WATER THAT BREAKS WALLS
A case study on the role of community and their networks in the operation of support in Kotturpuram after the 2015 floodsWater that breaks walls:
A case study on the role of
community and their networks in the operation of support in
Kotturpuram after the 2015 flood
‘Times of disaster can bring out the best and worst in people. The good thing that came forward was the increasing tolerance of people towards each other. During the flood walls
were broken. Literally but also figuratively’ Father Baskar
Bachelor Thesis Human Geography University of Amsterdam
Name student Floor Matla Student number 10599010
E-mail address flori.matla@hotmail.com Phone number +31 622956988
Supervisor dhr. J. V. Rothuizen Second reader dhr. dr. D. L. Arnold Date 24/06/2016
Preface and acknowledgements
This thesis was conducted as a part of my graduation of the Bachelor Future Planet Studies combined with a Human Geography major at the University of Amsterdam. The fieldwork for this thesis has been carried out in Chennai, Tamil Nadu from the 16th of April until the 13th of May, 2016. Prior to the fieldwork a twee week course on Coastal Management and Ecosystem Services was provided at the Anna University in Chennai. Together with 20 other students and the guidance of Maarten Bavinck and Jaap Rothuizen the course has been a good preparation of the actual field work. I am thankful for the stories of all lecturers and the excellent service of the Anna University staff.
During my fieldwork I also stayed of the guest house at the Anna University. I would like to thank the staff of guest house for making this stay comfortable. The research could not have been carried out without the help of my translator Kavita Sharma. Although our clocks were not always on the same track she did put in a lot of efforts to make the research a success. I would like to thank K. Subramanian for his work in the organization of our translators and going beyond his duties to make everybody happy. The experience of being in the field by myself has been a very educational but also inspirational one. It would not have been this great without the sincere corporation of all my respondents. I would like to make special thanks to Kanchana, Reeva and Gita who made me feel really welcome in Kotturpuram, my area of research. Back home in Amsterdam the most difficult part of the research took place, which involved the actual writing of the thesis, interpreting all the data and presenting it in an elaborate way. With the help of Jaap Rothuizen I was able to find a way to get my findings and ideas on paper. Also the encouragement of friends and family has been very helpful to motivate me during the writing process.
Index
1. Introduction ...4
2. Theoretical Framework ...6
2.1 Disaster and Support ...6
2.3 Social support networks ...7
2.3 External support ...10
3. Methodology ...11
3.1 Research design ...12
3.2 Research setting and population ...12
3.3 Research methods ...14
3.4 Limitations: validity and reliability ...17
4. Course of events: Floods in Kotturpuram ...20
4.1 Community composition of Kotturpuram ...20
4.2 Phases of crisis ...22
4.2.1 Pre-crisis ...22
4.2.2 Crisis ...23
4.2.3 Post-crisis ...27
5. Support networks in Kotturpuram ...29
5.1 Drivers of internal organization in the community ...29
5.1.1 Personal networks and altruism ...29
5.1.2 Contributions of groups...31
5.1.3 Local religious institutions ...32
5.1.4 Local political parties ...34
5.2 Support by external organizations ...35
5.2.1 Informal support by civil society ...35
5.2.2 Formal support by the government ...36
6. Theoretical analysis ...39
7. Conclusion ...42
Appendix ...42
1. Introduction:
The impact of floods in the city of Chennai and the crucial role of support
Floods are not uncommon in Chennai, Tamil Nadu. At least once a year the city a is confronted by heavy showers as a result of the monsoon season (Mishra, 2016). During the month of November 2015 Chennai even experienced some mild floods. But on the 1st of December 2015 Chennai experienced one of the worst flood events in the past decades (The Indian Express, 2015). According to reports in the Indian Express and the Hindu1, the floods have resulted in the death of about 400 people. The total costs for Chennai as a result of the floods lay somewhere between 2,2 and 3,9 billion dollars (or 14000-25,912 crore2) according to the ESN Economic Bureau (2016) and the Indian Express (2015). How a disaster of this size could occur is being debated. Some point to the role of climate fluctuations or climate change (Narasimhan et al., 2016), whereas others attribute the disaster to water logging (due to removal of waterbodies and increase in built environment) or administrative errors to respond in time (The Indian Express, 2015). Lavanya (2012) states that in order to mitigate the impact of future floods in Chennai structural changes in the infrastructure of the city and an increased awareness about the climate are both crucial.
