Floods and heavy rainfall wreak havoc in : The perception of causes, vulnerability and responsibility in media coverage of disasters

- an analysis of debates on climate change, development issues and responsibilities in media coverage of three in Kenya between 2013 and 2017

Bachelor thesis Uppsala University 2018 Author: Mikaela Ehrlin Supervisor: Frederike Albrecht Number of words: 13032

1 Table of contents

1. Introduction 3 2. Theoretical background 5 2.1 Climate change in Kenya 5 2.2 Development issues that causes vulnerability to climate change 7 2.3 Responsibility in relation to climate related hazards; international community, government and citizens 8 2.4 Media framing 8 2.5 Previous research on media framing of climate change in relation to natural hazards 9 3. Material 12 3.1 EM-DAT – the floods 12 3.2 Factiva – the news media 12 Table 1. 13 4. Method 13 4.1 Frame analysis and qualitative content analysis 13 4.2 Operationalization 15 5. Results and analysis 19 5.1 Is the author making any links between the floods/rainfall and its effects to climate change or global warming? If so, is climate change described to be the main cause of the /rainfall and its effects and to what extent is it actively framed as a climate change issue? 19 5.2 Are development issues described as a cause of the floods and its effects, and to what extent is it actively framed as a development issue? 19 5.3 What other issues are dominating the discourse of the articles? 21 5.4 Who is named responsible for action? The government, the international community or the citizens? 22 5.5 The articles that did not relate to a particular event 24 6. Conclusion 25 7. References 28

2 1. Introduction

Climate change is a global phenomenon that will affect all parts of the world in the form of higher temperatures, , and extreme events. Some countries have better capacity to withstand these changes than others. What is an accepted fact is that developing countries are the most vulnerable to climate change (UNFCCC 2017). One of these countries is Kenya, where climate change is already clearly visible in the form of natural hazards such as severe and floods. Media has an important role in educating and spreading information about climate change to the population (Miles & Morse 2006: 366, Tagbo 2010: 6). In the context of a developing country, many studies on media coverage of climate change have been made (Azmi et. al 2015, Takahashi & Meisner 2012, Cramer 2008, Tagbo 2010, Zamith, Pinto & Villar 2012, Gordon, Deines & Havice 2010). However, these studies mainly studied the extent of the climate change discourse in media in a general sense, and do not further elaborate on how the media discusses the relation between climate change and natural disasters. These kinds of studies have only been made in the context of industrialized countries. Therefore, this study would like to contribute to previous studies by investigating to what extent the media coverage relates natural disasters to climate change in the context of a developing country, namely Kenya. Bohensky & Leitch (2013) analyzed where the public attention was directed after a flood in Australia. They studied the media framings of the flood with a focus on the perceived links between the event and climate change, and the perceived roles of the government in managing the flood (p. 475). Their study showed that the media coverage on the floods included an elaborated discourse on climate change, including a belief that a link can be made between floods in general and climate change, denial of such relationship, confusion about the relationship, and belief that the particular flood is a proof of climate change (p. 479.). In comparison, Tagbo (2010) conducted a study where she interviewed journalists from Nigeria and South Africa, who argued that climate change may be an important development issue, but it is overshadowed in media by other issues like corruption or other political issues (ibid: 34, 37). Therefore, it is understandable that the climate change discourse in industrialized countries is very different from media discourses in developing countries, that may not be prioritized over other development issues. However, it is of interest to study whether this is true in the case of Kenya in the media coverage of floods. Hence, in addition to adopting a

3 similar framework as Bohensky & Leitch (2013) used to study if any links are made between the floods and its effects to climate change, it is of interest in this study to analyze if development issues are described as causes of the floods and its effects. Thereby, this study addresses if the Kenyan media identifies any development challenges that causes vulnerability to events that Kelman (2014) emphasizes is important in order to increase the capacity to withstand hazards (p. 120). Moreover, this study also analyzes who is named responsible to act when a disaster occurs. According to previous research, responsibility is often addressed to the government (Bohensky & Leitch 2013, Albrecht 2017, Ngcamu & Binza 2015, Barnes et al. 2008). In sum, this study aims to investigate to what extent and how the Kenyan news media relate extreme events in the form of floods and heavy rainfall to climate change, if development issues are described as causes of the flood and its effects, and who they name responsible to act. The analysis consists of Kenyan news media coverage on floods that occurred in 2013, 2015 and 2017 from English-speaking news media, namely Daily Nation, The East African, The Star and Business Daily. Frame analysis and qualitative content analysis are the methods used to conduct this study. This method was assessed as the most appropriate because the intention of this study was to analyze how the disaster situation is framed, and also to read the content of the news reports in its entirety and more intensively. In the following section, previous studies that relate to this study are more thoroughly elaborated to form a theoretical framework for the analysis. In the third section, the data material is listed. In the fourth section, the method that was employed to conduct this study is explained and discussed. The fifth section presents the results. The last section is the conclusion that ties together the study.

4 2. Theoretical background

2. 1 Climate change in Kenya It has become an accepted fact that the negative consequences of climate change are and will be affecting the developing countries more than other parts of the world (Kamau & Mwaura 2013: 153). The adaptive capacity to is low due to limited technological, institutional and financial capacity with low levels of economic development experienced across the continent (Awaour, Orindi & Ochieng 2008: 231). Nevertheless, climate change threatens to ruin decades of development efforts that have been achieved in most countries (ibid.). In Kenya, the impacts of climate change will be experienced in ways such as increased temperatures up to 48 degrees Celsius and rainfall variability by up to 20 % by the year 2100 (ibid.). Temperature rise is common across the country and rainfalls has become more irregular and intense (GoK 2010: 5). The Kenyan coast is one of the most vulnerable to sea level rise, where the second largest city, Mombasa, is situated (GoK 2010: 32). Mombasa has a history of disasters related to climate extremes including floods, which cause serious damage nearly every year as well as loss of lives (Awour, Orindi & Ochieng 2008: 231). Flooding is also a major problem in all informal settlements in Nairobi (ActionAid 2006: 4). Houses are built of weak, inadequate building materials, and migration has led to more houses being built close to rivers, meaning greater disruption when floods occur (ibid.). To many residents, the floods associated with El Niño are particularly severe (ibid.). Moreover, climate change is making weather less predictable, more uncertain and heavy rainfalls more likely (ActionAid 2006: 3). Many parts of the country experience unexpectedly heavy rainfall in mid-April which continues through the end of May, called the long rains, and from September to November, called the short rains, with some areas more prone to flood disasters (Nyakundi et. al. 2010: 347). Torrential rains have already caused destruction of various infrastructure in Kenya (GoK 2010: 38). Even though Kenya is evidently affected by climate change, Shisanya & Khayesi’s (2007) study showed that among the 132 respondents to their survey in the city of Nairobi, the respondents did not perceive climate change to be a very significant problem in relation to other threats to socioeconomic security in Nairobi (ibid: 281). The issue of climate change takes very little space in the context of unemployment, poverty, corruption, insecurity, street children and a degraded human condition in Nairobi (ibid: 281).

5 Tagbo (2010) investigated the media coverage of climate change in Nigerian and South African media, and also provides the perspectives of climate change journalists. There seems to be a consensus among the South African and Nigerian editors that climate change is not a front-page subject (p. 34). One assistant environment editor of the Nigerian The Guardian believes that climate change has become an important development issue but is overshadowed by other issues like corruption or other political issues (ibid: 34, 37). Moreover, a lack of understanding of the issues in climate change makes it difficult to report it, and there are also no more than two environment reporters on the newspapers included in Tagbo’s study, compared to the British The Guardian that has up to six environment reporters (ibid: 34). The priority given to environmental issues is strongly influenced by other development issues that the region is already dealing with and turbulent political story of many African countries (ibid: 35). At the same time as concerns about climate change started to be mentioned in Western media, about two decades ago, the agenda of South African media was political freedom and the fight for fair treatment (ibid.). Hence, climate change is a relatively new subject in many African media (ibid: 36). Publications on the subject in African newspapers and magazines are culled from foreign institutions and researchers, the problem is that much-needed local angles to the issue are often left out (ibid: 37). In relation, African citizens are said to be the least informed about climate change (BBC 2009: 2). However, research by the BBC World Service Trust Research were group discussions with people in Nigeria, Ethiopia, Ghana, Kenya, Nigeria, Senegal and Uganda showed a counter-perspective to the assumption that African citizens have more urgent issues to think about than their environment and climate change, as the people they interviewed are deeply concerned about the changing environment (ibid: 5). They argue that “the poorer people are, the more they understand just how vulnerable they are in shifts in the life support system that sustain them, and how poorly protected they are from the extreme consequences of a changing climate” (ibid.). Citizens need to have the right to information and how climate change will impact their lives and how they can adapt (ibid: 2). Moreover, African citizens have a major role in educating those responsible for climate change about its true impact and communicate their perspectives and experiences (ibid.). The media in Africa, together with schools, are critical sources of information on climate change (ibid.). People affected by climate change will be better able to adapt if they are informed about the issue, as farmers may change varieties of the seeds they plant, and floods and droughts need to be planned for, and livelihoods and businesses changed and adapted (BBC 2009: 2).

