in established organisations, and officials usually exercise their formal authority, some factors interfere with the exercise of authority and decision-making. Four problem areas arise (Quarantelli, 1988, pp. 380-381):
22
1) Personnel Burnout: Decision makers who remain on the job for too long will eventually give in to exhaustion or become inefficient as a result theoff. However, when they are re-placed and allowed to rest, the next person will lack certain information that has not been for-mally recorded, especially as part of the strategy may only exist in the mind of the predeces-sor. This can be dealt with, in part, in pre-planning, by ensuring a time-cap on work-shifts. 2) Organisational Authority Conflicts: It can be difficult to ascertain who has the organisa-tional authority in a certain situation and when there are new disaster tasks questions arise about which organisations have the authority to assume them. Once more, while to some ex-tent this can be dealt with in pre-planning, in reality the matter often has to be attended to in an ad-hoc manner.
3) Organisational Domain Conflicts: Authority and decision-making surrounding the perfor-mance of traditional tasks sometimes arises between established organisations and emergent groups. For example, in disaster relief, if a non-local welfare agency is mandated to act with the locals, it’s authority may be questioned and it may be viewed as an intruder to the domain of the local agencies. The same may occur if there is a rivalry between agencies, such as, be-tween the local police and a private security company.
4) Unresolved jurisdiction issues: While during non-crisis periods, unclear, or overlapping authority and responsibilities can be ignored, a disaster often cuts across jurisdictional bound-aries of local organisations, which can significantly slow down the pace when immediate ac-tion is of utmost order. A soluac-tion may be to obtain temporary consensus, with the under-standing that there will be no formal carryover into the recovery period.
According to Quarantelli, traditional command and control organisation makes incorrect as-sumptions about what is needed in a disaster. In contrast, coordination rather than control is desired for emergency situations, as it allows for loosening rather than tightening of the com-mand structure. Nonetheless, this is often difficult to achieve because of three major issues, outlined below (Quarantelli, 1988, pp. 382-383):
23
1) Lack of Organisational Consensus: Although, with or without a prior agreement, an under-standing of what coordination means is needed to act efficiently, coordination is neither self-explanatory, nor is it consensually perceived or valued. Although some organisations might see coordination as, at most, informing other groups of what they are doing, others may see co-ordination as the centralisation of decision-making to a particular agency or officer. This can lead to accusations from either side that one or more parties failed to live up to the stand-ard. As such, Quarantelli suggests that “organisational officials should be asking more than telling, requesting more than ordering, delegating and decentralising rather than narrowing and centralising at the height of the emergency” (Quarantelli, 1988, pp. 384).
2) Strained Organisational Relationships caused by New Disaster Tasks: When entities do not have practice in coordinating with one another, the integration of new tasks can cause coordi-nation problems.
3) Impact of Disaster Magnitude.: The larger the scope of the disaster, the more likely it is that this will inhibit the success of overall organisational co-ordination. This is because the number of entities involved, that have often never interacted with one another, different lev-els of structure in each, the mix of public and private organisations, can make it almost im-possible to reach total coordination during the emergency period.
Quarantelli notes that there are different ways to judge the consequences of not achieving to-tal organisational co-ordination, and that there may be a trade-off between efficiency and ef-fectiveness. While on the one hand, if efficiency is rated highly, lack of coordination may be seen as a serious problem. If effectiveness is seen as more important, then less coordination can be tolerated. Thus, to him, coordination is not an absolute good, as there can be relatively effective responses in disasters without there being a high degree of coordination.
However, although it is good to be aware that these limitations exhist, pre-planning should strive to allow entities to have maximum organisational coordination, while at the same time taking into account that most organisational problems will need to be approached as tactical issues during a situation filled with unknowns, rather than as something that can be dealt with solely pre-emptively.