Chennai has a high and dense population and therefore a lot of residents are exposed to the detrimental impacts of the raging water, especially the urban poor who are living along the riverside and in low-lying areas (Joerin, Shaw, Takeuchi, and Krishnamurthy, 2012). Chennai inhabits a lot of urban poor which increases the vulnerability of the city. However Dynes (2006:23) describes an approach in which the emphasis on human vulnerability would make place for the focus on human capability. Dynes wanted to point out that disasters do not only bring forward a lot of victims but also corporation and support. ‘The power of individuals and their networks
shouldn’t be underestimated’ Clarke (2006) points out. Beggs, Haines and Hurlbert
(1996) state that multiple studies on disasters have shown that one of the factors that allows individuals to recover quickly is the reliance on support. Aldrich (2008)
1
The Hindu (2015). Death toll in floods mounts to 347. Retrieved from: http://www.thehindu.com/news/national/tamil-nadu/death-toll-in-floods-mounts-to-347/article7973293.ece on 19/06/2016.
Janardhanan, A. (2015). Now Chennai struggles to lay its dead to rest. Retrieved from: http://indianexpress.com/article/india/india-news-india/now-chennai-struggles-to-lay-its-dead-to-rest/ on 19/06/2016
The Hindu (2016). Northeast Monsoon claimed 470 lives in Tamil Nadu: Jayalalithaa. Retrieved from:
http://www.thehindubusinessline.com/news/national/northeast-monsoon-claimed-470-lives-in-tamil-nadu-jayalalithaa/article8064661.ece on 19/06/2016
2
highlights that there have been several groups or communities in the aftermath of hurricane Katrina that recovered more quickly than others based on the presence of strong support networks. However support is quite a very broad definition and can come from various sources, each providing it in different ways. In the case of Chennai there had been various reports in the media criticizing the role of the government in the operation of support (The Frontline, 2015). Because formal support was apparently lacking questions about other forms of support started to arise.
Within this thesis the case of the December flood in Chennai is used to understand the operation of relief support within a community. A four week fieldwork has been carried out to talk with actual providers and receivers in order to understand how support comes about. The area of research that has been chosen is Kotturpuram, one of the areas severely impacted by the flood. Because of the size of the research only one community will be researched. In this way no general picture of the situation in Chennai can be made, however the research gives a very detailed picture for the selected community
The research will start by explaining the most important theoretical assumptions regarding the operation of relief support in community context. With this information a theoretical framework will be developed that helps to guide the research. Chapter 3 will talk about methodology, explaining the practicalities of the research which will give insight in how data has been derived and processed. In chapter 4 a setting of the scene is provided. Chapter 5 will discuss the results of the fieldwork. The results will be linked to the theoretical assumptions. Chapter 6 will provide a deeper analysis of the results in which will be discussed what is learned from this study. Lastly, a conclusion will provide a summary of the research.
2. Theoretical framework
The operation of relief support in times of disaster has been under the interest of researchers for a long time. Although not a new topic of research, support studies have gained renewed interest due to increasing concerns about environmentally induced disasters (Elliott, Haney and Sams-Abiodun, 2010). In addition, the transaction of support is a human process which is susceptible to changes over time and therefore requires continuing research. Currently, increased attention is paid to the role of communities and their networks and how these can contribute to reduce the vulnerability of a community (Perrow, 2007). Instead of regarding affected communities as passive victims they should be seen as active resources (Dynes, 2006). The emphasis on communities and their networks contrasts the traditional view of ‘top-down’ intervention, characterized by large institutions trying to
command and control the situation (Elliott, Haney and Sams-Abiodun, 2010). Yet there
is little research that discusses the interaction between the traditional and external relief operations and the internal community operations. Within this chapter a theoretical framework is created that discusses relevant literature related to the operation of support within a community and the role of external support. First of all the relation between disaster and support will be explained.