6 2. 2 Development issues that causes vulnerability to climate change Kelman (2014) emphasizes how the rhetoric of climate change being the biggest development challenges that faces humanity can overshadow other development challenges which need to be addressed simultaneously with climate change (ibid: 120). Focusing on climate change can depoliticize the challenges that the country face, such as community reconstruction and other long-term development challenges that could reduce vulnerability to hazards (ibid.). Kelman argues that nature provides input through environmental events which are potentially hazardous, but human actions have put people at risk without the capacity to deal with the hazards (ibid: 121). To highlight climate change rather than the vulnerabilities to all hazards disregards political root causes of the vulnerabilities and unresolved development challenges (ibid.). For example, the Maldives’ President from 2008-2010, Mohamed Nasheed, highlighted how climate change hazards threaten his country, which provided a convenient distraction from other development difficulties (ibid: 122, 123). Kelman emphasizes that “the key is not to ignore climate change, but to examine it in the context of vulnerabilities to all hazards” (ibid.). When rebuilding a community after a disaster, one needs to be aware to “build back better”, to not to rebuild it with the same vulnerabilities as before (ibid: 125). Moreover, Kelman puts forward one type of vulnerability called livelihood decisions, a vulnerability that tend to be exposed when a hazard manifests, but it is not the hazard in itself that is the fundamental problem (p. 121). This emphasizes political root causes that vulnerable people are facing, such as poor resource management, marginalization, exploitation and lack of choices available that put people at risk (ibid.). Bohensky & Leitch (2013) found in their study that residents advocated broad lessons for the region’s recovery and reconstruction which suggested that lessons for the future was being considered (2013: 483). Key infrastructure was such a vulnerability that was recognized, especially transport infrastructure to ensure that communities and industries are not left isolated (p. 484). Boykoff & Roberts (2007: 32) highlighted in their study that Mozambique have institutional and economic problems inherited from its colonial past, and the absence of basic protections resulted in a human disaster when the Connie, Elaine and Gloria struck in early 2000. Additionally, another country with colonial legacies is Honduras that was affected by Hurricane Mitch in 1998, illustrates why poorer nations with socioeconomic vulnerability and institutional fragility have a “structural disadvantage” in preparing for and responding to hazards (ibid: 18, 19).

7 Hence, flood hazards that affect Kenya are natural phenomena, but damage and losses from the floods are the consequence of human action (ActionAid 2006: 2). An example is with more and more people moving into cities, which aggravates hazards like flooding by restricting where overflow of can go, which intensifies the flood effects and risks (ibid.).

2. 3 Responsibility in relation to climate related hazards; international community, government and citizens Billet (2010) studied the representation of the climate change issue in Indian mass media, and found that India to a main part, 76, 3 percent of 156 articles, assigned responsibility for climate change to industrialized countries and framed the historical emissions of greenhouse gases as the cause for climate change (p. 8-9). Some articles argued that India must prioritize economic growth and poverty, and mitigation efforts should be undertaken by industrialized countries (ibid: 10). Current and nearby climate change was thus framed as being the international community’s responsibility (ibid: 10-11). Albrecht (2017) and Bohensky & Leitch (2013) had a main focus in their studies on how the media perceived the government’s role in disaster management after the floods in the UK and Australia respectively. Ngcamu & Binza’s (2015) study of the media framing of natural disasters in the Kwazulu-Natal province found that the media mainly reported that the government was responsible in the aftermath of disasters, with little responsibility being taken by individuals and communities (p. 115). In their study of media agenda setting during and after Hurricane Katrina, Barnes et. al (2008) found that the media mostly, 40 % of 1590 articles, emphasized the accountability of the federal government and its need to respond to a disaster (p. 604, 606). Less often was the individuals’, families’ and communities’ level of preparedness or responsibility addressed, only in 13,8 percent of the articles (ibid.). Thus, the media supported the view that the government bears the primary responsibility in the event of a disaster (ibid: 609).

2. 4 Media framing The news media serves several functions, such as informing and educating (Miles & Morse 2006: 366). The main purpose is to spread information to the population (ibid.). To those affected by a disaster, the media provides important information as objective observers in their reporting (ibid.). Miles & Morse (2006) argue that as the same information is available

8 to all, the “mass media enables common social experience in heterogeneous societies, in this sense, mass media is itself a form of social capital” (Miles & Morse 2006: 366). According to Scheufele & Tewksbury, framing is “based on the assumption that how an issue is characterized in news reports can have an influence on how it is understood by audiences” (2007: 11). Frames presents complex issues efficiently and in an accessible way (ibid: 12). Furthermore, Entman argues that frames “highlight parts of the information about an item that is the subject of a communication, and thereby elevating the highlighted parts in salience, meaning that one makes the information more noticeable, meaningful or memorable to the audience” (1993: 52). In short, frames are the themes that emerge in media representation of an issue (Tagbo 2010: 270). Tagbo argues that African media can have an important role in spreading information about the climate to effectively guide public debate and understanding about climate change (2010: 6). Moreover, regularly broadcasted communication about climate change is the first step towards developing coping mechanisms in Africa, and effective information efforts is crucial if African citizens are to gain capacity to respond to climate change impacts (ibid: 6, 17). Trumbo argues that the first step in solving a problem is to recognize that it exists, and representation in the news media is a crucial part of social problem recognition and definition, for example when it comes to climate change (1996: 269).

2. 5 Previous research on media framing of climate change in relation to natural disasters Studies of media framing of climate change in relation to natural disasters have, among others, been conducted by Bohensky & Leitch (2013) and Albrecht (2017). Bohensky & Leitch analyzed the public attention after the Brisbane flood in 2011 using the news media for their data material. Their result was that the public attention was directed to the causes of the events and lessons for minimizing future floods (p. 475). With their newspaper analysis, they focused on learning as an aspect of resilience in relation to two themes; firstly, perceived links between the flood and climate change and secondly, perceived roles of government managing the flood (Bohensky & Leitch 2013: 475). The authors explain well how the news media was useful for their study, as news media “both reflects and shapes public opinion by defining and limiting public discourse around key events and contributes to the sense making and informal learning of the community” (ibid: 476). The authors argue that disasters can provide windows of opportunity for learning as governments can help communities to become less vulnerable and correct flaws or social inequities while rebuilding (ibid.). On the other hand, disasters can also inhibit resilience as governments reinforce vulnerabilities when

9 they act hastily to restore normalcy to an affected region (ibid.). The main narratives about the relationship between the Brisbane flood and climate change was identified in the media sample as (1) belief that a link can be made between floods in general and climate change, (2) denial of a relationship between the flood and climate change, (3) confusion about the relationship and (4) belief that the 2011 Brisbane flood is proof of climate change (ibid: 479.). Even though their study found varying results between denial and belief of the relationship of the flood and climate change (ibid: 486), it shows that the discourse on climate change takes a great deal of space in the reports on the flood. Albrecht (2017) studied how media discourses can explain the significantly varying political consequences after natural disasters of similar severity. Albrecht compared media coverage of the 2005 and the 2015 floods in the UK, with a focus on framing and actors who express political claims. The coverage had an important difference; the satisfaction with the government increased following the 2005 floods, but the 2015 floods led to the resignation of the Environment Agency Chair and government sceptical public opinion polls (ibid: 131). Albrecht found that critics of the government decided not to blame the national government following the 2005 floods, which enabled the national government to exploit the disaster and be perceived as a credible disaster manager (ibid: 142, 149). Following the 2015 floods, critics held the government accountable for the consequences of the disaster and the lack of preparedness, which led to negative political consequences for the government (ibid: 142). Currently, there are several studies of media framings of climate change with the perspective of a developing country that looks at to what extent and frequency climate change is talked about in the media, and in what framing, but in a more general sense. For example, studies have been done of Malaysian (Azmi et. al 2015), Peruvian (Takahashi & Meisner 2012), Western Cape (Cramer 2008), Nigerian and South African (Tagbo 2010), a comparison between North- and South American (Zamith, Pinto & Villar 2012), and Mexican (Gordon, Deines & Havice 2010) newspapers and media. A few other studies mention how natural hazards have made the coverage of climate change increase in media and that the media recognizes the link to climate change. However, it has not been the main theme of the study and do not provide a deeper analysis of the matter. For example, Midttun et al. (2014) compared media framings of Norwegian, Chinese and Ghanaian newspapers. They found that the severe that affected a large part of China in 2010 was a major discursive theme that they found in the sample, and most of these articles referred to global warming as one possible cause for such extreme weather events (p. 1282). In Ghanaian newspapers, they found that in 2008 and 2010, the articles made sense of