24
2.5 EFFICIENCY AND EFFECTIVENESS IN AGILE CRISIS RESPONSE ORGANIZATIONS: VAN VEELEN ET AL. 2006.
Van Veelen et al. (2013) focus on the process of coordinating resources in a networked or-ganisation. The authors describe coordination as “the act of working together harmoniously”, implying that the agents within a multi-agent community (AAC), perform interdependent ac-tivities, which are a part of a larger plan (Van Veelen et al., 2013, pp. 203).
Similarly to Quarantelli, they note that agility can mean less coordination, and therefore im-pact the ability for the coalition to remain effective (Van Veelen et al., 2013, pp. 202). They distinguish effectiveness from efficiency, noting that while the former is the ability to com-plete as many jobs between the actors as possible, the latter regards the amount of jobs that are performed successfully. They therefore suggest that the solution to the issue of loss of ef-ficiency due to increased agility is to use automated approaches to coordination (Van Veelen
et al., 2006, pp. 207). The authors exemplify three different kinds of artificial coordination:
knowledge-based coordination; the negotiation-based system; and greedy coordination (Van Veelen et al., 2006, pp. 204-207).
In order to measure performance equations were developed to measure performance of the concepts of effectiveness and efficiency. Efficiency requires measuring the volume of the co-ordination effort of the actor. Thus, respectively, the equations are represented as:
FIGURE 4 EFFECTIVENESS (VAN VEELEN ET AL., 2006, PP. 206)
Where Ci(t) = 1 when casualty I is healed at time t and otherwise 0. Nc(t) is the total amount of casualties at time t. Thus, effectiveness of an allocation is the amount of jobs completed in ration to the total number of jobs.
25
FIGURE 5 COORDINATION EFFORT (VAN VEELEN ET AL., 2006, PP. 206)
Where x is the number of resources, y is the number of casualties, T is the total number of threads, Ci(x, y) is the amount of time thread I spent in coordination state, and S (x, y) is the amount of time thread, I , spent in the stable state.
Furthermore, Van Veelen et al., (2013) note that there are four major dimensions into which real life coordination strategies can be classified. These are:
1) Implicit versus explicit: While in implicit coordination strategies there is little to no ex-plicit inter-agent communication, generally meaning that agreement on coordination is shared by all participants, in an explicit communication strategy participants communicate with one another.
2) Dynamic versus static: Dynamic coordination allows alteration during the run time, which can either be done by fine-tuning a strategy, or by replacing one strategy with another. In static coordination, the configuration is the coordination is determined a priori.
3) Cooperative versus competitive: In a cooperative multi-agent community (ACC) the par-ticipants pursue an activity of mutual benefit, with individual preferences being a secondary priority; on the other hand, in a competitive AAC the participants are self-interested and pur-sue their individual goals, with the participation of the group being a secondary interest. 4) Centralized versus decentralised: In a centralised coordination strategy a separate group of participants handles coordination, while all the remaining participants obey their instructions, while in a decentralised coordination strategies, each agent has the capacity to coordinate and has functional problem solving abilities (Van Veelen et al., 2006, pp. 203, 204).
26 2.5. A REVIEW OF THE LITERATURE
Agility is still a very under-studied subject and much of the literature that can be found using research browsers, focuses either on its desirability, or connects it to other related but differ-ent, terminologies, without putting much emphasises on what the concept actually means. Holsapple & Li propose a unified work-design theory that they believe should be applicable to all organizations, since all organizations are affected by work-design. However, Holsapple & Li’s approach is geared towards a commercial setting, rather than a community crisis, re-quiring some adaption to fit the latter. This is because in a commercial context, decisions and agility can be implemented with less urgency than during a crisis, allowing companies to ade-quately consider cost-benefit situations. By comparison, crisis situations are fast paced, where urgency allows for very little time to study the available options, and where operational re-quirements can change at a moment’s notice.