2.1 Disaster and support
Disaster support is a distinct form of support, in which a lot of people are confronted with a similar fate and a similar need for support (Kaniasty and Norris, 2000). Although the impacts of a disaster are depended on the nature of the disaster a general pattern can be recognized in their course. Elliot, Haney and Sams-Abiodum (2010: 628) discusses that a disaster is not an event but ’a process that unfolds over time’. Three main phases can be identified in the course of a disaster, namely the pre-crisis, crisis and post-crisis phase (Barton, 1969). In all of these phases support can play a crucial role. Pre-crisis is concerned with preparation to and prevention of the disaster. However for the case of the Chennai flood there was little notice of a pre-crisis phase because for many the water came in unexpectedly. Mathbor (2007) concluded from multiple studies on natural disasters that people who live in the coastal areas are poorly educated and are at risk of being uninformed, especially communities with lower levels of physical, human and social capital. The crisis-phase is characterized by immediate evacuations of people in the affected area, in this phase there is a big threat to loss and damage. In the post-crisis phase the
immediate threat has disappeared but victims are confronted with their losses (Amaratunga and Haigh, 2011). During this phase aid and (inter)national attention are attracted. The support that is supplied involves the provision of food, the reparation of essential facilities and the cleaning of debris. It can be difficult to make exact definitions of phases as they might be overlapping (Amartunga and Haigh, 2011).
Disaster generates different needs for social support at different stages of crisis (Elliot, Haney and Sams-Abiodum, 2010). This implies that throughout the crisis the community has different needs. However these needs are not necessarily being met. Kaniasty and Norris (1995) consider the level of stress experienced by victims an adequate index for the relative need of support and will therefore be most likely related to the amount of help received. Norris, Kaniasty and Murrell (1996) expect that in a situation where many people need assistance the ones hit the hardest will be helped the most. Although relative needs can be a fair indicator for support provision and receipt Norris, Kaniasty and Murrell (1996) and Hurlbert, Haines and Beggs (2000) suggest that irrespective of needs some people are more likely to receive or provide support than others. Aldrich (2008) highlights that some communities have been able to recover quicker than others. As a possible explanation Aldrich points to the important role of a strong community network which has been neglected in many earlier studies. As mentioned by Dynes (2006) the power of social systems is neglected by regarding affected communities as victims instead of active resources. Haines, Hurlbert and Beggs (1996) were also in favour of regarding people as resources. They criticized the fact that a lot of prior studies focused almost solely on the receiving end of support, neglecting that receivers can also be providers. The next subchapter will explain the relevant theories about the operation of support initiated within the community by social networks.
2.2 Social support networks
Social networks are complex and not directly observable and therefore difficult to map. However there have been many researchers trying to understand the use of social networks in times of disaster. Social networks can best be understood through the term social capital. Because the term can be used in different contexts definitions of social capital differ. In this thesis social capital will be used in the context of relief support and will be described as consisting of ‘socially embedded
resources that actors draw upon through their social ties for instrumental purposes (in this case support)’ (Haynes and Hernandez 2008:60). Social capital, in other words,
expresses itself in social relations, not in individuals, which enable the extraction of support from other individuals. However not every relationship is the same. Putnam (2000;22-24) used the concepts bonding, bridging social capital to make a distinction between the different ways social ties can express themselves. With bonding social capital he refers to relationships with people who Putnam (2000: p) defines ‘bonding social capital’ as “inward-oriented networks consisting of relationships among socially
similar kinds of people”. In contrast ‘bridging social capital’ Putnam defines as
‘outward-oriented networks among socially different kinds of people’. According to Woolcock (2000) bridging social capital implies relationships among people different ethnic, geographical, and occupational backgrounds but with similar economic status and political influence.A third distinction that can be made is ‘linking social capital’ and refers to the and individuals’ relationships with institutions and other individuals who have relative power over them. These concepts help to explain the different ways in which social capital can be used to gain support in times of need.
Although social capital can differ individually the greatest strength of the social capital framework is found in the actions of individuals in relation to their community (Hawkins and Maurer, 2000). Beggs, Haines and Hurlbert (1996) were one of the first to point out that rather than individual characteristics studying community context offers a more systematic approach to understanding the operation of informal support. Instead of personal networks that provide broadly supportive relationships the research of Walker, Wasserman and Wellman (1993) shows that different ties and different type of ties provide different forms of support. Murphy (2007) highlights the importance of community organizations. The work undertaken by these organizations can be seen as specific examples of social capital networks in action. Stallings and Quarantelli (1985) suggest that emergent citizen groups in pre- and post-disaster periods are stimulated by the perception that a problem or issue is not recognized or acknowledged by others.