10 flooding and as effects of climate change, and climate change understandings in newspapers were largely expressed in terms of changes in Ghanaian weather (Midttun et al. 2014: 1284). Shanahan (2009) studied climate change coverage in several non-industrialized countries, such as Brazil, China, Honduras, India, Jamaica, Mexico, Mozambique, South Africa, Sri Lanka, Swaziland, Vietnam and Zambia. Out of these countries, China (p. 148), Jamaica (p. 150), Mexico (p. 150), Swaziland (p. 152), and Zambia more or less linked the climate change coverage to natural hazards they had been affected by. Billet’s study on India (2010) found articles on the two catastrophic floods in Mumbai 2005 and in Delhi 2007 that referred to the disruption of the regular monsoon cycle (p. 7-8). Osindo (2014) focused on Kenyan media, namely the Daily Nation and People Daily from January 2012 to June 2012, with a total of 127 articles on climate-related issues (p. xi). Osindo found that the “effects” frame was most commonly used, describing effects on lives and property and also in the form of natural hazards, with aspects of the earth’s climate as the cause (ibid: 31). However, the study did not go deeper into this matter. Examples of studies done on media framing of natural disasters are Ngcamu & Binza’s study on the Kwazulu-Natal province (2015), and Zaheer’s study on the role of mass media in Pakistan in disastrous situations (2015). Again, neither of these studies had a theme around climate change.

In sum, previous research has identified that in the media coverage of climate change in industrialized countries, it is not a question whether climate change is brought up, it is mainly a question of how it is debated. However, climate change discourses in developing countries, and in Kenya in particular, could be very different. As Tagbo emphasized, the priority given to environmental issues in African media might still be influenced by other development issues that the region is dealing with, and that the climate change issue is a relatively new subject (2010: 35, 36). In this respect, this study adopts the framework by Bohensky & Leitch (2013) to analyze whether any links are made between the floods that affected Kenya in 2013, 2015 and 2017 and climate change in the news media, while further investigating whether they link the floods to development issues that causes vulnerability to hazards that Kelman highlights is important in order to increase capability to withstand them (2014: 121), which in this study are infrastructural problems and livelihood decisions. In addition, as previous research has shown, there is often a discussion in media about responsibility in relation to natural hazards. Therefore, this study will explore who is named responsible to act when a disaster occurs in Kenya.

11 3. Material

3.1 EM-DAT - the floods The information about the floods in 2013, 2015 and in 2017 was collected using the EM-DAT database. This database contains a collection of data about different disasters, such as natural, climatological, technological and epidemic that has happened from the year 1900 until today. It provides information on the human impact of disasters, such as the number of people killed, injured or affected, and also disaster related economic damage estimates and disaster-specific international aid contributions (EM-DAT). It is possible to find information about disasters in only one country, which was a useful quality for this case study about Kenya. The duration of the disaster is also presented in EM-DAT, which together with two additional months after each end-date of the flood, constitutes the limit for the amount of data collected in Factiva. This date limit provided by EM-DAT serves mainly as an orientation around when an extreme event has happened in Kenya, with reservation for that similar events that may have happened within the additional two months may occur in the analysis due to the effects of the rainy season.

3.2 Factiva - the news media Factiva is an international news database, combining over 36,000 sources from 200 countries in 28 languages (ProQuest 2018). It provides access from a wide range of information from newspapers, newswires, industry publications, websites, company reports and more (ibid.). The English-speaking Kenyan news media that are accessible through Factiva about the selected floods are the Daily Nation, Business Daily, The Star and The East African.

To begin with, when searching for material in Factiva, the first key words were coded flood* OR * OR downpour* OR El Niño, in order to include variations of the weather event in the material. The material was extended with two additional months since there was an expectation to find discussions about the floods and rainfall after the immediate effects of the event and during the time of recovery and rebuilding. According to a study by Houston et al. (2012), mass media on average covers natural disasters for a shorter period of time than other issues (p. 619). This correlated with this sample as well, since the attention given to the floods became considerably less after a month, thus it is likely that the data material did not have to be extended further.

The three floods that are selected for this study are presented below.

12

Table 1.

Year Duration (EM-DAT) Extension of Analysis Number of articles (Factiva)

2013 March 10th - April 30th March 10th - June 30th 63

2015 December 1st - January 6th December 1st - March 6th 33 2016 2016

2017 May 8th - May 16th May 8th - July 16th 29

Material that has also been taken into account were articles that did not relate to any particular flood, and hence was not included in the matrix. However, it was noticed during the process of the analysis that it existed a discussion about climate change and natural hazards that did not necessarily relate to a particular event. Therefore, they were considered valuable to capture the discourse about climate change and natural hazards in its entirety during times of extreme weather events occurring in Kenya.

4. Method

4.1 Frame analysis and qualitative content analysis The methods used for this study is frame- and qualitative content analysis. According to Esaiasson et. al. (2017: 218), social phenomenon do not have a given meaning, they can be framed in different ways. In other words, they can be interpreted and represented differently. The authors give examples on questions that can be systematically asked to the data material, which are often about how the nature, the cause and solutions of a problem are framed (ibid.); (1) How is the problem framed, or, what sort of problem is it framed as? (2) Which kinds of actors are included in the framing and what action capacities are they represented with? (3) Who or what is represented as the cause of the problem? and (4) What action alternatives are made possible, given how the problem is represented? (ibid: 219). The framings in this study is explored by finding answers to what is described as the

13 cause of the flood in the news media, and to what extent it is framed as a climate change issue or a development issue. Further, this study also provides a qualitative content analysis to identify what other issues are dominating the media coverage of the floods and who is named responsible to act on the flood and its effects. Additionally, to identify to what extent that the news report relates the floods or rainfall to climate change or development issues, it also requires to include news reports that do not mention climate change or development issues. According to Esaiasson et. al. (2017), who calls this method qualitative text analysis, this method is suitable for studies that aim to read the content of the data material in its entirety and more intensively, and also to highlight certain aspects of the material (p. 211). Drisco & Maschi (2015: 88) explain that the focus of qualitative content analysis is to identify categories or themes that summarize the content and highlight key content in the data. According to Sandelowski (2000: 338), this includes an effort to understand the manifest, frequency and means, and also the latent content of data. Thus, latent content of this data material could for example be when a bridge is described to have collapsed because of the rains, where it is evident, but not explicitly expressed, that the bridge is not capable of withstanding heavy rain. A negative aspect of qualitative method is the risk of subjective interpretation of the material, that another researcher may interpret the material differently (Drisco & Maschi 2015: 91). According to Drisco & Maschi, it is central to qualitative content to make explicit how and why interpretations are made (2015: 91). Therefore, this study provides a matrix of the analysis with quotations from the articles included, which is available in the Appendix. Furthermore, another aspect with this particular study is the limited possibility for triangulation, which is the ambition to work with different approaches and with a variety of empirical data to minimize the risk of being biased in the analysis (Wodak 2001: 65). For example, it could have been fruitful to collect material from different types of social media in the study, or to have conducted interviews similar to Tagbo’s (2010) study, in order to gain further understanding of the public perception on the relation between floods or rainfall and climate change in Kenya today. Moreover, only a limited sample of the news media in Kenya could be analyzed in this study due to the limited time and scope of the thesis. A quantitative method could perhaps have been used in order to sort a wider sample of news media and see where climate change and floods and rainfall have appeared together in news media, and thereafter investigated those samples more thoroughly (Esaiasson et. al. 2017: 198). However, the aim with this study is to further explore the content of each news report during the time of a flood by also analyzing how the news reports address vulnerability and

14 responsibility towards the disasters. Therefore, climate change must not be mentioned in the article in order for issues on vulnerability and responsibility towards extreme weather events to be addressed. This could be expressed in many different ways. Also, climate change might be mentioned in an article, but not in relation to a particular flood or to a particular development issue. Thus, qualitative content analysis is chosen in order to understand these discussions by reading about them in its entirety. Kracauer (1952: 631) argues that quantitative methods may neglect qualitative explorations, which can reduce the accuracy of the analysis. Moreover, there is only English-speaking newspapers included in the study, and there are only four newspapers. However, Daily Nation, The Star, The East African and Business Daily belong to the group of Kenya’s main news sources, which could arguably generalize the public opinion in Kenya by the data material for this study. Lastly, to include the media coverage of the recurrent droughts could have been fruitful in order to broaden the scope of the climate change discourse in relation to natural disasters. However, the duration of the droughts in Kenya are difficult to limit, which is noticeable in EM-DAT that dated the droughts over a year or two years at a time. Thus, it would have been difficult to motivate a certain time frame, or to use all data available in Factiva about droughts. However, this study analyzes three different floods and have a few years apart, which motivates that the data included in the study is large enough to measure the climate change discourse in the Kenyan news media, along with debates on development issues and responsibility in relation to natural hazards. Drisko & Maschi (2015: 92) argue that qualitative research generally seeks to be meaningful in context rather than universally applicable. However, the generalizing ambition with this study is to explain what is payed attention to in news media when a weather related disaster has occurred in the context of a developing country that is vulnerable to the impacts of climate change.

