1
: How did the policy narratives which were framed by traditional media influence the policy strategies of politicians during, and in the aftermath of Greek bush fires?
: Alkis Makrygiannis : s2161559 : alkhs.mak@gmail.com : 15.04.2020 : Dr. W. Jong :. Dr. J. Matthys
3
1. Introduction ... 4
1.1 Relevance of the Research ... 6
2. Theoretical Framework ... 6
2.1 Crisis Communication... 7
2.2 Media ... 7
2.3 Main Models of the Study ... 8
2.3.1 Media Coverage Crisis Exploitation Characteristics by Olsson et al. (2015) ... 9
2.3.2. Theoretical construct of the Narrative Policy Framework (NPF) ... 10
2.3.3 Crisis Exploitation by Boin et al. (2009)... 12
2.4 Conceptualization & Expectations: An Extended Media Crisis Exploitation Model ... 14
3. Methodology ... 20
3.1 Overall Strategy ... 20
3.2 Case Study Design ... 20
3.3 Operationalization ... 21
3.4 Data Collection and Analysis ... 25
3.5 Limitations of the Research... 27
4. Analysis ... 29
4.1 Case Study: The 2018 Athens Wildfire ... 29
4.2 Presence of Policy Narratives ... 30
4.3 Media Coverage Characteristics in Newspapers ... 32
4.3.1 Style of Reporting ... 32
4.3.2 Objectivity of the News ... 33
4.3.3 Tone of Reporting in Newspapers... 35
4.3.4 Framing of Newspapers ... 36
4.4 Policy Outcomes ... 39
4.4.1 The Wildfire as a framing contest ... 39
4.4.2 Policy Effects... 41 5. Conclusion ... 49 5.1 Summary ... 49 5.2 Findings ... 50 5.3 Limitations ... 53 5.4 Future Recommendations ... 53 Reference List ... 54
4
1. Introduction
Unpredictable, catastrophic and extremely stressful. These are just a few everyday words that come into people’s minds when asked to label a crisis, a disaster, sometimes a “fiasco” or a “scandal”. These types of events tend to trigger social unrest in the public domain, which in turn leads to extensive political ramifications (Olsson et al., 2015). Big crises usually come with the public demand for bigger answers, that are expected to shed light on every and each detail of these major tumults. In search of such answers, the crisis becomes a “hot potato” among the actors involved, in some cases by producing losers rather than winners (Boin et al., 2015). Large-scale crises in particular, have the power to either bolster or hinder careers of officeholders and public leaders by questioning political responsibility and leadership performance (Pile and Prins, 2018, Olsson et al., 2015). According to Boin et al. (2009: 82), these major disruptions create breeding ground for “actors inside and outside government to redefine issues, propose policy innovations and organizational reforms, gain popularity and strike at opponents”. Officeholders exploit the dynamics of crises, by providing their own dominant picture of the situation, and divert public attention in an effort to shape certain favorable policy and/or political outcomes (Boin et al., 2009). In this context, at one end of the spectrum we have the crisis per se, and at the other end the changes in political / policy processes. What lies at the middle however, as an mediating factor, is the power of mass media to produce certain narratives, in the context of forming public opinion or becoming actively involved in the policy process (Arnold, 1990; Crow and Lawlor, 2016). Framing is better understood as “a cognitive concept that refers to the influence of transmitted messages on individual knowledge and beliefs” (Crow and Lawlor, 2016: 476). Simply put, frames shape the way that some events are being reported, having the potential to put the spotlight on certain aspects of the focusing event, manipulating the consensus gentium and thus, to likely shape a policymaker’s agenda. Media set the stage for these “frame contests” between actors to manifest, but sometimes media themselves can influence the policy process (Crow and Lawlor, 2016).
Based on the above, the overall goal of this research is to shed academic light on the interaction between policy narratives operationalized by mass media and policy outcomes of large-scale crises, such as disasters. In order to achieve this, three main models will be put forward: The first one, the Narrative Policy Framework (NPF), focusing on the role of media and policy narratives in the public policy process (Shanahan et al., 2016; Crow and Lawlor, 2016). According to the NPF, narratives are of utmost
5
significance for highlighting dynamics, tactics, actor typology and behavior in the realm of public policy decision making (Crow et al., 2016). A Narrative can be defined as “a story with a temporal sequence of events unfolding in a plot that is populated by dramatic moments, symbols, and archetypal characters that culminates in a moral to the story” (Jones and McBeth, 2010: 329). Such narratives can drive public opinion and foreshadow governmental policy-making (Jones and McBeth, 2010). Crow and Lawlor (2016), appended the discussion of mass media influence on public policy to the NPF framework, by providing theoretical tools that broaden the NPF to include the role of media framing and narratives in policymaking. In addition to the NPF, Olsson et al. (2015) introduced a four-dimensional model characterizing crisis exploitation coverage by mass media. Their framework assesses journalistic routines of reporting and contains the following aspects: “interpretive or descriptive journalism, strategic game or issue framing, and balanced reporting and biases in actor presentations” (Olsson et al., 2015: 159). This, in conjunction with the work of Crow and Lawlor (2016), will set the basis for a tailor-made classification framework of media frames and narratives, and their impact on public policies during disasters.
The third framework that this study will explore is the crisis framing and exploitation model by Boin et al. (2009). In their article “ Crisis exploitation: political and policy impacts of framing contests” (2009), the authors identify a series of ‘framing contests’, shaping the association between the so-called “crisis exploitation” game, played by political leaders in an attempt to impose their own narrative, as well as the political and policy outcomes of crisis events (Boin et al., 2009). The authors also recognized the significant role of mass media as an actor with the potential to influence the policy-making process, and exploit certain crisis narratives (Boin et al.,2009). As per the fact that this research is dedicated to examining policy narratives in the context of disasters, and through the lens of the above-mentioned NPF framework, the framing contest model of Boin et al. (2009) will be operationalized in a complementary manner in order to track and analyze crisis policy impacts, rather than political outcomes.
This explorative paper will focus on how crisis coverage by Greek traditional media contributed to the employment of different policy narratives during the 2018 Attica Wildfires. The combined theories will be applied to a single-case study design of the aforementioned recent wildfire disaster that took place on July 2018 in the suburbs of Athens, and resulting in the death of 102 people (BBC, 2019). The goal is to analyze policy processes and actor behavior in the context of a wildfire crisis. This leads us to the following central research question:
How did the policy narratives which were framed by traditional media influence the policy strategies of politicians during, and in the aftermath of Greek bush fires?