The common assumption that is made is that the access to social support is associated with strong and homophilous ties (Hurlbert, Haines and Beggs, 2000). These are ties that are related to bonding of social capital. In early studies, Drabek et al. (1975) also found that kin remain the most common source of informal aid after disasters. Hulbert, Haines and Beggs (1996) also conclude their research by saying that a dense network is also more profitable than a scattered network. However Aldrich (2008) highlights that many members of one’s immediate social network are often clustered geographically implying that they are all affected and in a vulnerable state. Aldrich (2008) found that as a result people reliant only on local help received
less support than those able to reach beyond their community for assistance. Elliot, Haney and Sams-Abiodum (2010) explain one’s immediate network does not have to be geographically close but depends on both local and trans-local ties. Hawkins and Maurer (2000) found that close ties (bonding) were important for immediate support, but bridging and linking social capital enabled longer term survival and the rebuilding of the neighbourhood and community.
But community networks do not only have to facilitate support because of their pre-crisis structure. According to Haines, Hurlbert and Beggs (1996) disasters can also create ad-hoc networks which brings individuals closer together, referred to as altruistic community. Fritz (1961) introduced the concept of the ‘altruistic community’ to contrast the idea that the period during and after disaster is characterized by disorder, conflict and self-interest. Haines, Hurlbert and Beggs (1996) and Kaniasty and Norris (1995) refer to this concept to better understand the way people behave in non-routine situations such as a disaster. The idea of the ‘altruistic community’ highlights cooperative and altruistic behaviour as outcomes of the disaster situation. The concept might explain why some people acted the way they did; for instance certain victims have shown more efforts because they possessed a certain skill or resource even though there was no personal gain (Kaniasty and Norris, 1995).
Kaniasty and Norris (1995) discuss that the altruistic community would not last because the need for support will increase over time. Kaniasty and Norris conclude a ‘pattern of neglect’ can emerge in which black, older or less educated victims are excluded from altruistic communities. Hawkins and Maurer (2000) found that people gave and received assistance across racial and socio-economic lines. Although they also found reports of ethnic minorities (Afro-American communities) who experienced that socio-economic status and race remain significant barriers in support operation. In the case of India minorities also exist; a lot of discrimination is still found within the caste system (Ariyabandu and Fonseka, 2009). Patterns of neglect could therefore arise discriminating individuals based on caste.
Besides support coming from inside the community, support can also emerge independent of social ties. This will be referred to a external support.
2.3 External Support
External support comes about independent of social ties. External support however cannot be treated as a single flow of support. One flow can be classified as formal support and is related to the more traditional top-down view. Elliot, Haney and Sams-Abiodum (2010) state formal relief assistance has received a lot of criticism. Formal institutions would be too slow or would provide inadequate support. However another flow of external support is becoming increasingly relevant. This support is coming from civil society and can be seen as informal support. Although little studies make this distinction I consider it important because they are different in nature. They have different motives and can make different contributions.
Informal external support is becoming increasingly present as a consequence of social media. Gao, Barbier, Goolsby and Zeng (2011) studied the impact of social media on different cases of natural disasters. They discuss that social media has recently played a significant role in natural disasters as a means to collect information and visualize data needed for the operation of disaster relief. Gao et al. (2011) refer to the power of crowdsourcing through social applications such as Twitter, Facebook, Flikcr etc for the case of the Haiti earthquake in 2010. ‘In just 48 hours, the Red Cross received US$8 million in donations directly from texts which exemplifies one benefit of the powerful propagation capability of social media sites’. In the case of Chennai there have been reports on the big role social media played for the operation of relief efforts. According to an item of the Hindustan Times (2015) constant updates on Facebook, Twitter and WhatsApp have been contributing to the spreading of information on accessible roads and services. Although there can said to be a positive correlation between social media and relief effort, the role of social media does not provide ‘an inherent coordination capability for easily coordinating and sharing information, resources, and plans among disparate relief organizations’ (Gao et al. 2011; p. 10). According to Ghassabi and Zare-Farashbandi (2015) media coverage can play a positive role regarding disaster management as it can encourage lots of people to provide humanitarian assistance. However the information that is emitted to the audience needs to be dealt with carefully. Besides an activator media can also be a distractor. Ghassabi and Zare-Farashbandi also discuss that mass media can create panic among victims and non-victims resulting in a chaotic situation. An appropriate representation of the help that is needed is crucial to initiate an adequate support system. Media can there have influence on the community context to the extent that it can create organization within the community or lead to disorganization.