4.2 Operationalization The news reports about the floods are the units of analysis that are collected with the help of Factiva. To find the answers for a study with the help of content analysis, it is necessary to concretize the overall problem formulation with more precise analytical questions that will be asked to the material (Esaiasson et. al 2017: 216). The analytical questions and the analytical framework for this study are outlined below.

15 1. Is the author making any links between the floods/rainfall and its effects to climate change/global warming? If so, is climate change/global warming described to be the main cause of the flood/rainfall and its effects and to what extent is it actively framed as a climate change/global warming issue? 2. Are development issues mentioned as cause of the floods and its effects, and to what extent is it actively framed as a development issue? 3. What other issues are dominating the discourse of the articles? 4. Who is named responsible for action? Is it the Kenyan government, the international community or the Kenyan citizens?

These questions have an open approach to the material. That is, the answers to these questions are determined by what is found in the content of the data material (ibid: 223). However, the analysis is limited around these analytical questions, together with the categories that are illustrated in the analytical framework below.

Article about flood/rainfall/downpour/El Niño

Development issue main Climate Other cause change -Livelihood issues main cause decisions -Infrastructural problems

Who is responsible to act? The government? The international community? The citizens? None?

The content of the articles was read through, analyzed and sorted in the matrix that is available in the Appendix, with the categories Date, Author, Newspaper, Headline, Links to

16 climate change/global warming, Climate change as main cause, Development issue as main cause, Other issues and Responsibility. Firstly, if an article is describing climate change as the cause for the flood or rainfall and framed actively as a climate change issue, it could for example be described that the heavy rain that causes floods in the country is a result from climate change, and that measures need to be done in order to mitigate further impacts of climate change. Secondly, in the column for development issues, it is also highlighted whether or not the article is actively framed in a way that represents infrastructural problems or livelihood decisions as the cause of the disaster. Infrastructural problems are thus the faults in the construction of bridges and roads (Bohensky & Leitch 2013: 483) as well as drainage and housing (ActionAid 2006: 4). Livelihood decisions are political root causes such as marginalization and lack of choices available that put people at risk (Kelman 2014: 121). An active frame would thus blame the flood or the number affected on construction of roads or that people do not have anywhere to go despite warnings, whereas a non-active frame would blame the heavy rains for the floods or that a bridge collapsed. Third, is the column where other dominating issues like water borne diseases or other effects of the flood or rain are categorized, which also includes those articles that mainly describe the course of events. Lastly, the analysis is expected to find a discussion in the news reports about responsibility in relation to the floods and its effects. These are categorized as government, citizens, international community or none. Previous research has shown that the government is mainly named responsible for the floods and its effects, such as Ngcamu & Binza (2015) and Barnes et. al (2008). If citizens are named responsible is measured if they are asked to move from a risk-prone area or do other action is order to be prepared or mitigate for a disaster (Barnes et. al 2008). A study’s validity depends on whether the study actually empirically measures what it aims to measure theoretically (Esaiasson et. al 2017: 58). This study aims to measure debates on climate change, development issues and responsibilities in media coverage of floods in Kenya in 2013, 2015 and 2017. It is done by identifying what is communicated through the news reports about the floods to the citizens. As mentioned, the media is an important source of information on climate change in Africa (BBC 2009: 2). Informing about the erratic weather and the floods in relation to climate change or issues of vulnerability in the news reports, may broaden the knowledge about what needs to be done in order to withstand these hazards, or else, these disastrous events will continue. As Tagbo argued, regularly and accurately broadcasted communication about climate change is the first step towards developing coping mechanisms in Africa (2010: 6).

17 The analytical questions are chosen in order to address the key concepts in the theoretical background. By asking if the articles contain any links between the floods and climate change and describes it as the main cause, it is possible to measure the extent in which climate change is talked about in relation to the floods or rainfall and will thus answer if there is an equivalent to Bohensky & Leitch’s (2013) results in the context of a developing country. Moreover, by asking what other development issues that is named the cause of the floods, it reflects to what extent the root causes of the disaster are addressed, that Kelman (2014) argues needs to be addressed simultaneously with climate change in order to decrease vulnerability to hazards. By actively framing the problem as being caused by an issue of infrastructure or livelihood decisions, they are also representing what needs to be done in order to decrease vulnerability to future hazards. Lastly, the question about who is named responsible for the flood and its effects will identify the public opinion about who they think is to blame, which in previous research often is laid on the government. Dissatisfaction with the government expressed in the media can lead to political consequences, which Albrecht showed in her study (2017), or pressure the government to make communities less vulnerable and correct flaws while rebuilding (Bohensky & Leitch 2013: 476). The reliability of a study has to do with how it could be repeated by someone else and receive the same result, using the same method by either the same researcher at different times, or by other researchers at approximately the same time (Teorell & Svensson 2007: 59). To lower the risk of low reliability for this study, the analytical framework is developed in a clear and literal way that leaves less room for differentiated interpretation of the material. In order to not risk excluding relevant material, the choice of keywords is generic in order to find a wide range of material. However, the imminent risk of interpreting the material differently at a different time is still present. An attempt to reduce this risk has been done by what Esaiasson et. al (2017: 208-209) call a reliability test, which means that the researcher recodes a smaller part of the material after some time has passed to check if the coding of the material matches both times. The more matches, the higher reliability. This kind of test was done a few weeks apart from starting the process of analysis, since the test had to be done during the writing process of this study. Seven articles from each year were re-coded in the categories Links to climate change, Climate change main cause, Development issue main cause with the subcategories infrastructural problem and livelihood decisions, and lastly Responsibility. The result is attached to the Appendix with the parts that did not match marked in bold.

18 5. Results and analysis

5.1 Is the author making any links between the floods/rainfall and its effects to climate change or global warming? If so, is climate change described to be the main cause of the flood/rainfall and its effects and to what extent is it actively framed as a climate change issue?

The answer to the first analytical question is that out of the 125 articles included in the matrix, only three made links between the floods or rainfall and its effects to climate change and actively framed them as climate change issues (Ecaat 2013, Daily Nation 2013a, Karumbu 2015). Thus, the framework used by Bohensky & Leitch (2013) on the media coverage of floods in an industrialized country had only limited applicability in the context of a developing country. According to the results of this study using this approach, the media coverage on the floods in Kenya coincides with what Tagbo (2010) showed in her study, that environmental issues are still strongly influenced by other development issues that many African countries are dealing with (2010: 35) that manifests when a hazard occur.

This could have resulted in a conclusion that Kenyans are unaware of the effects of climate change, because this approach overlooked the climate change discourse that exists in relation to natural hazards, but not necessarily in relation to a particular event. These discourses are presented further in section 5.5.

Moreover, one article described a citizen’s concerns of the weather becoming more unpredictable than before (Daily Nation 2013e): “We want the rains to come so that people can plant crops and we have food as a country. But they have become so unpredictable and changed pattern. In the past, they would mostly rain at night, but now we cannot tell. They start anytime,” said Mr Wamae.”. This coincides with the study by BBC World Service Trust Research (2009) that found that among the people they interviewed, including people from Kenya, that people are deeply concerned about the changing environment and understand how vulnerable they are in shifts of the life support that sustains them (p. 5).

5.2 Are development issues described as a cause of the floods and its effects, and to what extent is it actively framed as a development issue?