In contrast to Holsapple & Li, C2 Agility Theory and NATO SAS-085 offer a very complete bibliography on the subject of agility, more specifically applied tin the military context and, more recently, to complex endeavours (which can include firefighting). However the main criticism of C2 Agility theory is that it is, to date, too much of a “fuzzy concept” desirable but with insufficient measurability or praxis. This is acknowledged even by the authors of NATO SAS-085, who see the C2 Agility Case Study Template and the C2 Agility Generic Evidence Table as most adequate for use in theoretical research of the concept of C2 Agility itself, ra-ther than to learn about the events to which C2 Agility is applied.
Finally, although Quarantelli (1988) described many of the core ideas for agility and coordi-nation during a crisis, the concept of agility was not used directly, but rather implied. During the emergency stage of a disaster, both overly specific and overly generic regulations can hamper implementation of plans during a disaster, especially when considering that its partic-ularities will be unknowns until it happens. As such, being prepared should not mean the same as being rigid during operations, with the author instead recommending that the com-mand chain be loosened, if appropriate, during the operation (Quarantelli, 1988, 374-375). The inclusion of this author’s work was nonetheless addressed and defended in its respective sub-section.
27
Van Veelen et al (2006) was addressed to discuss the dichotomy between efficient and effec-tive coordinating strategies and how these affect agility.
Almost all of the authors agree that agility is not a good onto itself but rather, something that has to be appropriate considering the situation itself (NATO SAS-085, 2014, pp. 34; Van Veelen et al., 2006, pp. 202; Quarantelli, 1988, pp. 384). The exception is Holdapple & Lim who do not mention this aspect at all. In the case of NATO SAS-085, the authors
acknowledge that the project began with the objective of finding a “one-size fits all” edge/ag-ile Control and Command Agility model. However, in time, it became clear that this was not always the best model (NATO SAS-085, 2014, pp 46). Furthermore, added agility can be costly, because while having more options means that is more likely that the appropriate one is available, if too many options are considered, then entities become indecisive, slow to re-spond, or prone to error (NATO SAS-085, 2014, pp. 53-56). This is seen for Quarantelli and Van Veelen as being a dichotomy between efficiency and effectiveness; when one is in-creased, the other tends to decrease. While Holsapple & Li do not mention this dichotomy, they describe what they call the concept of agility effectiveness, which, for them, is seen in terms of four measures, timeliness, flexibility, relevance and affordability, which need to be present both in the ability to be responsive and in alertness (Holsapple &Li, 2008, pp. 8-10). While in principle, a further search could be carried out, it is clear that much of the existing literature on agility either relates to task allocation via automatic management systems (van Veelen et al., 2008). These are neither a method for analysis, nor can they be applied a
poste-riori. The rest of the literature measures agility using theoretical models for research purposes
in simulations where data is very strictly measured (Arminoff & Johanson, 2007; Devine et al 2015; Jobidon et al., 2013), and which therefore would be very difficult if not impossible to measure in real-life situations, as these would not provide sufficient information.
28 CHAPTER 3 – METHODOLOGY
1. RESEARCH DESIGN
The previous section of this thesis established that there was a lack of literature regarding agility, what it means to be agile, and how to identify agility. As such, I could find no single method, theory, or framework that seemed to fit the needs of this thesis and, thus, could an-swer my research question: “What are the factors that interfere with agility in emergency re-sponse to wildfire”. Consequently, the approach I decided on involves a two step-process: first, agility will be assessed in the case of Pedrógão Grande by based on the work by Holsap-ple & Li, with some input from NATO SAS-085; after this, I will try to identify the factors that inhibited agility, using the framework presented by Quarantelli.
While the second step contributes largely to answering my research question, the first aims to strengthen my research by exploring the concept of agility in the context of the Pedrógão Grande fires. The components that comprise agility will be studied so as to determine which ones were achieved or not. This is done because, without such a procedure, the analysis of which factors compromise agility in fighting forest fires would necessarily be vague without discussing what it means to be agile in this particular case study. Thus, this two-step pro-cesses allow conclusions to be drawn linking the factors that inhibit agility with the compo-nents of agility per se.