3. Methodology
Having discussed the theoretical assumptions that form the background of this study brings me to explaining the practical aspects of my research. This chapter gives insight in how the fieldwork has been carried out, how data was acquired and how the acquired data will be dealt with. The choices that have been made regarding the location, population and research methods will be justified. Lastly, the encountered limitations of the fieldwork will be discussed. The fieldwork in Kotturpuram took place from the 16th of April until the 13th of May 2016.
3.1 Research design
The research conducted in Kotturpuram takes the form of a case study. According to Bryman (2008: p. 66) a case study research is concerned with ‘the
complexity and particular nature of the case in question’. Here the case in question is the
operation of support in the community of Kotturpuram after heavy floods. According to Yin (2009) case studies are useful to study a particular social process and help explain why and how it happened. A case study focuses on one specific process which makes it is easier to collect more detailed and in-depth information (Bryman, 2008). The subject of this study is the operation of support for the case of a single community. To guide the research a research question was developed based on the theoretical assumptions that social networks are an important element in the operation of support. The research question that was developed is the following:
How have social networks within Kotturpuram facilitated support operations throughout the disaster process and what has been the influence of external support?
The theoretical assumptions that different phases of crisis generate different needs, personal networks can facilitate support and external support influences community organization were taken into account. They have been used to form sub-question that can help to answer the main sub-question. The following sub-sub-questions have been formulated:
1. How have needs driven the operation of support throughout the different phases of crisis? 2. To what extent did pre-existing networks facilitate the operation of support?
3. What has been the influence of support from external organizations on the operation of support in Kotturpuram?
The formulated questions will be answered according to a mixed-method approach. The several techniques that were conducted will be explained under paragraph 3.3. First the research setting and population will be introduced.
3.2 Research setting and population
The research setting that was chosen is a Housing Board in Kotturpuram. The Housing Board was formerly a place for people to settle down temporarily but currently most of the houses are owned. The area is often referred to as Kotturpuram, although Kotturpuram officially captures a larger area including several smaller wards. In this thesis the name Kotturpuram is used to refer the Housing Board area only. The Housing Board is situated in southern Chennai along the Adyar, one of the rivers that overflowed as a consequences of the high level of water discharge (The Indian Express, 2015). The picture below was taken during the 2nd of December (figure 1). The severity of the impact clearly indicate the need for support in the area. Beside the severity of the impact the choice for Kotturpuram was also based on the people that inhabit the area.
The population of Kotturpuram is relatively poor and inhabited by lower-income households however there is quite some diversity in the population. Although most houses look similar there are some distinct differences among blocks. Other signs of diversity in the population are the multiple places of worship that are present. Because the neighbourhood is geographically closely-knitted it makes it easier to cover a large and diverse set of respondents which increases representativeness of the case. Because the area of Kotturpuram is relatively small it is an accessible area for the scope of this fieldwork. The images below (figure 2;3) show satellite images of Kotturpuram. One picture was taken on 19th of march 2015 (before the flood) and one was taken on the 21st of February 2016 (after the flood.). In the first image a red marking is used to show the boundaries of the research area. Both images are provided with a scale in the left bottom corner. The images clearly show differences in the built environment. It is visible that several homes that were located along the river have disappeared. It is however not clear to what extent the flood directly influenced this difference since material directly before and after the flood is not available or unclear.
Figure 1: Kotturpuram on the 2nd of December. Source: The new Indian Express (2015). Retrieved from:
http://www.newindianexpress.com/photos/nation/Shocking-Pictures-of-Devastated-Chennai-in-Floods/2015/12/03/article3159072.ece on 24/06/2016
Another factor that has determined the choice of Kotturpuram was my assumption that the community has a dense network. Because this study deals with the operation of support through social networks a population with a lot of close connections might be easier to research. The following paragraphs will explain how the research in the area took place.