The analysis found that the effects of the floods are mainly described to be caused by infrastructure that cannot withstand the heavy rain; 97 articles of 125 in total. If it concerns an

19 infrastructural problem is evident even though the article is not framed as such, for example (Wokabi 2013): “The heavy downfall has washed away some roads while rendering others impassable thereby cutting the smooth supply of the farm produce to retail markets.”. Articles with no active framing accounted for approximately 56 percent of the articles concerning infrastructural problems. Those articles that framed or partly framed the flood as caused by infrastructural problems accounted for approximately 39 percent of the articles concerning infrastructure. Articles that were framed as infrastructural problems could for example be expressed in the following way (Daily Nation 2013b): “Besides the human toll, there is also the impending tragedy of property destroyed by or floodwaters on account of poor workmanship on the part of owners or public institutions like those responsible for ensuring that towns have working drainage systems and roads are in generally motorable conditions.” The number of active frames did not increase during the years. In 2013, the number of articles with active frames were 39 per cent in comparison to 36 percent that did not have any active frames. In 2015 and 2017, the number of articles that did not have an active frame outnumbered those articles that did, 45 percent in comparison to 36 percent, and 51,7 percent in comparison to 37 percent respectively. This may imply that the role that the Kenyan journalists have when reporting on a particular flood is to mainly serve as objective observers and describe the course of events (Miles & Morse 2006: 366). However, out of the articles that actively framed infrastructure problems as causing the floods, approximately 30 percent of them consisted of opinions of citizens that were interviewed (Githinji 2015, Waweru 2016), letters published in the newspaper (Obuto & Onyino 2013, Wangari 2013), and the rest from the section called “The Cutting Edge” in Daily Nation that can be found in a section in the newspaper called Blogs & Opinion1. Hence, as there seem to be many articles that are being framed as infrastructural problems by citizens, or at least not mainly journalists, may imply an awareness among the population of what can be done to decrease vulnerability towards hazards. This again correlates with that African citizens do understand how vulnerable they are to changes in the climate (BBC 2009: 2, 5). In addition, it also serves as another counter-perspective to Shisanya & Khayesi’s (2007: 281) study that found that climate change is not perceived to be a significant problem in relation to other threats to socioeconomic security in Nairobi. An awareness among the population about these issues may have developed since their study was made.

1 For more examples, see Appendix

20 Among the articles that emphasized livelihood decisions, the problem could be that people live in risk-prone areas (Bii, Karanja, Nyassy, Odongo & Ponciano 2013, Kibui, Otieno, Kipsanga, Munene 2013, Ecaat 2013, Jamenya 2015, The Star 2017e), or lack of choices to go when they were told to (Odongo et. al 2013), and articles that were about the Dadaab refugee camp that were in risk of (Adow 2015, Migori 2015, Mwangi 2015). That most of these articles were about people living in risk-prone areas correlates to what Kelman (2014) highlights in his article, a fundamental problem and a root cause of vulnerability (p. 121) that Kenya is yet to overcome in order to decrease the number affected by floods or rainfall and its effects. However, the results of this study show that the news reports in Kenya do not coincide with Kelman’s (2014) theory concerning small island developing states, since the news reports about the floods in Kenya mainly address development issues instead of climate change. Thus, most of the articles report on issues related to development but do not actively frame them as such. However, even among articles that did not have an active development frame had a tendency to name the government as responsible to act, which is further elaborated on in section 5.4. This may imply that the common perception among the Kenyan people is that damage and losses from floods are consequences of human action (ActionAid 2006: 2) and that Kenya is capable to act upon decreasing vulnerability to natural hazards, even though it may not be explicitly expressed in the articles. Although, in support for this argument, in one chronicle, Musumba (2013) described the situation in this way: “So far, 36 Kenyans have been killed by floods. Sadly, any person or institution speaking after the deaths of these Kenyans with regard to what could have been prevented is simply whistling in the . (…) How do some things in our country go so horribly wrong all the time? How many droughts and famines and deaths from preventable causes must we tolerate?”

5.3 What other issues are dominating the discourse of the articles?

Most of the other issues that dominate the discourse of the articles are mainly the risk or the spread of waterborne diseases that comes with the rains and floods (Daily Nation 2013c, Nyassy & Shilitsa 2013, Bii et al. 2013, Oketch et al. 2013, Nyassy 2013, Adow 2015, Migor 2015, Damary 2015, Astariko 2015, The Star 2017a, The Star 2017b, The Star 2017c, Mkanyika 2017, The Star 2017d). Further is the food shortage and economic losses that are consequences of flooding of crops and harvest (Daily Nation 2013d, Sayagie et al. 2013, Daily Nation 2013e, Daily Nation 2013f, Okewo 2013, Wafula 2013, Karanja 2015,

21 Marindany 2015, Alal 2015, The Star 2017e, The Star 2017f, The Star 2017g) along with food prices increasing when for example floods disrupt transport and thereby create economic losses (Wokabi 2013, Otuki 2015). Food shortage and economic losses is also a great problem when it comes to drought, and this may be the reason why agriculture is a great part of the climate change discussions in Kenya overall (Wafula 2013, Gibendi 2013, Kaino 2016, Odunga 2013, Wakabi 2013). When rains pound cities, businesses like open air markets are also disrupted, spoiling business for both traders and buyers as the markets is where many people get their daily needs (Daily Nation 2013g). Schools are also disrupted due to damaged buildings or that the classrooms are used for shelter for displaced people (Odhiambo & Nyassy 2013, Ochieng 2016).

This result implies that many important issues were not adequately measured due to that the operationalization was not outlined in a way that included these issues as development issues, even though they should be treated as such. Therefore, the study’s results encourage future research to outline a different approach. For example, one could identify the various development challenges that Kenya has, and from that point of view, explore how media coverage of these development issues relate to vulnerability to natural hazards and climate change, and what they discuss needs to be done to solve these challenges. For example, an issue that is very relevant in the context of natural hazards is economic losses when floods destroy crops and other livelihoods. An author suggested that to solve this, rural communities must resettle from low-lying flood prone areas (Ecaat 2013). Thus, this approach would enable to explore discourses on how to act towards decreasing vulnerability to hazards and climate change overall. Perhaps a research question could have asked what the direct impacts of the floods have on Kenya’s economic development, and what action is said to be needed to overcome these challenges, which would have included infrastructural problems such as roads, bridges, drainage and housing that withstand the heavy rains. As many livelihoods are disrupted because of the rains and floods due to diseases, economic losses and food shortages, these issues should indeed be interpreted as development issues. This approach may also have included the issues that are presented in section 5.5, that further explore the many debates that exists in Kenya about climate change, natural hazards and mitigation efforts, that could have been overlooked by the approach used for this study.

5.4 Who is named responsible for action? The government, the international community or the citizens?

22 The absolute majority of the articles named the national government or the county governments as responsible to deal with the floods and its effects. The international community was only mentioned once (Mkanyika 2017). The citizens were described as responsible to relocate from risk-prone areas to mitigate the number affected (Daily Nation 2013c, Sayagie et al. 2013, Bii et. al 20132). One article highlighted that citizens will be charged with attempted suicide if they refused to move to safer grounds (Owiti 2015). One time, the citizens were blamed responsible when it came to not responding to warning signs (Daily Nation 2013c), and another when they were asked to insure their crops against climatic changes to avoid economic losses (Daily Nation 2013a).

What the government was named responsible for was, among other things, to bring relief supplies and resettle displaced people (Nyassy & Shilitsa 2013), to repair infrastructure that were destroyed with capacity to withstand all weather standards (Suter 2013), to come up with long-term solutions to decrease the number of disasters that relate to the rainfalls that is a perennial problem (Oketch et al. 2013), and to be better prepared to respond to disasters (Ecaat 2013).3 There were many citizens that were dissatisfied with the government for not providing the services that the citizens pay taxes for, or that the government had received funds to deal with the effects of El Niño in 2015 that many asked for (Odeny 2015, Ndanyi 2015, Mbuva 20154).

That the government received the blame for the disasters correlates with previous studies, that the media reflects upon the government’s role in disaster management in countries like the UK (Albrecht 2017), Australia (Bohensky & Leitch 2013) and the US (Barnes et al 2008), and the Kwazulu-Natal province in the study by Ngcamu & Binza (2015), that name the government as mainly responsible for the aftermath of disasters. That the international community was only mentioned once may be because of the absence of a climate change discourse, in comparison to what Billet (2010) found in the representation of the climate change issue in Indian mass media. This matter is encouraged to be further explored in the articles that did not relate to these particular floods, where discussions about climate change are more prominent. That the majority of articles had to do with problems such as inadequate infrastructure may be the reason, as already mentioned, that it is widely recognized that Kenya is vulnerable to hazards, as well as it is understood that action can be

2 For more examples, see Appendix 3 For more examples, see Appendix 4 For more examples, see Appendix

23 taken to decrease these vulnerabilities. By pressuring the government to take action, capacities can be increased in order to withstand future hazards.