ASSESSING AGILITY
As aforementioned, for the first step, aspects were taken both from NATO SAS-085 and from Holsapple & Li, although more weight was placed on the latter.
29
• The definition of agility adopted throughout this thesis is based on that of Holsapple &Li1. Thus, agility will be defined as “the ability to converge alertness to change (recognis-ing opportunities/challenge) – both internal and external - with the capability to use resources in a responding (proactive/reactive) manner”. This definition was chosen as it seemed the most complete, recognising that agility requires both being able to notice a change in circum-stances that require a change in response, as well as the capacity to enact such a response. In contrast, most other conceptualisations such as those in Lai2 (2018) and NATO SAS-0853 do not emphasize awareness in their definitions of agility, instead highlighting the ability to be responsive. However, these authors recognize the importance of such a connection in their work, even though alertness or the idea of alertness is not outright included in the definition itself. In this thesis, I support that alertness, and its associated cognition mechanisms, are a key part of agility.
• In keeping with this decision, and for the sake of consistency, the components of agil-ity evaluated will be as defined by Holsapple & Li, although adapted to fit crisis manage-ment. The literature search I carried out showed that there was no theory that adequately ex-plained the components of agility in a crisis management context. Holsapple & Li’s work-de-sign provides a perspective of agility from an organizational standpoint, which is suitable in the case of organization of emergency response, so long as there is adaptation to the particu-larities of crisis management. Furthermore it describes the components of agility in great de-tail, in a way that lends itself well to my case study. Ideally, all of the aspects described by Holsapple & Li would have been achieved. As mentioned earlier, the authors’ description of
1“the result of integrating alertness to changes (recognising opportunities/challenges) – both internal and
envi-ronmental – with a capability to use resources in responding (proactive/reactive) to such changes, all in a timely, flexible, affordable, relevant manner”
2“Agility amid uncertainties at the organization level in this paper is defined as “to cope or ingeniously adapt to
a set of circumstances” (Preston, 1001, p. 88) or as “devising resourceful solutions to intractable problems” (Meyer, 1998, p. 572)
3“Agility is the capability to successfully effect, cope with and/or exploit changes in circumstances” (NATO
30
agility resulting from both alertness and response-ability is in keeping with views expressed by several other authors, although not necessarily encompassed in their definitions of agility. While the enablers of agility of NATO SAS-085 could have been chosen as metrics, I opted against this as I believe that these are less tangible concepts than those given by Holsapple & Li, and require a great deal of extrapolation to be measured. For example, “flexibility” to some extent will necessarily be present in emergency response, but it would be problematic to measure such an intangible concept without resorting to guesswork as it implies estimating all possible measures that a team could have taken, and why others could not have been adopted. Adding to these issues, NATO SAS-085 considers that the enablers of agility cannot be considered components of agility, and acknowledges that these concepts have a large amount of over-lap and are highly interdependent (NATO SAS-085, 2014, pp. 55), thus mak-ing them nearly impossible to isolate and measure. For these reasons, I support that NATO SAS-085’s enablers would have been unsuitable to measure agility in this case-study. • Nonetheless, based on the NATO SAS-085 generic evidence table, the analysis will be done by phases. This allows for a sense of urgency that otherwise would have been lack-ing, as, unlike the work-design perspective of Holsapple & Li, which sees agility as the func-tion of strategic, operafunc-tional and episodic agility, a crisis situafunc-tion is necessarily heavily im-pacted by its episodic nature. Furthermore, as a crisis situation is necessarily rapid and com-plex, and what is a suitable action in one moment may not be so in the next, studying the event by phases allows to focus on how entities re-organize to face new requirements. • Finally, this thesis simplifies Holsapple &Li’s model by disregarding its measures of agility effectiveness (timeliness, relevance, flexibility and affordability) in relation to each of the components of response-capability and alertness. Instead, the operationalization section of the methodology will try to incorporate these ideas as much as possible, but will instead use indicators for alertness and response-capability that are more in line with the specificities of the case-study.