Figure 3: a satellite image by Google Earth (2015) of Kotturpuram before the 2015 floods
3.3 Research methods
This thesis is based on qualitative research. As mentioned earlier a mixed-method approach was used during the fieldwork. The research methods include observations, semi-structured interviews, interviews with key informants and an adjacent literature study.
Literature study
Prior to and after the fieldwork a literature study has been conducted. Part of this study was to create the theoretical framework that provided the background of the research which was crucial before going in the field. After the fieldwork literature has been used to strengthen the theoretical framework and use it to explain findings in the field.
Observations
Observations have been useful to get an idea of the population composition and their characteristics such as economic status, based on housing and household possessions. Participatory observations also have been useful to gain people’s trust. According to Kawulich (2005) participatory observations are characterized by having an open, nonjudgmental attitude, which I adopted. Participatory observations were
almost made constantly even during interviews. Observations have led to insight in the composition of the community, the impact of the flood and have therefore contributed to the generation of new questions within my research. Because social networks are not visible most of the information had to be acquired through interviews.
Semi-strucured interviews
Semi-structured interviews were held with in total 35 respondents. These respondents included people that were considered part of the community of Kotturpuram and have stayed there during the floods. These people will be represented as members of the community rather than ‘independent’ individuals. Regarding privacy issues and the use of common or difficult names, I choose not to refer to any. People may be identified according to their gender, age and place of residence within the community.
Prior to the interviews my translator would introduce my research to the approached respondents to comfort them and let them understand the context. The interviews would take the shape of conversations in which a topic list was used to structure the content. The topic list that was used initially had been based on earlier interviews with flood affected respondents. During the interviewing the topic list had been under constant adaptation. Answers of respondent led to new questions about topics which were not included in the list and therefore the topic list was adjusted regularly to include these new insights. The item list that was used during most interviews can be found in the appendix (reference). According (Bryman, 2008: 12) semi-structured interviews are useful to enable a more open-minded approach for the researcher, so that concepts and theories are a result of the acquired data. Using this method has made me more flexible in order to discover relevant data.
Respondents of the interviews were approached according to a sampling method. According to Bryman (2008) sampling of population within a case study remains important because the respondents need to be relevant to the study. Selecting the best sampling method depends on the nature of the research. A maximum variation sampling was mainly used to select respondents. This would give the most insight in the community composition and network. According to Bryman (2008: 419) maximum variation sampling is used ‘to ensure as wide a variation
as possible in terms of the dimension of interest’. Respondents were selected on gender,
age and place of residence within the area. But convenience of who was available also played a role. Although maximum diversity was sought significantly more
women were approachable during the day time since most of the men were at work. Therefore the research session often took place from 3 p.m. till 8 p.m. to include women as well men as in the session. Although more women than men were interviewed I do not consider this as a big constraint for my research. Respondents were not only selected based on certain characteristics. Snowball sampling was also used to a small extent. Snowball sampling is technique in which respondents propose other potential respondents who have had the experience or characteristics relevant to the
research (ibid.). This technique was considered because as Bryman (2008: 424) points
out it enables to capitalize on and to reveal the connectedness of individuals in networks. Since networks are the main concept of this research using this method can be very beneficial however the scope of this research was too small to capture entire networks and therefore the technique has been used primarily to understand specific support operations.
Because the research population was not selected randomly a sampling bias might exist. According to Bryman (2008:188) human judgment often affects the selection process of respondents, making some members of the population more likely to be
selected than others.
Interviews with key-informants
Besides interviews with inhabitants interviews with key-informants were held. Key informants were used to gain specific knowledge which would have taken a long time to extract from the normal respondents (Bryman, 2008). This information had to deal with the operation of support among different community organizations. Names of the key informant were taken (with permission) to make references easier and to avoid confusion among the different informants. The list below will give an overview of the key-informants:
Name Gender Organization
Kanchana F Social worker DMK
Amdu M Hindu temple + social work DMK
Kamani M DMK secretary Kandan M AIADMK secretary Pushpanani F Police office Satish M Volunteer Gomati F Volunteer Syed M Mosque 1 Sabajahan M Mosque 2
Father Baskar M Roman catholic church Jarifa F Protestant church Jabaseelan + youth
group
M Protestant church Sapathi M Hindu temple
Table 1: Key-informants by name, their gender and the organization they belong to.