5.5 The articles that did not relate to a particular event

The articles that were not included in the matrix were the ones within the date limit and included the words flood*, rain*, downpour* or El Niño but did not relate to the floods selected for the analysis. However, it was noticed during the process of analysis that these articles reflect a discourse on climate change and natural hazards that necessarily did not have to do with ongoing extreme events. Many discussions were about Africa as a whole (Yeda 2013, Kaberuka 2013, Wandati 2016), for example how the Paris deal fell short on Africa’s needs (Angena & Wang’ombe 2015, Dyer 2015, Wandati 2015b) and a reflection about Trump’s choice to withdraw from the Paris Agreement (Sebunya 2017)5. Several articles focused on Kenya alone.6 In sum, Kenya’s future plans to tackle climate change are said to invest heavily in strengthening the community resilience to the changing climate, massive forest protection and afforestation programs, water harvesting and conservation, climate- smart agriculture7, establishment of renewable energy sources and energy conservation (The Star 2017h). Moreover, several of these articles were about mitigation efforts towards the extreme weather that has been done or are planned in Kenya, for example reinforcing river dykes to stop yearly flooding (Wafula & Ramenya 2013, Wanga 2013, Ochieng 2015)8. Another counter-argument to Shilanya & Khayesi’s (2007) study is represented by a chronicle that linked unemployment to natural disasters. Mang’eni (2013) argues that “Youth poverty is multifaceted and complex; it is as a result of socio-historical reasons, natural disasters such as floods and droughts, compounded by environmental degradation.”.

5 They could discuss climate change in general (Giugale 2013) or East Africa particularly (Abdallah 2016) 6 For example, how the drought affects the Kenyan economy (Ugangu & Kinyangi 2013), expectations that Kenya will be the first African country to have a climate change law in place (Njagi 2015), how water- harvesting technologies could prevent wasting rainwater (The Star 2017i), and how climate change will worsen the El Niño phenomenon (Mutambo 2016) 7 Climate change greatly influence the ability of households to improve incomes and meet their food needs (Wafula 2013, Gibendi 2013, Kaino 2016). There are plans to instruct farmers in how to adapt to warmer temperatures (Odunga 2013, Wakabi 2016) 8 Other examples are repairing roads (Situma 2013, Otuki 2013a, Wekesa 2016), building of underground cables (Irungu 2013, Otuki 2013b), mapping out risk-prone areas (Daily Nation 2013h), construction of dams to both control flooding and supply water (Kang’aru 2013, Mureithi 2013), mitigation against cholera spreading (Otieno 2015), the set-up of the Situation Room in Baringo (Kwama 2016), fix open sewers to decrease floods in the cities (The Star 2017j), and construction of resilient water systems to protect people from climate change (Wandati 2015a)

24 These examples show that Kenya have managed or plan to reduce the damage caused by the rains. These could be interpreted as efforts to “build back better”, the rebuilding of a community after a disaster without re-creating or exacerbating vulnerabilities (Kelman 2014: 125), and that lessons for the future are considered (Bohensky & Leitch 2013). The prominence of climate change discussions within this date-limit may be due to that these events evoke these discussions, which was noticed in the previous studies on media coverage on climate change in developing countries as well (Midttun et al. 2014, Shanahan 2009, Billet 2010). Again, that these debates are held outside the context of a particular flood, could emphasize the point that news reports mainly focus on the course of events (Miles & Morse 2006: 366). However, this large number of articles with a broad range of subjects indicate that Kenya may have come a long way in the debates about natural hazards and climate change. This shows that the news media make efforts in providing information about these issues to the citizens, which was according to Tagbo (2010) crucial in order to gain capacity to respond to impacts of climate change (p. 17).

In sum, this analysis of the discourses on climate change, development issues and responsibilities in media coverage of the floods in Kenya in 2013, 2015 and 2017 shows that the framework used by Bohensky & Leitch’s (2013) had limited applicability to the context of Kenya. A different approach would have further elaborated on debates on climate change, natural hazards and mitigation efforts that might not relate these discussions to a particular flood, as well as development issues that was overlooked in the analysis. The media coverage on the floods mainly address infrastructural problems. Although many articles actively framed the floods as an infrastructural issue, most of the articles did not contain such frames. However, as most of the articles named the government responsible to act to address these issues, even articles without an active frame, indicate that the common perception among Kenyans is that they are vulnerable to hazards, but that actions can be taken to decrease such vulnerabilities.

6. Conclusion

This study aimed to investigate to what extent the Kenyan news media relate extreme events in the form of floods and heavy rainfall to climate change, if development issues are described as causes of the floods and its effects, and who is named responsible to act. The floods that occurred in 2013, 2015 and 2017 were selected for the analysis, and the data

25 material consisted of news reports on the floods from the English-speaking news media Daily Nation, The East African, The Star and Business Daily. Previous research has shown that the media discourse on climate change in industrialized countries are very different from the media discourse in developing countries. This is due to that environmental issues are strongly influenced by other development issues that countries in for example Africa are dealing with. However, as Kenya is a country that is already visibly experiencing climate change in the form of severe droughts and floods, it was expected to find a similar discussion about the floods in relation to climate change which has been found in the context of industrialized countries. Therefore, a framework used by Bohensky & Leitch (2013) to investigate the media framings of floods with perceived links between the event and climate change in Australia was adopted for this study. The results indicate that this framework had only limited applicability in the context of Kenyan news media on the floods. Only a few links were made between the floods and climate change in the news reports. The main focus of the news coverage on the selected floods overlooked various discourses on climate change, natural hazards and mitigation efforts that took place in parallel with the ongoing floods and heavy rainfall but did not necessarily relate these issues to a particular event. Moreover, it was investigated whether development issues, namely infrastructural problems and livelihood decisions, were described as causes for the floods. The results show that the majority of the articles mention these development issues but did not identify them as core reasons for the negative impacts of the floods. This may imply that the attention was given to other problems that relate to vulnerability to hazards that could not be adequately measured with the operationalization used for this study. Lastly, the government was named responsible to act on the floods and its impacts in most of the articles. Studies made with qualitative methods with interpretive elements should always be treated with reservation, since another researcher may interpret the material differently and make slightly different conclusions. In this study, this could be the case in terms of what development frames are found and what other issues are presented in the articles, while the results that are more certain to also be found by others are the extent of the climate change discourse and that the government is mainly named responsible to act. The first main finding drawn from this study is that to adequately measure media discourses on climate change and vulnerabilities to hazards in the context of a developing country, future research is encouraged to firstly identify relevant development issues that a country has, and from this perspective, explore how media coverage of these development

26 issues relate to vulnerability to natural hazards and climate change, and what they discuss needs to be done to solve these challenges. Thus, this approach would enable to explore discourses on how to act towards decreasing vulnerability to hazards and climate change overall. The second main finding is that although many articles actively did not actively frame the floods as infrastructural issues, most of the articles named the government responsible to act to address the floods or rainfall and its effects. This may indicate that the common perception among Kenyans is that they are vulnerable to hazards, and that damage from floods are consequences of human action. However, that Kenya is also capable of acting towards decreasing these vulnerabilities and gain resilience to withstand these hazards. To conclude, studies that aim to explore debates on climate change in the context of a developing country should show consideration to other development issues that causes vulnerability to hazards. Investigating the climate change issue in those contexts in the same way as have been done in an industrialized country may overshadow discussions that we need to take part of in order to understand its true impacts.

27 References

ActionAid International. 2006. Climate change, urban flooding and the rights of the urban poor in Africa. Key findings from six African cities. Report. London: International Emergencies and Conflict Team. https://www.actionaid.org.uk/sites/default/files/doc_lib/urban_flooding_africa_report.pdf (Accessed 2018-04-04).

Albrecht, Frederike. 2017. The Social and Political Impact of Natural Disasters. Investigating Attitudes and Media Coverage in the Wake of Disasters, Uppsala: Acta Universitatis Upsaliensis.

Awour, Cynthia Brenda; Orindi, Victor Ayo; Adwera, Andrew Ochieng. 2008. Climate change and coastal cities: the case of Mombasa, Kenya. Environment and Urbanization. 20 (1): 231-242.

Azmi, Nor, Nor Azikin Mohd Omar, Safawati Basirah Mohd Zaid, Zanirah Wahab, & Akmah Yusof. 2015. "Media Portrayal on Global Climate Change: an Analysis of Malaysian Mainstream Newspapers." Studies in Media and Communication [Online]. 3 (1): 73-78. (Accessed: 2018-05-02).

Barnes, D Michael; Hanson L, Carl; Novilla M.B. Len, Meacham, T Aron; McIntyre, Emily; Erickson, C Brittany. 2008. Analysis of Media Agenda Setting During and After Hurricane Katrina: Implications for Emergency Preparedness, Disaster Response, and Disaster Policy. American Journal of Public Health 98 (4): 604-610.

BBC (2006). Kenyans drown after flash floods. BBC News. November 13. http://news.bbc.co.uk/2/hi/africa/6144052.stm (Accessed 2018-04-20).

BBC (2009). Policy briefing # 3. Least responsible, most affected, least informed: Public understanding of climate change in Africa. BBC World Service Trust: UK. http://downloads.bbc.co.uk/worldservice/pdf/wstrust/PB_climatechange_web.pdf (Accessed 2018-05-03).

Billet, Simon. 2010. Dividing climate change: Global warming in the Indian mass media. Climatic Change 99 (1): 1-16.