6
1.1 Relevance of the Research
To time, the academic literature focusing on the integration of the NPF for interpreting media exploitation and policy outcomes during crises is unprecedented. The interdisciplinarity of the NPF model allowed a gradual amount of researchers from various fields to address the framework’s main research objective: the role of narratives in the policy process (Shanahan et al., 2016). Pierce, Smith-Walter, and Peterson (2014 as cited in Crow et al., 2017) examined 19 peer-reviewed articles, that used the NPF but focused on environment- and energy-related topics. Crow et al., (2017), were the first to apply the NPF in disaster policy contexts by examining policy problems used by various actors during two wildfires in Colorado, USA. None of the existing NPF literature however, distinguishes between media actors and other policy actors. By answering the aforementioned research question, this study contributes to the fields of crisis communication, crisis management, political communication, and public policy in the following ways: (i) by introducing the utilization of NPF models into Crisis Management and Crisis Communication; (ii) by creating a comprehensive framework for analyzing media framing, narrative strategies and policy outcomes; (iii) it contributes to empirical literature by conducting an observational research focusing on a recent large-scale disaster; (iv) and by adding to the studies of Boin et al., (2009), and Olsson et al.,(2015) with the integration of media actors and policy strategies.
From a societal perspective, this research project provides valuable insight for crisis managers, public policy practitioners, as well as journalists. As it has been briefly discussed above, frames can shape policy and political outcomes during crises. Understanding the practical significance of crisis exploitation by mass media and its impact on policy making can be beneficial for all actors involved in the media narrative-policymaking nexus. Regarding the Attica Wildfires per se, the findings of this analysis could add to the knowledge of Greek journalists, crisis managers, and officeholders when faced with suchlike crisis events.
2. Theoretical Framework
This section is dedicated to presenting all relevant theories to our main research objective. In this part, the research will dive into unravelling the concept of NPF, and the significance of identifying dominant policy narratives in our data sample, as well as the complimentary frameworks from Olsson et al. (2015)
7
on media framing, and Boin et al. (2009) on policy impacts in the aftermath of crises. Above all else, some key definitions on crisis communication, media, and crisis exploitation by media actors will be provided:
2.1 Crisis Communication
Research on the field of Crisis Communication has traditionally focused on the private sector, with numerous corporate case studies which covered issues related to image restoration, responsibility attribution, and trust (Olsson, 2014; Jong, 2020). From a corporate point of view, crisis communication evaluates the use of crisis response strategies and communication tactics needed to preserve the organization’s reputation after the occurrence of a crisis (Barton 1993; Benoit 1995; DeVries and Fizpatrick 2006; Hearit 1996, 2001 as cited in Kim, 2019). These notions have been further elaborated and integrated into the Situational Crisis Communication Theory (SCCT), which asserts that
“corporations should strategically respond to a crisis situation based on an understanding of how the public attributes responsibility for the crisis” (Kim, 2019: 354). Coombs (2007), further added to the model three prevailing clusters that shape crisis responsibility (victim, accidental, and intentional). Each crisis cluster can help the crisis manager assess the level of crisis responsibility and thus, estimate the organization’s reputational score (Coombs, 2007).
Moving from the private to the public domain, the academic literature mainly focuses on public leadership tasks and performance rather than organizational reputation. Existing theoretical studies provided valuable insight, by introducing various models for measuring leadership performance and the impact of framing contests on crisis events (Boin et al. Hart, 2005; Boin et al., 2013; Boin et al., 2009). Boin et al. (2005), constructed a theoretical framework with five public leadership tasks for crises: sense making, decision making, meaning making, terminating, and learning. Drawing from this model, Jong et al. (2016), conducted a systematic review of 34 empirical articles, evaluating leadership tasks and behavior effectiveness of mayors and governors during, and in the aftermath of crises.
8
The role of media in the context of crisis communication is of paramount importance. To start with, media can be defined as “ a means of mass communication- the platform or technology through which messages are disseminated” (Crow and Lawlor, 2016: 474).A distinction based on the platform through which messages are being transmitted can be made; direct communication mainly refers to user-generated information channels, and the so-called web 2.0 applications including search engines, social networks, audiovisual content sharing et cetera (Huang et al., 2010). Traditional media coverage however, remains the main source of information dissemination, and should not be overshadowed by the rapid spread of new media ( Crow and Lawlor, 2016). For the purposes of this study, traditional press (online and printed) will constitute the unit of analysis. Ultimately, media can be perceived as the ‘device’ of news spread, but also as a group including many different actors such as journalists; editors and reporters. An additional definition provided by Crow and Lawlor (2016: 474), designates media as “ a nested system of interacting and dependent organizations”.
Overall, the relationship between media, media actors and the public policy process has drawn academic attention in recent studies (Crow and Lawlor, 2016). Scholars highlight the significance of storytelling in analyzing policy/ political outcomes, as valuable tools for researching policy agendas, asserting that media can play the role of actors in the policy process (Shanahan et al., 2016; Crow and Lawlor, 2016). Olsson et al. (2015), made a rough distinction between inside-the-media and outside-the-media news events. The latter refers to events framed as crisis-related from the actors reported on in the news; inside-the-media on the other hand, involves frames and narratives perceived as crises from journalists and news organizations themselves (Olsson et al., 2015; Frandsen & Johansen, 2010; Frandsen & Johansen, 2016). Yet, it remains academically unclear whether those actors deploy suchlike coalition frames intentionally, or whether they influence frames by co-incidence “simply acting as conduits of information from coalitions” (Crow and Lawlor, 2016: 474).
2.3 Main Models of the Study
This study seeks to add to previous research on public policy outcomes and crisis framing by traditional media by creating an extended NPF model based upon two existing frameworks; framing contests and
9
political/ policy outcomes by Boin et al. (2009), and the four-dimensional media crisis exploitation model by Olsson et al. (2015). Even though Boin et al. (2009) did not approach crisis exploitation from the perspective of media acting as frame-producers, both of these models acknowledge the fact that traditional media are not always just observers who simply set the “arena” for frames to be
communicated: According to Boin et al. (2009: 96), “ the rival interpretation is that the media pursue their own agenda in crisis reporting, and that the crisis communication performance by any of the actors matters less than the degree to which the color of the frame they put forward fits with the pre-existing biases of the main media outlets”. Olsson et al., (2015) further added that certain types of journalistic norms (i.e. game framing, and interpretive journalism) can strengthen/weaken the political actor’s ability to impose their dominant frame and thus, weaken their support.
2.3.1 Media Coverage Crisis Exploitation Characteristics by Olsson et al. (2015)
Boin et al. (2009) successfully linked the framing contest to certain policy outcomes, without however investigating the extent to which media can generate or affect framing contests, that lead to certain policy outcomes. This research is based upon the assumption that these outcomes can be influenced by media framing. In a very similar vein, Olsson et al. (2015) introduced four analytical dimensions of news coverage that necessitate the following characteristics: interpretive or descriptive journalism, strategic game or issue framing, and balanced reporting and biases in actor presentations.