Knowledge about who were considered key-informants came about gradually during the first week of research. Respondents were asked about the social structure of the community, whether they were part of internal organizations or knew people that were. The interviews with key-informants would give insight in the social organization within Kotturpuram and were therefore an essential part of the research. Contact with key-informants was often sought through respondents that were familiar with these people. This would make it easier to approach certain individuals. Especially political figures had to be approached carefully because of their authority in the area. Not only the approach required a careful position the questions posed also required some sensibility. Asking questions that are too upfront or considered inappropriate may cause the respondent to act in a more reserved way. Therefore I choose to use unstructured interviews in which I would let the respondents talk freely and from there would ask new questions that would guide the course of the conversation. This was also useful to discover new insights. Based on their answers and the considerations of my translator I would ask follow-up questions appropriate to the standards and values of the key informant. Bryman (2008) warns for the risk to view reality through the eyes of a key informant. Because information from key informants is more easily accessible the researcher might unconsciously start to rely on it.
3.4 Limitations: validity and reliability
The most important limitation of this research included the available time of four weeks to carry out the fieldwork. Speaking with 35 inhabitants of Kotturpuram gave me some insight in what have been important factors that influenced the operation of support but no statistical analysis could have been made. Generalization of a small population sample is difficult which limits the external validity of the research (Bryman, 2008).
Not being able to speak Tamil, the local language, required the assistance of a translator. Working with a translator had a significant influence on the research. Not being able to understand the questions posed and answers provided led to a certain
extent of content loss. Translations were quite often summaries of the actual answers, possibly leaving out relevant quotes. The information provided by my translator might have been her own interpretation of the answers, this limits as well the external and internal validity of the research. On the other hand having the corporation of a translator also had its advantages. As I was not very familiar with the local culture having the assistance of someone that is familiar helps me to explain more of the context. Also my translator had been active in the support operations during the 2015 floods which enabled some conversations with volunteers that had been active in Kotturpuram. To increase internal validity triangulation was applied. Besides the fact that mixed-methods are useful to require diverse data it is a way of cross-checking findings, referred to as triangulation (Swanborn, 2005). Besides empirical data from observations most of the data is based on the perceptions of respondents. Because this data might be considered subjective using different methods and interviewing a diverse population creates a higher internal validity. Another influence on the internal validity of the research was that a lot of interviews resulted in group sessions, which made it more difficult to extract an unbiased answer from the approached respondent. However the interaction with a large amount of people also helped me to become familiar with the community. Figure 4 gives a lively representation of an average interview.
I experienced that my own appearance influenced the position of respondents. People in India or more specifically in Chennai, Tamil Nadu are not very familiar with the presence of foreigners especially my white skin colour was fascinating for the local population. People had mixed responses to my presence but all were very willing to participate. However some respondents wondered whether I was a social worker and would provide them with support. This assumption could have made them respondent in a different way which would also limit the validity of the research. Researchers with a different position or appearance might have received different answers, therefore the external reliability is limited.
One of the contextual factors that might have impacted the external reliability of the research was the fact that the fieldwork took place during the election time. In Tamil Nadu political involvement is really high and therefore some answers might have been given to be politically correct. They would do this because very often respondents were accompanied by a lot of others.
4. Course of events: Floods in Kotturpuram
In this chapter an overview is given of the different phases of disaster the community of Kotturpuram went through. Observations together with the stories of respondents have been used to put together a timeline which includes important events regarding the operation of support after the flood. The concepts of pre-crisis, crisis and post-crisis will be used to structure the timeline. The distinction among the different phases has been made based on my own interpretation.
4.1 Community composition of Kotturpuram
The homes found in Kotturpuram are not very diverse. Most of them are three-storey flat apartments and the other homes are single storey buildings. According to several respondents the area is inhabited by around 3500 households and 10.000 individuals. According to personal observations and information received by Kanchana, one of the key informants, a classification of four housing sectors could be made within Kotturpuram. On the map (figure 7) below an overview is given of these sectors which have been named: H-block, R-block, Chitra Nagar and the Watertank Colony. The last one, the Watertank Colony, contains newly built homes that are not entirely finished. One female respondent who stayed in the building during the flood mentioned that ‘eventhough they are already assigned they are not yet
inhabited, however some people stay there or use it to store stuff’. Although the H- and
R-block looked similar the area of Chitra Nagar seemed relatively poorer. This led to the assumption that social-economic differences exist within the community. Respondents were selected based on their place of residence within the area, from every sector households were selected for interviews. From the information provided and observations made during these interviews it could be concluded that there were clear differences in income-level and material possessions. Facilities in Chitra Nagar were very poor, there is no access to water inside the homes and no proper sanitation. ‘We have to walk for about five minutes when we want to use to toilet or need to
wash’ one respondent in Chitra Nagar mentioned. In the H- and R-block all
households had access to water.