Bohensky, Erin L. Leitch, Anne M. 2013. Framing the flood: a media analysis of themes of resilience in the 2011 Brisbane flood. Regional Environmental Change 14 (2): 475-488.

Boykoff, Maxwell T & Roberts, Timmons J. 2007. Media Coverage of Climate Change: Current Trends, Strengths, Weaknesses. Human Development Report 2007/2008: United Nations Development Programme (UNDP).

28 http://hdr.undp.org/sites/default/files/boykoff_maxwell_and_roberts_j._timmons.pdf (Accessed 2018-05-03).

D. Guha-Sapir, R. Below, Ph. Hoyois - EM-DAT: The CRED/OFDA International Disaster Database – www.emdat.be – Université Catholique de Louvain – Brussels – Belgium. (Accessed 2018-05-03).

Drisco, James W & Maschi, Tina (2015). Qualitative Content Analysis. Included in: Drisco, James W & Maschi, Tina (eds.) Content Analysis. Oxford University Press. p. 81-120.

Entman, Robert M. 1989. How the Media Affect What People Think: An Information Processing Approach. The Journal of Politics. 51 (2): 347-370.

Entman Robert M. 1993. Framing: Toward Clarification of a Fractured Paradigm. Journal of Communication 43(4): 51-58. https://www.unc.edu/~fbaum/teaching/articles/J- Communication-1993-Entman.pdf (Accessed 2018-05-03).

Esaiasson, Peter; Gilljam, Mikael; Oscarsson, Henrik,; Towns, Ann & Wängnerud, Lena. 2017. Metodpraktikan - konsten att studera samhälle, individ och marknad. 5 uppl., Stockholm: Wolters Kluwer.

GoK 2010. National Climate Change Response Strategy. Government of Kenya. https://cdkn.org/wp-content/uploads/2012/04/National-Climate-Change-Response- Strategy_April-2010.pdf (Accessed 2018-04-04).

IFRC (2010). Emergency Appeal. Kenya: Floods. International Federation of Red Cross and Red Crescent Societies. June 11. http://www.ifrc.org/docs/appeals/10/MDRKE012EA.pdf (Accessed 2018-14-20).

Houston, J Brian; Pfefferbaum, Betty. & Rosenholtz & E, Cathy 2012. Disaster News: Framing and Frame Changing in Coverage of Major U.S. Natural Disasters, 2000-2010. Journalism and Mass Communication Quarterly 89 (4): 606-623.

IFRC (2013). Emergency Appeal. Kenya: Floods 2013. International Federation of Red Cross and Red Crescent Societies. April 22. http://www.ifrc.org/docs/Appeals/13/MDRKE025EA.pdf (Accessed 2018-04-20).

IFRC (2016). Emergency Plan of Action (EPoA) Kenya: Floods. International Federation of Red Cross and Red Crescent Societies. January 4. https://reliefweb.int/sites/reliefweb.int/files/resources/MDRKE036.pdf (Accessed 2018- 04- 20).

29 Kamau, Julius W. Mwaura, Francis. 2013. Climate change adaptation and EIA studies in Kenya. International Journal of Climate Change Strategies and Management 5 (2): 152-165.

Karimi, Martin (2015). WFP Assists Families Displaced By El Niño Floods. World Food Programme. December 22. https://www.wfp.org/stories/kenya-wfp-assists-families-displaced-el-ni%C3%B1o-floods (Accessed 2018-04-20).

Kelman, Ilan, 2014. No change from climate change: vulnerability and small island developing states. The Geographical Journal 180(2): 120-129.

Kracauer, Siegfried. (1952). The Challenge of Qualitative Content Analysis. The Public Opinion Quarterly 16 (4): 631-642.

Midttun, Atle; Coulter, Paddy; Gadzepko, Audrey and Wang, Jin. 2015. Comparing media framings of climate change in developed, rapid growth and developing countries: Findings from Norway, China and Ghana. Energy & Environment 26 (8): 1271-1292.

Miles, Brian & Morse, Stephanie. 2006. The role of the news media and natural disaster risk and recovery. Ecological Economics 63 (2): 365-373.

Ngcamu, Bethuel Sibongiseni; Binza, Shakespeare Mzikayise. 2015. Media framing of natural disasters in Kwazulu-Natal Province: Impact of Contigency plans. Journal of Governance and Regulation 4 (4): 110-117.

Nyakundi, Hellen; Mogere, Stephen; Mwanzo, Isaac; Yitambe, Andre. 2010. Community perceptions and response to flood risks in Nyando District, Western Kenya. Journal of Disaster Risk Studies 3 (1): 346-366.

Obura, Fredrick (2017) Floods to affect parts of Rift Valley, Western, Central, Nyanza regions. The Standard. October 11th. https://www.standardmedia.co.ke/business/article/2001257069/meteorological-department- warns-of-flash-floods (Accessed 2018-05-01).

Osindo, Joel Michael. 2014. The framing of climate change issues in Kenyan print media: a comparative analysis of the Daily Nation and the People Daily newspapers. Masters degree. Nairobi: University of Nairobi/School of journalism and mass communication. http://hdl.handle.net/11295/77990 (Accessed 2018-05-02). Shisanya, C. A. & Khayesi, M. 2007. How is climate change perceived in relation to other socioeconomic and environmental threats in Nairobi, Kenya? Climatic Change 85 (3): 271- 284.

30 ProQuest (2018). Factiva: About. https://proquest.libguides.com/factiva (Accessed 2018-05- 03).

Sandelowski, Margarete. 2000. Whatever Happened to Qualitative Description? Research in Nursing & Health 23 (4): 334-340.

Scheufefel, Dietram. A & Tewksbury, David. 2007. Framing, Agenda Setting, and Priming: The Evolution of Three Media Effects Models. Journal of Communication 57 (1): 9-20.

Tagbo, Evelyn. 2010. Media coverage of climate change in Africa. A case study of Nigeria and South Africa. Reuters Institute Fellowship Paper. Oxford: University of Oxford/Reuters Institute for the study of journalism. https://reutersinstitute.politics.ox.ac.uk/sites/default/files/research/files/Media%2520Coverag e%2520of%2520Climate%2520Change%2520in%2520Africa%2520A%2520Case%2520Stu dy%2520of%2520Nigeria%2520and%2520South%2520Africa.pdf (Accessed 2018-05-03).

Teorell, Jan & Svensson, Torsten. 2007. Att fråga och att svara. Malmö: Liber.

Trumbo, Craig. 1996. Constructing climate change: claims and frames in US news coverage of an environmental issue. Public Understanding of Science 5 (3): 269-283.

TW (2017). Heavy rain, deadly floods affecting Kenya. The Watchers. November 10th. https://watchers.news/2017/11/10/heavy-rain-deadly-floods-affecting-kenya/ (Accessed 2018-05-01).

UNFCCC. 2017. Article/12 Oct, 2017. Developing Countries Need Urgent Support to Adapt to Climate Change. United Nations Climate Change. https://unfccc.int/news/developing- countries-need-urgent-support-to-adapt-to-climate-change (Accessed 2018-05-11).

Wodak, Ruth (2001). The discourse-historical approach. Included in: Wodak, RP, & Meyer, M (eds) Methods of Critical Discourse Analysis. London. Sage Publications. p. 63-93.

Zaheer, Lubna. 2016. Natural Catastrophes and Role of Pakistani Mass Media. Journal of Political Studies 23 (1): 207-232.

Zamith, Rodrigo; Pinto, Juliet; Villar, Maria Elena 2012. Constructing Climate Change in the Americas: An Analysis of News Coverage in U.S. and South American Newspapers. Science Communication 35(3): 334-357.

Newspaper articles

31 Abdallah, Halima (2016). Developing cities urged to embrace climate change resilience methods. January 9. The East African.

Adow, Mohamed (2015). Cholera kills seven at Dadaab, 540 taken ill. December 18. The Star. Factiva.

Alal, Maurice (2015). K’Oyoo wants floods fund released. December 7. The Star. Factiva.

Andae, Gerald (2015). Climate change storm to cut tea production by 9pc in five years. December 10. Business Daily. Factiva.

Astariko, Stephen (2015). 20,000 displaced by floods in Tana River. December 01. The Star. Factiva.

Bii, Barnabas; Karanja, Samuel; Nyassy, Daniel; Odongo, Ponciano (2013). Diseases threaten flood victims. April 8. Daily Nation. Factiva.

Damary, Rita (2015). Three cholera cases in Nakuru, water hawking lorries blamed. December 1. The Star. Factiva.

Daily Nation (2013a). Farmers told to insure their animals, crops. April 17. Daily Nation. Factiva.

Daily Nation (2013b). The Cutting Edge-Sunday May 5 2013. May 4. Daily Nation. Factiva.

Daily Nation (2013c). Two drown as floods death toll rises. April 29. Daily Nation. Factiva.