The table below summarises the ideas presented above regarding how agility and its compo-nents will be analysed in this thesis.
31
COMPONENT Phase 1 Phase X...
Alertness
Strategic Foresight
Systemic Insight
Response capabilities
Value evaluation
Coordination & Integra-tion
Organisational Learning
Reconfiguration
Table 1: Template table with the components of Agility
In the case of Pedrógão Grande, four distinct time phases of firefighting where identified. This is not a mathematical division where each of the temporal parts is equal, but rather within each phase the actors followed more or less a given approach to emergency response. The first phase can be summarized as “initial response”, an early phase where the fire could have been controlled, had it been tackled promptly and with sufficient means. The second phase, starting at 19h00 on the 17th June marks approximately the time where the climatic conditions became exceptionally harsh due to the interaction between the lighting and the fires. According to those on field, at this point it became almost impossible to directly try and mitigate the fire. The overall strategy was subsequently changed in order to prioritize protect-ing civilians and their property (ANPC 2, 2017, pp. 61, 62). The third phase was a stabiliza-tion period from the 18th June when organizations began to finally regain control of the
situa-32
tion, and finished on the 22nd.June. On this day, by 07h41 in the morning, the fire was consid-ered in conclusion, meaning that its main combustion points where extinct, though small fo-cus points were still being eliminated or kept under vigilance to make sure they would not re-active (CEIF, 2017, pp. 85; DECIF, 2016, pp 29) . The last phase lasting from the 22nd to the 25th June will not be analysed as it for the most part comprised of vigilance and extinguish-ing small and occasional re-ignitions.
In short, agility will be defined by alertness and response-capability. Alertness will be consid-ered reached if at least 50% of its components are accomplished; Response-capability will be considered reached if at least 50% of its components are accomplished. Agility will only have been accomplished if there is both alertness and response-ability; if there is just one, the re-sponse will not be considered agile.
IDENTIFYING THE FACTORS THAT CONTRIBUTED TO A LACK OF AGILITY
The second aspect, which aims to answer the research question, will identify the factors that contributed to a lack of agility. This analysis will be done based on Quarantelli (1988) and made to fit the specificities of emergency response for fire fighters.
As stated in the theoretical framework, Quarantelli (1988) does not provide a theory on agil-ity; rather, the author documents organizational behaviour at the emergency stage and the problems that occur during community disasters. That being said, and as defended in that sec-tion, I found this paper to be very compatible with the concept of agility, which can be sum-marized in the following sentence found in the text: “Good crisis management, to a consider-able extent, is the application of tactics which are specifically relevant to the situational con-tingencies of a given community disaster” (Quarantelli, 1988, pp. 46). Thus, I find this sen-tence, which sensitizes many of the general ideas of the article, is highly compatible with the conceptualization of agility as promoted in this thesis, and defend that Quarantelli’s idea of “good crisis management” can be equated with mine of “agile” crisis management.
Adding to this, the scope of Quarantelli’s framework is perfect for the analysis of organiza-tional problems of organizations responding to wildfires, as it focuses on “consensus type community crisis generated by natural or technological events”, excluding conflict situations
33
and non-community disasters (Quarantelli, 1988, pp. 373). This makes it suitable for the re-search conducted in this thesis.
As such, the following table is produced which states the problems that can interfere with agility in crisis management:
Problems encountered...
Communication Process A) Intra-organizational information flow
B) Information flow between organisations
C) Information flow from organizations to the general public
D) Information flow from the public to the different organisa-tions
E) Information Flow within different systems of information
Authority and Decision Making
F) Personnel Burnout
G) Organisational Authority Conflicts
H) Clashes over organisational domains
I) Unresolved Jurisdiction Issues
Coordination structure J) Lack of Organisational Consensus
K) Strained Organisational Relationships Caused by New Dis-aster Tasks
L) Impact of Disaster Magnitude
34
The Pedrógão Grande fires were a crisis situation and, an atypical situation, therefore jeop-ardizing the ability to use pre-conceived ideas on how to respond. It would seem unfair to state that a factor was inhibiting agility if it occurred but once, as it would be almost impossi-ble to avoid committing a single error during a crisis. As such, in order to allow some leeway, given the extremeness of the circumstances I will attempt to measure the strength of how each factor interfered with agility. Accordingly I have developed the following classification: five to nine accounts of one type of factor will indicate that the factor was problematic; ten to fourteen will indicate that it was very problematic; over fifteen that it was chronic.