• Framing : Frames are crucial for communicating to the public an issue at stake, and can potentially affect the audience’s perception (Olsson et al., 2015). Being one of the most
commonly applied frames, the strategic game frame is inextricably linked to the so-called media logic, “characterized by features such as dramatization conflict, and personalization of the news” (p.162). This frame puts the crisis in a normative context, and seeks answers such as why did the crisis happen or what is the prior crisis history of the officeholder. According to the authors this framing type can generate distrust among political actors. Issue framing on the other hand, solely focuses on news covering the crisis per se, or at what has been said or done by officeholders regarding the crisis event.
• News Objectivity: Here, Olsson et al. (2015) distinguish between balanced and unbalanced reporting. Balanced reporting is about the amount of media coverage that a political actor receives, in comparison to his political counterpart. Unbalanced reporting occurs when media show preference to government actors or vice versa.
10
• Style of Reporting: The third dimension relates to interpretive and descriptive style of reporting. Simple observation and factual reporting of the news (what, where, when, and who) adheres to descriptive style of reporting. “In shifting focus to why, the journalist becomes an analyst, rather than an observer of political events” (Olsson et al, 2015). In the latter case one should expect crisis reporting to follow an interpretive style.
• Tone of reporting: Finally, this dimension involves positive or negative tone of reporting. Negativity can be further distinguished into frame-based, and actor-related, dependent upon the journalistic focal point. Frame-based negative tone refers to the instance when the focusing event as a whole is being reported negatively; usually this occurs when journalists cover highly politicized events such as scandals, fiascos or conflicts. Actor-related occurs when journalists choose to report certain actors in a negative tone. The later occurs when journalists “portray themselves as independent actors in their own right and, by doing so, fulfill the idea of the fourth estate” (Olsson et al., 2015: 163)
2.3.2. Theoretical construct of the Narrative Policy Framework (NPF)
This research is intended to analyze dominant media policy narratives, and policy outcomes during, and in the aftermath of the 2018 Attica bushfires. So far, we have covered both the policy outcomes of crisis exploitation based on the model of Boin et al. (2009) as well as the four dimensions of news coverage by Olsson et al. (2015). At this point, the conceptualization of the NPF during crises is important for two prevailing reasons: (a) for identifying dominant wildfire-related policy narratives on newspapers, and (b) for coding these narratives by turning traditional media into characters (actors) in the policy outcome process. First however, some key elements of the NPF will be presented.
The initial development of the NPF could be attributed to the lack of empirical studies in the field of policy studies. According to Crow at al. (2017: 629), “narratives are important, but seldom empirically tested in policy scholarship . . . they hold potential for illuminating dynamics, beliefs, and actor behavior within the policy process”. The central idea of narratives is about influencing processes, such as
policymaking (Crow et al., 2017). The NPF includes three levels of analysis. The micro level, corresponds to the individual-level and thus, analyses how narratives influence individuals, and vice versa. The meso level NPF, which is also set as the level of analysis of the current study, involves groups of actors
constructing and communicating through different narratives in order to shape certain policy outcomes (Shanahan et al., 2016; Crow et al., 2016). At last, the macro level of analysis focuses upon a larger scale
11
and “centers on policy narratives that permeate institutions, society, and cultural norms” (Shanahan et al., 2016: 334)
There are several elements that may be included in a policy narrative. These are: (a) the setting, which corresponds to the place or situation where the policy event manifested (i.e. Attica, Greece), (b) the plot or policy problem, (c) the policy solution, and (d) the characters (usually categorized as heroes ,victims and villains) (Crow et al., 2017). Characters can have the form of individuals, agencies, various groups, or even non-human entities (Shanahan et al., 2018). Policy narratives may be ‘imperfectly constructed’, meaning that there is no requirement for all of these elements to be included in the framework (Crow et al., 2017). According to Shanahan et al. (2018: 336), “a policy narrative must have at least one character and a reference to the policy in question”.
12
2.3.3 Crisis Exploitation by Boin et al. (2009)
Boin et al.’s (2009) framework on crisis exploitation through framing, attempts to connect the dots between communication frames that might be put forward by leaders during crises, and certain political/ policy outcomes. According to the authors, crisis exploitation is the “purposeful utilization of crisis-type rhetoric to significantly alter levels of political support for public office-holders and public policies” (Boin et al., 2009:83).
Following the crisis, political leaders will attempt to impose one out of three framing contests, each of them having specific political and policy stance characteristics. Boin et al., (2009) conceptualized this “crisis communication checklist” by introducing the following model:
Figure 1 (Source: Boin et al., 2009:84)
• No Crisis: Figure 1 shows three different postures. The first type, denial, is usually being deployed by actors during unpreventable disasters, such as “Acts of God”. The imposed frames are being operationalized as if there is no crisis. There is no blame attribution, and policy stance remains as is.
• Crisis as Threat: The second frame acknowledges the existence of a crisis event, that can prove harmful both for the society as well as for the status quo players. The attribution of blame is evident, and the officeholder needs to diffuse it, usually by exogenizing the crisis.
13
• Crisis as opportunity: As the phrase itself indicates, unlike the two aforementioned crises, this frame sees the crisis event as an opportunity to promote certain policy and/or political
outcomes. In other words, proponents of this type will try to politicize the crisis, and attempt to change the status quo.
Up to this point we have seen that crisis events give rise to various frames, what Boin et al. (2009) described as “conflicting interpretations reflecting the existing plurality of values and interests, as well as the ‘scripted competition’ between governments and oppositions” (p.88). Apart from the unfolding these frames, the authors demonstrate an additional model on policy outcomes of crisis exploitation:
Figure 2 (Source: Boin et al., 2009:90)
Figure 2 suggests that four types of policy outcomes may result, depending on the ‘power-balance battle’ between status quo players, and change advocates:
• Box I: Both parties adopt an aggressive stance. This can either end up in a protracted stalemate or in a fundamental policy change
• Box II: Status quo players continue to adopt a defensive stance towards policy adjustment(s). Policy change advocates demand gradual policy reform. This policy outcome can either lead to a dead end or to incremental policy adjustments
• Box III: Coalition advocates press for a paradigm policy shift, and government proceeds to substantive changes in an attempt to pacify pro-change players
14
• Box IV: Negotiated package of moderate policy reforms
2.4 Conceptualization & Expectations: An Extended Media Crisis Exploitation Model
Hitherto, this paper has demonstrated some relevant theories to our main research objective. The following extended conceptual model encompasses these frameworks, and provides all basic guidelines needed for conducting this research. Table 1 illustrates the linkage between dominant policy narratives in news outlets, and expected policy outcomes. Therefore, our general expectation is that news media coverage, will act as an intervening variable between policy narratives (independent variable), and policy outcomes (dependent variable). Particularly, this research expects that each dimension of media coverage will lead to certain policy outcomes. The Extended Media Crisis Communication Model (Table 1) concisely illustrates these expectations. Additionally, based on the empirical findings of Olsson et al. (2015), and Boin et al., (2009), the following set of supplementary expectations has been created:
• A depoliticized style of media coverage (predominance of issue frames, descriptive journalism, balanced reporting or unbalanced reports covering the status quo players) is expected to increase support for status-quo players (government), whereas ideological/ political subjectivity and biased frames (game framing, interpretative style of reporting, unbalanced reports covering the change advocates) is expected to sway public opinion against status quo players.