Daily Nation (2013d). The Cutting Edge-Wednesday April 24. April 23. Daily Nation. Factiva.

Daily Nation (2013e). Traders counting losses as heavy rains set in. April 15. Daily Nation. Factiva.

Daily Nation (2013f). Heavy rains and lack of cheap inputs disrupt planting in Narok. April 2. Daily Nation. Factiva.

Daily Nation (2013g). Traders counting losses as heavy rains set in. April 15. Daily Nation. Factiva.

Daily Nation (2013h). Counties told to map out disaster areas. June 6. Daily Nation. Factiva.

Dyer, Gwynne (2015). Sad, Paris Deal Won't Stop Global Warming. December 15. The Star. Factiva.

32 Ecaat, Justin. 2013. Why weather is key in economic growth agenda. June 2. Business Daily. Factiva.

Gibendi, Ramenya (2013). Farmers abandon rice growing due to poor returns. May 13. Business Daily. Factiva.

Giugale, Marcelo (2013). GIUGALE: Dilemma for countries in quest for ‘green growth’. June 14. Daily Nation. Factiva.

Irungu, Geoffrey (2013). Kenya Power to spend Sh100bn on network upgrade. April 17. Business Daily. Factiva.

Jamenya, Joseph (2015). Storm causes mudslide in Hamisi, scares 10 families. December 14. The Star. Factiva.

Kaberuka, Daniel (2013). Water everywhere, and Africa still goes thirsty. March 23. Daily Nation. Factiva.

Kaino, Bethwel (2016). Policy To Cushion Farmers Is Timely. January 12. The Star. Factiva.

Kang’aru, Wachira (2013). Coming soon: Kenya’s largest fresh water dam. June 3. Daily Nation. Factiva.

Karanja, Samuel (2015). At least 100 killed as El Niño tails off. December 23. Business Daily. Factiva.

Karumbu, Immaculate (2015). Kenya set to benefit from Sh8.5bn El Niño fund. December 4. Business Daily. Factiva.

Kwama, Kenneth (2016). Baringo 'Situation Room' Saves Lives. January 20. The Star. Factiva.

Mang’eni, Ben T (2013). MANG'ENI: Empowering youth at the county level. April 30. Daily Nation. Factiva.

Marindany, Kurgat (2015). Kaijado roads cut off by heavy floods. December 15. The Stra. Factiva.

Mbuva, Andrew (2015). Four saved from Machakos flood in River Athi Islands. December 18. The Star. Factiva.

33 Migori, Katy (2015). Cholera strikes 500 in Kenyan refugee camp amid heavy rains. December 17. The East African. Factiva.

Mkanyika, Lucy (2017). Floods wreak havoc at the Kenya coast. May 8. The East African. Factiva.

Moraa, Jackline (2013). Board seeks Sh8bn more to repair roads. May 17. Daily Nation. Factiva.

Moora, Jackline & Mwajefa, Mwakera 2013. Hundreds left out in the cold by landslides. April 22. Daily Nation. Factiva.

Mureithi, Francis (2013). Sh4.8 billion dam to end the water woes of Baringo residents. May 13. Daily Nation. Factiva.

Musumba, Linda (2013). MUSUMBA: Why we are being buffeted at every turn. May 08. Daily Nation. Factiva.

Mutambo, Aggrey (2016). BYANYIMA: Govts, global partners must join hands to avert humanitarian crisis. February 24. The East African. Factiva.

Mwangi, William (2015). Seven dead, 541 admitted in Dadaab cholera outbreak. December 17. The Star. Factiva.

Ndanyi, Mathews (2015). Floods displace more than 3,000 in Baringo as rivers burst banks. December 18. The Star. Factiva.

Njagi, David (2015). Kenya may get climate change law before christmas. December 18. The Star. Factiva.

Njoroge, Kiarie (2015). City Hall blames El Niño for garbage pile up. December 29. Business Daily. Factiva.

Nyassy, Daniel (2013). 8,000 evacuated from floods. May 1. Daily Nation. Factiva.

Nyassy, Daniel & Shilitsa, John (2013). Seven contract water-borne diseases. April 28. Daily Nation. Factiva.

Obuto, Alex & Onyino, Dennis (2013). Identify areas prone to floods to avoid disasters in the future. Daily Nation. Factiva.

34

Ochieng, Gilbert (2015). No more floods, says Namwamba. December 4. The Star. Factiva.

Ochieng, Gilbert (2016). 200 pupils stay at home due to rain damage. January 6. The Star. Factiva.

Odeny, Manuel (2015). Israel donates food, water tanks to Migori El Niño victims. December 28. The Star. Factiva.

Odhiambo, Moses & Nyassy, Daniel (2013). Floods crisis may delay opening of new school term. May 5. Daily Nation. Factiva.

Odongo, Ponciano; Otieno, Elisha; Ochieng, Lillian; Suter, Philemon; Muyanga, Philip & Ngunjiri, James (2013). Floods kill six as heavy rains season sets in. April 1. Daily Nation. Factiva.

Odunga, Dennis (2013). Scientists want farmers involved in climate change plans. June 20. Business Daily. Factiva.

Oketch, Angela; Okewo, Everline; Wafula, Linet; Shilitsa, John & Karanja, Samuel (2013). Secondary school shuts down after flooded classes pose disaster threat. May 12. Daily Nation.

Okewo, Everline (2013). Farmers appeal for inputs to replant fields after floods. April 2. Business Daily. Factiva.

Otieno, Julius (2015). 87 suffer cholera in the Eastlands. December 17. The Star. Factiva.

Otuki, Neville (2013a). Roads authority seeks Sh10 bn for Nairobi repairs. June 5. Business Daily. Factiva.

Otuki, Neville (2013b). Kenya Power replaces cables to lower costs. May 13. Business Daily. Factiva.

Owiti, George (2015). Don't await airlift from rains-officer. December 17. The Star. Factiva.

Sayagie, George; Obirie, Moraa; Kimani, Joyce; Okeyo, Verah & Moraa, Jackline (2013). Two more die in landslides. April 23. Daily Nation. Factiva.

Sebunya, Kaddu (2017). Trump's burning foot a wakeup call. June 9. The East African. Factiva.

35

Situma, Evelyn (2013). Rongai traffic jams may worsen as road repairs start. April 26. Business Daily. Factiva.

Suter, Philemon (2013). Poor roads delay planting in Marakwet. April 3. Business Daily. Factiva.

The Star (2017a). Water-borne diseases alert issued, Naivasha latrines now flooded. May 13. The Star. Factiva.

The Star (2017b). 361 families living along River Nzoia displaced and 20 houses submerged. May 13. The Star. Factiva.

The Star (2017c). epidemic hits Mombasa as rains pound county. May 10. The Star. Factiva.

The Star (2017d). Cholera outbreak in Naivasha, one admitted. June 4. The Star. Factiva.

The Star (2017e). Rains give farmers hope as others count losses. May 9. The Star. Factiva.

The Star (2017f). [Video] Shocking footage of Embu mudslide emerges, crops worth Sh3m destroyed. May 24. The Star. Factiva.

The Star (2017g). Rains both a blessing to some, a curse to others. May 18. The Star. Factiva.

The Star (2017h). Interview: We've restored 850 k hectares of degraded forests by involving communities-PS. June 19. The Star. Factiva.

The Star (2017i). Invest in hometanks to harvest rainwater. May 9. The Star. Factiva.

The Star (2017j). Shahbal pledges tough drugs war, capital punishment for barons in manifesto. June 30. The Star. Factiva.

Wafula, Linet (2013). Rice farmers face losses on lack of ready markets. April 3. Daily Nation. Factiva.

Wafula, Linet & Remenya, Gibendi (2013). Now floods menace is under the bridge in Budalang’i. April 22. Daily Nation. Factiva.

Wandati, Zeynab (2015a). Kenya inks $200mn dam construction deal at Paris talks. December 3. The East African. Factiva.

36 Wandati, Zeynab (2015b). Compensation the sticking point in climate change talks. December 7. The East African. Factiva.

Wandati, Zeynab (2016). Humanitarian agencies warn of famine as El Niño wreaks havoc. January 9. The East African. Factiva.

Wanga, Justus (2013). Ambitious plans for lakeside city. May 12. Daily Nation. Factiva.

Wangari, Samuel C (2013). Harvest rainwater and use it for irrigation in dry seasons. April 17. Daily Nation. Factiva.

Wekesa, Chrispinus (2016). City Hall sets aside Sh300m to repair roads. January 7. The Star. Factiva.

Wokabi, Charles (2013). Floods push up food prices. April 17. Daily Nation. Factiva

Yeda, Vivienne (2013). New RCC in Kampala gives region the chance to take charge of climate change. May 4. The East African. Factiva.

37

38