2. CASE STUDY
SINGLE CASE STUDY
In order to answer the research question, I will use explanatory research. Two theories from the literature review will be used together, Holsapple & Li’s work-design theory as well as Quarantelli’s theory on organizational behaviour at the emergency stage of community disas-ters. While the objective of this thesis is to understand which factors interfered with agility in the emergency response to a wildfire, which will be answered by using Quarantelli’s frame-work. Holsapple &Li’s will also be used to give more strength to the analysis and to under-stand which aspects of agility were in fact not reached. In short, the two theories complement each other in regard to the objectives of this thesis.
Agility is identified as being the dependent variable, as it is the variable that is being studied
and its performance is contingent on the independent variables, while the independent varia-bles can be summarized as the factors that constrain agility. The objective of this research is to identify, therefore, the independent variables, which will be accomplished by resorting to Quarantelli.
The research will be a qualitative single case study. A case study was chosen as it seemed the best way to answer the research question, as it allows me to “apply an intensive approach on a single phenomenon, studied in its own context to achieve greater detail” (Swanborn, 2010, pp. 3). This choice is therefore justified as an in-depth analysis of a particular case will allow to discern which factors interfered with agility, based on empirical evidence. While, ideally,
35
multiple case studies would be performed in order to make the research more replicable, this would not be feasible given the time and space constraints of this thesis.
CHOICE OF CASE SELECTION
These Pedrógão Grande fires were selected for three reasons. First, the event warrants a deeper investigation because of the sheer magnitude and consequences, amounting to 67 deaths and an estimate half a billion Euros in societal costs (Diário de Notícias, 2017). It is therefore important to do a deeper investigation in to what happened and whether the re-sponse could have been better, in order to try and prevent in the future any similar situation from occurring. Furthermore, this may also provide some closure for those who were af-fected by the situation. Secondly, these fires were greatly impacted by the abnormal weather conditions, the horizontal convective currents that occurred due to the interaction between the fires and the thunderstorm. As this conjugation of events was completely unexpected both for decision makers and those on field, there was no preparation for such a scenario or any con-tingency plan to fall back on. This means that the Pedrógão Grande fire lends itself well for the study of agility. Finally, these are the first fires in Portuguese history to have minute to minute descriptions of the operational response. Addressing the research question will fill an important knowledge gap regarding crisis management in Portuguese wildfires (Fernandes et
al., 2015, pp. 8; Beighley & Hyde, 2018, pp. 14).
3. DATA COLLECTION AND ANALYSIS
ORIGIN OF DATA
As the research is based on studying agility during the Pedrógão Grande fires, and, subse-quently, analysing if and what factors compromised agility, the data comes mostly from three types of sources:
36
• Evaluation of the fires that took place between 14 and 17 June 2017 in mainland Por-tugal (CTI4 – Independent Technical Committee)
• The Fire Complex of Pedrógão Grande and neighbouring counties initiated on June 17, 2017 (CEIF5 – Center for Study on Forest Fires).