• Narratives dominated by factual-based, and unbiased journalism will show stability across dimensions characterized by purely descriptive coverage (issue frames, descriptive journalism), while narratives dominated by journalistic subjectivity, biased analyses, and personal
assessments will show stability across dimensions characterized by journalists’ speculations, comments and melodramatic reports (game framing, interpretative news reporting).
15
• Narratives that appear more frequently in newspapers, and address policy-related problems might foreshadow future effects on policies
16
Independent Variable: Policy Narratives
Intervening Variable: Media Coverage Dependent Variable: Policy Outcomes Narrative Policy Framework (NPF) by
Shanahan et al. (2015)
Four dimensions of Media Coverage by Olsson et al. (2015) Crisis exploitation Framework by Boin et al. (2009)
Identified Policy Narratives Allocated
dimension(s) of media coverage
Allocated type(s) of media coverage Expected Policy Outcome(s) based on government’s policy framing strategy
#1,2,3,4 Style of reporting Descriptive Resist Policy
Change: Policy stalemate; politically imposed incremental adjustment Contain Policy Change: negotiated incremental adjustment Interpretative Resist Policy
Change: Policy stalemate or politically imposed paradigm shift Contain Policy Change: Major and swift rhetorical/ symbolic change; more incremental substantive change
#1,2,3,4 Framing Issue Framing Resist Policy
17 stalemate; politically imposed incremental adjustment Contain Policy Change: negotiated incremental adjustment Game Framing Resist Policy
Change: Policy stalemate or politically imposed paradigm shift Contain Policy Change: Major and swift rhetorical/ symbolic change; more incremental substantive change #1,2,3,4 Objectivity of the news
Balanced Resist Policy
Change: Policy stalemate or politically imposed paradigm shift Contain Policy Change: Major and swift rhetorical/ symbolic change; more incremental substantive change Unbalanced Resist Policy
Change: Policy stalemate; politically imposed
18 incremental adjustment Contain Policy Change: negotiated incremental adjustment
#1,2,3,4 Tone of reporting Positive Tone Resist Policy
Change: Policy stalemate; politically imposed incremental adjustment Contain Policy Change: negotiated incremental adjustment Negative Tone Resist Policy
Change: Policy stalemate or politically imposed paradigm shift Contain Policy Change: Major and swift rhetorical/ symbolic change; more incremental substantive change
➢ Style of Reporting: This dimension refers to the amount of interpretative or descriptive journalism in the news article under examination. Here, the uncertainty and surprise
characterizing a crisis event usually comes with a lack of available information to be transmitted to the public (Olsson et al., 2015). Media have no other option rather than being descriptive and relying on governmental sources. On the contrary, higher levels of interpretative journalism
19
are expected to be traced when media become contributors to debates, rather than factual observers of the crisis event (Crow et al., 2016; Olsson et al., 2015).
➢ Framing: This dimension refers to the extent to which media employ strategic game or issue frames. The former refers to common journalistic practices of employing so-called episodic frames. This type of framing strategy is characterized by high levels of “dramatization, conflict, and personalization of the news” (Olsson et al., 2015: 162). In this scenario we expect to see substantial policy changes. Issue framing is purely descriptive, and focuses on the thematic coverage of the actual crisis events, favoring governmental players to maintain their policy status quo.
➢ Objectivity of the news: Objectivity of news reporting is closely linked to balanced coverage (Olsson et al., 2015). This dimension describes the common assumption that political actors should be given equal opportunities to media access. According to Olsson et al (2015: 162), this is often measured by “studying the amount of coverage given to different political actors”. In the case of newspapers, balanced reporting refers to the equal appearance of official sources (i.e operational crisis staff). Contrastingly, unbalanced reporting refers to the journalistic preference to feature certain political actors on a grater scale. This is expected to lead to significant alterations to policy outcomes, in favor of status quo (governmental) actors.
➢ Tone of Reporting: Simply described, this dimension refers to the journalist’s stance towards the (governmental) actor. Negative reporting is therefore expected to produce extensive policy rearrangements, and positive tone will influence policy outcomes towards the status quo policy preference.
20
3. Methodology
3.1 Overall Strategy
This research is rather explorative, and it is dedicated to analyzing the relationship between policy narratives deployed by newspaper outlets, and policy outcomes; while simultaneously applying these dimensions to a recent Greek wildfire disaster. Therefore this thesis situates itself between two main focal points: From a theoretical standpoint, it applies the concept of NPF research into the field of crisis management and crisis communication, by conceptualizing an extended NPF model that can be used for further research; from the empirical perspective it operationalizes the provided framework by applying it to the 2018 Attica Wildfires crisis. This enables Greek practitioners to utilize this study as a means of strategic-level evaluation of public policy processing with regard to traditional media influence. Additionally, this research model provides a solid empirical basis for future case studies on disaster management, with regard to the influence of policy narratives -presented through the lens of traditional news media- on policy outcomes.
3.2 Case Study Design
Following the words of Robert Yin (2009:14), a case study can be defined as “an empirical enquiry that investigates a contemporary phenomenon in depth and within its real-life context, especially when the boundaries between phenomenon and context are not clearly evident”. Accordingly, the chosen case- the 2018 Attica Wildfires- will be investigated in the context of policy outcomes attributed to crisis exploitation strategies by Greek traditional media. The main research question of this paper requires the prosecution of qualitative content analysis on Greek newspaper articles focusing on the selected case. Newspapers will constitute the unit of analysis for this research. The level of analysis required to conduct an NPF study -which is different than the unit of analysis- is set as the meso-level.
Going back to the initial definition provided by Boin et al. (2009:82) on crisis exploitation during crises, “disruptions of societal routines and expectations open up political space for actors inside and outside government to redefine issues, propose policy innovations and organizational reforms, gain popularity and strike at opponents”. The Attica Wildfires have been deemed by numerous print media channels as highly politicized, and created a lot of social and political controversy in the mass media (BBC, 2019).
21
According to a recent report by Greek prosecutors “there was an absolute lack of communication, chaos and a collapse of the system, culpability among state services that shared responsibility, criminal
mistakes and omissions during the handling of the wildfires" (BBC, 2019). What makes the Attica Wildfires particularly relevant for academic analysis, is the expectation to find a plethora of policy narratives, as well as a large abundance of newspaper sources with strong indications of leadership and media framing.
3.3 Operationalization
Table 2: Operationalization Table
Concepts Theory Definition Indicators Data
Sources
Policy Narratives Shanahan
et al. (2015) a narrative is an exercise in storytelling that presents the reader with a far more explicit set of moral evaluations and recommendations for treatment (Crow & Lawror, 2016: 481) Sentences or paragraphs qualify as narratives if they mention the Athens Wildfires and also reference it in terms of at least one of the four dimensions of media coverage. -News articles included in our article Database (N=66)
Media Coverage Olsson et
al. (2015)
Media reporting of crises can affect an actor’s crisis communication performance by influencing storytelling - -News articles containing at least one coded narrative (N=55)
22 through certain coverage characteristics (Olsson et al., 2015; Boin et al., 2009) Media Coverage Characteristic #1 Style of Reporting
Interpretive Journalism that focuses more on explaining why an event has occurred. Journalists become analysts rather than factual observers of a debate. News articles dominated by journalistic comments, prognoses, and speculations on the wildfire’s trajectory
Descriptive Journalism that
focuses on the four classical W questions (what, where, when and who) (Olsson et al., 2015:162). Simple, and factual description of the event. News articles containing information related to Wildfire situational updates, or purely informative scientific and judicial reports Media Coverage Characteristic #2
Framing Issue News articles
focusing on the actual crisis. Purely descriptive journalism (Olsson et al., 2015) News articles putting the spotlight on the actual event and its characteristics as they occurred Game In contrast to issue framing, game framing Articles depicting the Crisis
23
manifests when journalists exploit stories by subjectively covering the crisis.
management of the Wildfire as a highly politicized strategic activity Media Coverage Characteristic #3 Objectivity of the news (Balance of Sources)
Balanced Balance of sources
refers to the amount of coverage given to governmental and/or oppositional actors within an article. Equal media coverage is an indication of balanced reporting (Olsson et al., 2015). Articles covering both the government AND opposition Unbalanced Articles dominated mostly by governmental or oppositional sources Articles covering the government OR opposition Media Coverage Characteristic #4 Tone of reporting
Positive When journalists
adopt a positive stance towards events as a whole, and/or actors involved in the handling of the wildfire crisis. Positive tone can manifest through support, competence or progress (Olsson et al., 2015). Involved actors (e.g. Government, Civil Protection, Police and Fire Services) who are framed in a positive manner
24
Negative Involved actors or
events who/which are portrayed in the news in a negative manner (Olsson et al., 2015). Involved actors (e.g. Government, Civil Protection, Police and Fire Services) who are framed in a negative and condemnatory manner
Policy Outcomes Boin et al.
(2009) Media effects on “program content and delivery modes” of policies related to the wildfire crisis; the nature and degree of policy change in the wake of an emergency/ disturbance (Boin et al., 2009:83). News articles containing information about policy solutions (policy changes) with regard to the the coded policy problems (narratives) -News articles whose main subject is the Athens Wildfire of 2018. -Official statements by governmental actors
25
3.4 Data Collection and Analysis
Deductive content analysis focuses upon using prior formulated theoretical considerations -which direct and limit the scope of analysis- and connecting them to textual data (Marek et al., 2015). Deductive categorization will be therefore performed by operationalizing the Integrated Media Crisis Exploitation Model. In order to identify existing policy narratives, textual data will be collected from three different Greek newspapers. Only articles whose main subject is the Wildfire will be included in our database. Data will be collected from the 23rd of July 2018 -which was the date of the fires onset-, and up until November 2019. According to Crow et al. (2017), policy narratives require long-term observation, otherwise scholars are likely to miss the emergence of policy debates during low level seasons of media attention. Thus, the time frame will be extended beyond the actual crisis timescale. Policy outcomes will be verified by examining newspapers as well as published scientific reports and official inquiries (e.g. judiciary reports).
According to Shanahan et al., (2015: 341), “newspapers are often the best early source for policy narratives about local issues or issues with a particular geographic domain such as disasters, given that policy actors form groups much later in the policy process”. Three large daily Greek newspapers have been chosen as news agencies susceptible for analysis: Kathimerini (Greek: Η Καθημερινή), Proto Thema (Greek: Πρώτο Θέμα), and Efimerida Syntakton (Greek: Εφημερίδα των Συντακτών). In terms of daily circulation, both kathimerini and proto thema have a weekly circulation of around 60.000 copies (Eurotopics, 2019). Efimerida Syntakton has a lower circulation of approximately 7.000 weekly copies (Eurotopics, 2019). With regard to ideological diversity, newspapers can be distinguished between right-of-center (Kathimerini, Proto Thema), and left-right-of-center ( Efimerida Syntakton) (Eurotopics, 2019). Even though this paper does not seek to understand how media ideology matters to story framing during crises, our decision to analyze newspapers with different political orientations will allow us to get a broader, more accurate, and less subjective understanding of potential differentiations in media coverage characteristics between newspapers. Given that during the 2018 Wildfires the Greek
government was led by the Coalition of Radical Left party (SYRIZA), it is the belief of the author that left-wing newspapers may have adopted a moderate stance in news stories where left political sources (party leaders, MP’s, government officials) appear. The following combination of greek keywords has been used to search for news articles through each newspaper’s online database: Μάτι [mati], φωτιά [fire], πυρκαγιά [wildfire]. Articles matching the search criteria will form our article database. Then, the articles will be imported into MAXQDA Qualitative Data Analysis Software. We decided to use software
26
for our data analysis for two main reasons: (i) to increase the validity of our findings and (ii), to speed-up the data analysis process, given the limited time scope of this research.
Each article will be coded via MAXQDA for the presence of narratives and certain media coverage characteristics. To do so, nested coding will be performed. This involves going through the data twice and thus, assigning primary and secondary codes; primary coding involves an inductive data analysis process, which will be held out by reading the imported articles and identifying prospective policy narratives. This paper acknowledges that there might be more than a single narrative seen in each article, as the responsibility/ accountability surrounding the wildfire – which is the central plot arcing across narratives- may feature numerous “sub-plots”. Based on their respective indicators in the operationalization table, these elements will be identified in each document, and will then be qualified as narratives. The first four narratives with the highest incidence in our article database will be
susceptible to secondary coding. During the primary coding process, specific words or phrases with direct connotations to certain narratives will not be sought out. Due to the nature of the content of policy narratives, the process of identification will follow a more interpretive style. As acknowledged in the work of Crow and Lawror (2016), “Policy narratives are predicated on a culturally subjective understanding of linguistic usage and (often) culturally specific or emotionally laden tropes. While still intersubjective, these “human” elements of storytelling are more interpretive and difficult to associate with specific vocabulary as frames have been” (p.479).
The next step of the analysis is the assignment of secondary codes to the four prevalent policy
narratives. Unlike primary coding, this process is mostly deductive; secondary coding will be applied to narratives, by allocating them to corresponding types of media coverage. We also acknowledge that not all narratives match to certain framing strategies. For example, when measuring the balance of political sources we only coded news articles which included evaluations, debates or analyses of governmental and/ or oppositional actors. After tracing and analyzing the independent, and intervening variables, these will be compared with the policy outcomes. Ultimately, the results will be cross-checked against the initial expectations.