• Pedrógão Grande – The Challenge of a Catastrophe (CEIPC6 – Center of Studies on Interventions in Civil Protection)
Assessment reports produced by the entities themselves
• Report on the Fire in Pedrógão Grande (ANPC7- National Association for Civil Pro-tection)
• Provisional Report of Rural Fire - 2018 – 1st January to 1st July (ICNF8 – Institute for Conservation of Nature and Forests)
• Pedrógão Grande – Final Report (GNR9 – National Republican Guard)
Audits and investigations on the participating entities
• Answer to the question nº 4560/XIII/2-ª, 23rd of June 2017 – Forest fire occurrence in Pedrógão Grande (Cabinet to the Prime Minister)
• Final Report (ANPC )
4: Comissão Técnica Independente
5 Centro de Estudos de Incêndios Florestais
6Centro de Estudos e Intervenção em Protecção Cívil 7Autoridade Nacional de Protecção Cívil
8Instituto da Conservação da Natureza e das Florestas 9 Guarda Nacional Repúblicana
37
• Document Answering the Analysis of the Report on the Fires in Pedrógão Grande by ANPC (ANPC)
• Audit on SIRESP (IGAI10 – General Inspection on Internal Administration) • Study on the Functioning of SIRESP – Part 1 (Institute of Telecommunication11)
The reports that are produced by the entities involved in the firefighting are, for the most part, seen as reliable. This can be justified as there have been several legal processes investigating their quality that verify that all the necessary information that should have been provided by each entity was accounted for (CEIF, 2017; CTI, 2017). The single exception is the first Re-port by the ANPC, the National Authority for Civil Protection, which is the institute responsi-ble for fire response coordinating, and all firefighting activity and has been the subject of an intense investigation surrounding the destruction of important documents. These include the tactical situation chart, the strategic plan of action, and the cell information chart (Mandim, 2018).
Nonetheless, while these no longer exist, the audit on ANPC’s role in Pedrógão has been re-leased (see above) and is an extremely thorough document, featuring lengthy transcripts from the on-field operatives regarding the command and control of operations, and the issues they faced regarding logistics and communications, among others. This allowed me to analyse the data directly from a reputable source, as this audit (ANPC, 2017b; ANPC, 2017c) has been the subject of much legal scrutiny. The audit also allowed me to read interviews of over two dozen individuals who participated in the response, including high-ranking commanders which would otherwise be impossible for a student such as myself to acquire independently. The data compiled in the CTI, CEIF, and CEIPC reports are also highly reliable and was cross-linked by experts. However, the interviews conducted for their elaboration are not available to the public.
10Inspecção-geral da Administração Interna 11Instituto de Telecomunicação
38
Given the time restrictions for elaborating this Masters thesis, it is unlikely that I would have been able to gather additional information via interviews, given that the Pedrógão Fires have become a sensitive subject due to matters relating to accountability.
In addition, I have had to survey the legislation relating to civil protection in Portugal, in par-ticular that relating to the response to wildfires. For this, the main laws are the Law-Decree nº134/2006, implementing the SIOPS (Integrated System for Rescue Operations), last up-dated in 2013, Law nº27/2007, establishing the Base Laws on Civil Protection, and the first and second DONs (Nacional Operational Directives), the latter of which refers to the DECIF (Special Device for Combating Forest Fires).
DATA ANALYSIS
The research will use qualitative content analysis to analyse the data. Qualitative content analysis is the interpretation of textual data through a systemic process of coding and the identification of themes and processes, and it was chosen as it fits my data, which is text based, and my research type, which is qualitative.
The data regarding the indicators by Holsapple & Li was coded using numbers (1, 2, 3…), in relation to the different phases of the firefighting at Pedrógão Grande. When a component of agility was not relevant to the phase it was registered as “not applicable”. In the case of a component with multiple indicators it was considered affirmative only if all parts manifested, or if the data did not indicate/specify that each of them failed.
The data regarding the indicators by Quarantelli was coded as (A, B, C…). Given that crisis situations are highly time-sensitive and hundreds of decisions are often made in an hour, in order not to be overly harsh on the response I decided to only count a problem once it had oc-curred over five times during the course of the entire operation. If it ococ-curred five times, it would be considered problematic; over ten, very problematic; over fifteen, chronic.
All of the variables can be found below, as the next section is devoted to the operationaliza-tion of the key concepts of the research.