27
3.5 Limitations of the Research
According to Yin (2003), research validity comprises of three essential elements; construct, internal, and external validity. Construct validity is about establishing correct operational measures which spring from the study’s concepts. Thus far, all key concepts have been ramblingly presented and analyzed; the three key theoretical considerations have been combined into a tailor-made, media crisis exploitation model (Table 1).
Internal validity refers to whether the outcomes of the research (dependent variable) are due to the manipulation of the independent variable. The causal relationship between the variables has been carried out by presenting and reviewing all relevant literature. The body of knowledge of this study led us to specific and justified expectations illustrating the linkage between variables.
External validity on the other hand, is strictly related to generalizability. A common issue of single-case research designs is that conclusions are difficult, if not impossible, to become generalized in view of the fact that findings are all based on a case-specific scenario. Yin (2003, 2010), distinguished between analytical and statistical generalization. The latter refers to capacity of the research to produce statistical generalizability of research findings. In other words, statistical generalization can be defined as the extension of experimental results from a study sample to the population as a whole (Yin, 2010). Hence, generalizing the findings of this particular research to a broader population can be problematical in terms of statistics. Contrarily, Robert Yin (2003), describes analytical generalization as a process of transferability of findings from the analysis back to the conceptual framework: “ Analytic generalization is not generalization to some defined population that has been sampled, but to a theory of the
phenomenon being studied, a theory that may have much wider applicability than the particular case studied… it resembles experiments in the physical sciences, which make no claim to statistical representativeness, but instead assume that their results contribute to a general theory of the
28
in order to test the Extended Crisis Exploitation Model (see Table 1). In this vein, external validity will be improved , and findings may be generalized analytically.
In order to further enhance validity, this study payed special attention to the representativeness of samples. In an effort to mitigate bias, coded newspapers have been represented by distinguishing between right-of-center to left-of-center, based on ideological diversity. This provided us with a more balanced sampling frame (newspapers), which makes the entire study less subjective, and thoroughly ameliorates the above described statistical generalization of the research.
As it was previously mentioned, there are no specified words or phrases linking to certain codes; the process of identifying potential policy narratives will be carried out in an interpretive manner, partly based on coder’s discretion. This can be perceived as a major limitation, making it rather difficult to test the reliability. However, this study is based on an analytical, and discursive operationalization table with a clear set of indicators. Besides, the use of MAXQDA software, mitigates suchlike concerns, by providing the researcher with accurate and specific measurements illustrated in quantitative summary tables.
29
4. Analysis
In this chapter, all gathered data will be presented and analyzed with knowledge of the theoretical framework and the combined theories in the Extended Crisis Communication Model. The purpose of the research is to empirically evaluate the extent to which media framing can influence policy narratives, leading to certain policy outcomes. For the analysis, the results of all coded articles (N=55) will be allocated based on the four dimensions of media coverage as presented by Olsson et al. (2015). Following the presentation and discussion of the results, policy outcomes for each identified narrative along with their respective media coverage characteristics will be examined. A brief summary of the Wildfire crisis timeline will be provided beforehand.
4.1 Case Study: The 2018 Athens Wildfire
At 1657 hrs (local time) the fire service received the first phone call of the fire ignition point in a forest surrounding the settlement of Kallitehnoupolis located on Penteli mountain (5km to the West of Màti coastline, in the Eastern suburbs of Athens). The predominance of stormy west-northwest wind gusts and dense vegetation in the affected area, pushed the fire towards Neos Voutzás, and finally reached the densely populated coastal settlement of Mati and Rafína. The timeline of the wind direction during the disaster clearly indicates that the fire travelled swiftly from the starting point to the coastline within 90 minutes. According to the scientific report of the National and Kapodistrian University of Athens:
The wildfire, classified as a crown fire, aided by very strong winds of an approximate velocity of 90km/h, crossed the main highway of the area (Marathonos Av.) within a very
short time (estimated approx. 30 to 40 min. depending on the location), and moved to- wards the settlement of Mati, where it reached the sea with a fire front of approximately 1
30
area. The difficult weather conditions, the area morphology and the town configuration resulted in the entrapment of a great number of people ( Lekkas et. al., 2018).
The scientific team ended up with 15 preliminary concluding statements, among which and for the purposes of our study we distinguished the one dealing with crisis response: On the basis of testimonies that are systematically analyzed, it is evident that the population near the beachfront did not receive early warning information from the emergency services or civil response in the form of an evacuation order; the only available information at the time was either from individuals who were voluntarily evacuating the western part of the settlement, either through social media users in the form of
audiovisual content and text posts (Lekkas et al., 2018). The massive flames in the beachfront village of Mati trapped hundreds of people in their cars and homes, resulting in 102 deaths (BBC, 2019).
4.2 Presence of Policy Narratives (see Table 2, p.17)
Our analysis has shown that the core theme in news reports surrounding the wildfire was related to the responsibility/ accountability of officeholders, and decision makers. The majority of the articles in our dataset address questions about why did the fire occur, and who is to blame for the escalation of the crisis. Identified policy narratives in this dataset range from acts of force majeure (strong winds leading to the quick spread of blazes) to acts of inadequate crisis management. In particular, policy narratives included long-term problem discussions around the impact on health, and environment (4%),
compensations for fire ravaged population (8%), or more rarely discussions around the lack of citizen alertness during the wildfire (1%). The four dominant policy narratives included in the analysis were: (1) Civil Protection Measures have not been activated (21%), (2) Mismanagement by police and fire services (31%), (3) Unlawful Construction (11%), and (4) Fire Compensations (8%).
Figure 4 represents the number of identified dominant narratives per newspaper. The results indicate that kathimerini (N=28) had the largest narrative aggregation among newspapers. This could be mainly attributed to the fact that among coded articles, kathimerini has the highest daily circulation (see. Figure 4). Narrative presence and type varied therefore depending on the medium of publication, and possibly, on the ideological alignment of newspapers. Efimerida ton Syntakton for example (N=13), has a lower circulation of approximately 6.500-8.000 copies per day, however policy problems did not “traditionally” focus on the lack of response and recovery during, and after the wildfire; policy narratives included a
31
broader context of problem definitions, some of them focusing on scientific reports highlighting the lack of civilian self-protection, and situational awareness, or blame games accusing the right-wing opposition of politicking by “using” mourning and tragedy for political expediency. Efimerida ton Syntakton is situated within the center-left wing, and thus at the time of the crisis it might have taken a pro-government stance, while keeping a neutral position against the common frame of inadequacy among the crisis management system . Therefore, the presence or absence of suchlike policy problems could be partly associated with the political orientation of newspapers.
Figure 4: Narrative Acquisition by Newspaper
0 2 4 6 8 10 12 14 16
kathimerini proto thema efimerida syntakton
Narrative acquisition by Newspaper
Civil Protection measures have not been activated Mismanagement by police and fire services Unlawful Construction fire compensations
32
4.3 Media Coverage Characteristics in Newspapers
(see Table 2, p.17)4.3.1 Style of Reporting
As Table 3 above illustrates, Style of Reporting has been measured for each of the four dominant narratives, based on the amount of descriptive or interpretative storytelling. By comparing style of reporting among narratives, it is clearly evident that in all four narratives the majority of coded newspapers followed a factual-based reporting, based on describing the actual event as it unfolded. According to the results, the mismanagement by police and fire services received the largest percentage of descriptive journalism, whereas the reports on unlawful construction in the bushfire-affected areas had a lower deviation between styles of reporting. More precisely, 31 % of newspapers provided evidence-based explanations, judicial interpretations, and in some cases scientific reports regarding the lack of coordination between first responder agencies during the wildfire. The following citation is an illustrative example of descriptive journalism based on a scientific publication, regarding the
organizational mismanagement of authorities : “the study held by Dr. Synolakis and his team indicated that the original size of the fire at East Attica has been wrongly calculated and underestimated by the Fire Department and the General Secretariat for Civil Protection, with most firefighters operating on the fire-front of Kineta, while it is noted that aerial support was minimal.” (Souliotis, 2018).
To the contrary, a smaller number of articles deviated from the factual-based, and observational standpoint of news reporting; journalism here produced a mix of facts, analysis and personal opinion. This has been explicitly emphasized by Kathimerini in an article related to arbitrary construction in the affected areas: “Arbitrary construction and climate change. The government tried to put the blame for the disaster and the deaths of 100 people to unlawful construction and climate change.. This is their cheapest excuse. Given that the entire country is full of arbitrary buildings and also considering the fact that there is nothing to be immediately done against the threat of climate change, a large number of
Table 3: Descriptive and Interpretive Style of Reporting in Newspapers (%)
Style of Reporting Narratives
Civil Protection measures have not been activated
Mismanagement by police and fire services Unlawful Construction fire compensations Descriptive 21.8 30.9 9.1 10.9 Interpretative 9.1 12.7 5.5 0 Σ=55 (100%) 17 (30.9 %) 24 (43.6 %) 8 (14.5 %) 6 (10.9 %) N=55 (100%) 17 (30.9 %) 24 (43.6 %) 8 (14.5 %) 6 (10.9 %)
33
casualties in every natural disaster should then be considered as a normality from now on. This is wrong ! “ (Drimiotis, 2018). Based on the above table, and according to the initial expectations of the study the content of news coverage was largely free from comments and biased analyses; news media acted as conduits of policy information. The percentage of actor-driven, and interpretative narrative strategies was low among all dominant narratives.
4.3.2 Objectivity of the News
Table 4: Balance of Political Sources in Newspapers (%)
Note: Column percentages based on the sum of coded segments (row “SUM”). Only news items where political sources appear are included in the analysis
For this dimension of media coverage, only 14 (25,5%) out of 55 documents were susceptible for further coding. The analysis here looks at views, statements or direct quotations of political elements within articles, in order to measure the balance of media coverage between governmental and oppositional actors. According to Olsson et al. (2015), natural disasters are expected to produce balanced levels of representation among political sources. Table 4 illustrates this trend; In looking at the balance between actors, it becomes apparent that short-term policy problems (Civil Protection measures have not been activated, Mismanagement by police and fire services), received the highest levels of oppositional coverage. A probable explanation is that the right-wing opposition pressed the government for answers in the light of a series of mistakes and omissions make by the authorities during the handling of the wildfires. Kyriakos Mitsotakis, leader of the New Democracy opposition party, said the following: “But public figures have no right to hide during a crisis. (A leader) has an obligation to take a position, and be next to people in trying times [...] I really wonder how some people.. can sleep today, and continue in exercising their duties” (Euractiv, 2018).
Sources Narratives
Civil Protection measures have not been activated
Mismanagement by police and fire services Unlawful Construction fire compensations Government 14.3 7.1 14.3 28.6 Opposition 14.3 14.3 7.1 0 Σ = 14 (25.5%) 4 (28.6%) 3 (21.4%) 3(21.4%) 4 (28.6%) N= 55 (100%) 17 (30.9 %) 24 (43.6 %) 8 (14.5 %) 6 (10.9 %)
34
Contrarily long-term policy problems, such as addressing the risk of unlicensed residential building in the wildland-urban interface, or the discussion related to fire compensations was mainly dominated by governmental sources. Apparently, the left government of SY.RIZ.A led by Alexis Tsipras tried to shift the focus away from the operational deficiencies of the crisis management system. In an effort to mitigate public anger, Greek PM promised to enforce a list of measures, including “ a crackdown on unlicensed construction which is thought to have worsened the fire and blocked off escape routes for residents fleeing toward the sea” (Euractiv, 2018). On the 26th of July, 3 days after the wildfire, Greek Minister of Defense Panos Kammenos visited the fire-ravaged settlement of Mati and made the following
statement: “This (Arbitrary construction) is a crime of the past, because at the coastline of Mati, the majority of all these houses have been built without permission. Residents have occupied the coastline without complying with the legislation. After such a tragedy, it is time to understand that failure to comply with the Law is dangerous for them (residents) and their families as well” (Kathimerini, 2018). It is worth mentioning that due to the low number of documents in which political sources appear, and also because of the fact that this research investigates exclusively only three national newspapers, political actors are expected to appear to a larger extent than what is shown in our analysis. Hence, a larger sample might have produced a more limpid overall picture of balance distribution between sources.
35
4.3.3 Tone of Reporting in Newspapers
Table 5: Tone of Reporting in Newspapers (%)
The tone in media presentations is one way of “measuring the framing power of political actors” (Olsson et al., 2015: 168). As demonstrated in Table 5, we can see that a large majority of coded articles ( 89% in total) have been reported on in a negative tone. According the Operationalization Table, negative coverage manifests when involved actors (e.g. Government, Civil Protection, Police and Fire Services) are being framed in a negative and condemnatory manner. In particular, for the narratives dealing with the failures of the Greek Civil Protection as well as the inadequacies of first responder authorities, none of the articles in our dataset received a positive attribution. Here, all coded newspapers have adopted a critical, but factual-based stance against the SY.RIZ.A government and state authorities. The following quotation from the left-wing Efimerida ton Syntakton abstracts these results: “Operational dysfunctions, errors and omissions have been mainly traced in actions of coordination, communication and
co-operation among the forces of Civil Protection . The magnitude of disconnect among state agencies was so big and chaotic, that no one really knew who the head of the ground forces was.” (Efimerida
Syntakton, 2019).
Tone of reporting in articles related to fire reimbursements appears to be more evenly distributed between actors. We found this especially interesting in the case of a long-term narrative related to recovery measures taken by the government. The explanation for this result is that all coded data has been collected from the date of the crisis event (July 23rd, 2018) and up until November 2019. Within this time period, national elections have been held in the country, and from July 2019 onwards, the
Tone of Reporting Narratives
Civil Protection measures have not been activated
Mismanagement by police and fire services Unlawful Construction fire compensations Positive 0 0 1.9 5.8 Negative 30.8 44.2 13.5 3.8 Σ= 52 (94.5%) 16 (30.8%) 23 (44.2%) 8 (15.4%) 5 (9.6%) N= 55 (100%) 17 (30.9 %) 24 (43.6 %) 8 (14.5 %) 6 (10.9 %)
Note: Column percentages based on the sum of coded segments (row “SUM”). Only news items where frame-based or actor related sources appear are included in the analysis
36
center-right political opposition of Kyriakos Mitsotakis and his New Democracy party, won a majority in the Parliament, and formed a new government. In light of the above, many articles in our database have been published after the July 2019 elections, where the disaster relief measures of the current conservative government have been reported on in a positive tone. Specifically, all three (5,8%) of the positively flagged articles have been written after July 2019, and they were all related to the actions taken from the newly-elected government in favor of the population concerned. By extension, the tone in newspapers covering the four prevalent narratives of the wildfire has been overwhelmingly negative for the government of Alexis Tsipras who was in charge during the crisis event. In accordance to Olsson et al. (2015), the high amount of frame-based negative coverage strongly mirrors the polarization it caused among political actors, and indicates that the crisis has been framed by the media as a highly politicized natural disaster. Politicization in this context, was not triggered by media coverage. Instead, the absence of actor-related journalism bears testament to the fact that news media acted as simple conduits of information.
4.3.4 Framing of Newspapers
Table 6: Framing of Narratives in Newspapers (%)
Frames Narratives
Civil Protection measures have not been activated
Mismanagement by police and fire services Unlawful Construction fire compensations Issue 30.8 40.4 9.6 5.8 Game 1.9 5.8 5.8 0 Σ = 52 (100%) 17 (32.7 %) 24 (46.2%) 8 (15.4%) 3 (5.8%) N= 55 (100%) 17 (30.9 %) 24 (43.6 %) 8 (14.5%) 6 (10.9 %)
37
The final dimension of media coverage is framing, “an essential feature for understanding news media characteristics” ( Cappella & Jamiesson, 1997; Entman, 2003; Iyenger, 1994; Semetko & Valkenburg, 2000; as cited in Olsson et al., 2015: 161). Here, we coded items as issue frames if the news stories were factually represented, and policy problems were simply described as they unfolded. Articles coded as game frames ”portray politics as a race with winners and losers at the expense of substantial policy issues” ( Olsson et al., 2015: 162). As it can be seen from Table 6, media coverage of all policy narratives focused on the event (issue framing), instead of putting the spotlight on political power plays, or
exploiting the disaster for political purposes. Here, narratives related to the shortcomings of the
response system are dominated by issue frames. Similar to the results presented in the dimension ‘style of reporting’, journalists framed the wildfire with a focus on the respective actors falling short in
performing their tasks in the line of duty. One might argue that the intense criticism towards the errors and omissions of both emergency service senior officials (heads of Police & Fire departments) and local, regional or national decisionmakers (mayors, governors, ministers, and the Greek PM), should produce higher levels of game frames for each identified narrative. According to the analysis of media reports, the overall picture of the State’s crisis response during, and in the aftermath of the wildfire indicates that the majority of newspapers provided a clear, unbiased and apolitical picture on the inadequacy on handling the crisis. Issue framing here highlights that journalism focused on a factual-based description of the errors and omissions surrounding the wildfire. The following citations from Proto Thema and Kathimerini emphasize this:
1. There was a 30-minute delay in mobilizing the fire brigade for the fire at Ntaou Pentelis. 2. a helicopter that was monitoring the fire in Kineta, was later dispatched to the Wildfire in Mati, however it eventually ran out of fuel.
3. The Fire Department's operations center was in total disarray and reacted wrongfully when evaluating the magnitude of the Wildfire, despite the fact that there have been reports from officers coordinating the ground forces in Mati.
4. meteorological data showed that the wind gusts in the area had a speed of 70 km / h and not 110km / h as government officials claimed
5. The settlement of Mati was not on the Fire brigade's list with the 36 areas of Attica that include arbitrary construction in forestall territories. As a result, no special protection
38
6. Greek Police made a fatal mistake by setting up diversions which drove people into Poseidon Avenue in Mati when Marathon Avenue was closed.
7. The regional coordinating body convened only after the Wildfire.
8. The report contains detailed information on the actions that should be taken by senior officers on the basis of the existing legislative framework
(Proto Thema, 2018)
Heavy responsibilities also arose against officials of the General Secretariat for Civil Protection and the Regional and Local Government. As it is already known, the regional coordinating body convened only after the disaster, no evacuation procedures were initiated, and residents were not timely informed about the imminent danger of the fire. Even though the integrated emergency communications service (112) has already been tested, it was not launched by the General Secretariat for Civil Protection.
In addition, independent wildfire experts identify inactions in taking all necessary
precautionary measures, such as road cleaning from debris, and removing fuel in case of a wildfire.
39
4.4 Policy Outcomes
(see Table 2, p.17)4.4.1 The Wildfire as a framing contest
In order to answer the main research question of this study, and proceed with the presentation of policy changes for every narrative, it is of utmost importance to identify the framing strategy that the Greek Government engaged in. As discussed earlier, Boin et al. (2009) proposed three dominant types of framing contests that manifest in the wake of unpredicted events. Among them, the ‘Crisis as Threat’ frame is being adopted by political actors in cases where the existence of an imminent and critical threat to the collective good creates the need to “defend the agents (incumbent office-holders) and tools (existing policies and organizational practices) of that status quo against criticism” (Boin et al., 2009: 84).
Following the deadly Wildfire, the first official joint press statement was given by the Minister of Public Order and Citizen Protection, Nikos Toskas and the Government Spokesperson of the Hellenic Republic, Dimitris Tzanakopoulos. Toskas categorically denied the occurrence of any “noticeable” errors or omissions on the part of the Fire Brigade or the Police. Instead, he spoke of criminal acts and arson, claiming that there is evidence to support that conclusion:
"First, I would like to express my deepest sorrow for the loss of so many innocent people. It is our responsibility to protect the natural environment. Concerning the investigations carried out by the fire and police departments, I would like point out that we have serious and substantive evidence of criminal activity, arson. A gradual amount of fires occurred within a very short time in the area of Penteli. There are several issues that concern us. It’s no secret that there has been a finding which led us to carry out fast-track investigations. We are also concerned about the latest arson point in Nea Makri, which was detected 300 meters from another fire front. I would like to emphasize to the extreme climatic conditions throughout the region without forgetting the self-sacrifice of all firefighters, police and volunteers who helped in reducing the number of casualties[...] As far as I am concerned, I take full political responsibility. I am still trying to figure out who is to blame, but I can reassure you that no grave operational mistakes have been made. Of course, the losses are great and they have surely shocked all of us, and above all those people who were
responsible for managing this situation.”