Besides economic status differences in social status were also present and were also related to place of residence. ‘The H-block is occupied by Brahmins, the R-block is a
mix of Brahmin and lower caste and Chitra Nagar is occupied by lower castes’. Although I
did not ask people about their caste I experienced that several respondents living in the H-block mentioned that they were Brahmins. Assuming that Kanchana made a
12% 14% 74%
Religion in Kotturpuram (N=35)
Muslim Christian Hindu
fair statement about the caste division a possible relation between caste and the economic status can be identified.
Religion is another way in which diversity within the population is expressed. The majority of the respondents we spoke to is Hindu (74%), followed by Christian (14%) and Muslim (12%) as can be read from the pie chart below (Figure 5). Of all religions a place of worship is present in the area. I did not find any signs that people of different religions live segregated from each other. ‘Here it is not the case that Muslims
live in one part and Christians in another, we live together’ one female respondent said.
Figure 7 gives an overview of the different places of worship within the area.
Figure 5. An overview of the religious composition in Kotturpuram. Soruce: Floor Matla
The last distinction that can be made is based on political preference. There are two main parties of which AIADMK is the ruling party and DMK the opposing party. Difference in political preference appears to create some distance among people. When I asked a female respondent whether she thought political parties were cooperating, she responded firmly with: ‘No. They do not even talk to each other’. Kamani the secretary of DMK in Kotturpuram emphasized the power of politics ‘When the one comes to power, the other will go to jail’. When asking people if there are a lot of conflicts within their community they generally say that conflicts happen but that those remain small and are not long-lasting. They did not specifically refer to political conflicts.
Gaining insight in the community composition helps to understand the context of the community network and the possible influence it has on the operation of support. The context helps to explain the pre-crisis structure of the community which is crucial to understand how support is operated within it. The next paragraphs will give an overview of how the different stages of crisis impacted the community.
Figure 6. A map providing an overview of the important locations in the area including a legend on the right. Source: Floor Matla
4.2 Phases of crisis
In this chapter an overview is given from the impact of the flood on the community of Kotturpuram. The overview will be structured according to the different phases of crisis, namely pre-crisis, crisis and post-crisis. The overview will include events from the first of December up till the period of April and March in which my research took place.
4.2.1 Pre-crisis
December 1st: During the day the heavy showers had appeared, resulting in water reaching knee-level. According to some respondents the police had given warnings about more water coming in, although some indicated that there had been no
warning at all. The clarity of the warning is therefore considered doubtful. More importantly the warning has not been effective since no one was prepared or had left the area. The only preparations that were taken by people were stocking of valuables in higher places and some people also moved to relatives living on higher levels. According to, Pushpanani, one of the police officers who had been active in the provisioning of warnings said people were asked to leave but refused to. ‘Warnings
were made in several parts of Kotturpuram using speakers’ Pushpanani indicated that
people had underestimated warnings as a consequence of a similar but milder flood that occurred earlier in November. According to (Coombs, 2007) the pre-crisis phase deals with precautionary action; in the case of Kotturpuram the only preparations that were made included stocking of valuables in higher places and few people had moved to relatives living on higher levels.
Around midnight significantly more water started to come in, brought in by the Adyar. The banks of the Adyar had overflowed causing the water to enter at a very high speed. The source and severity of the water flow were very unexpected. In addition one female respondent told me that the water entered from a different side than usual. The respondent explained to me that usually when floods happen in the area water enters from the side of the bridge; in this case water entered from the opposite side (as indicated in Figure 6). On the map can also be seen that the people in Chitra Nager were the first to be affected. All of the respondents indicated that water started to come in around midnight, about 12.30 a.m. During this time almost everybody was at home and asleep. However the people in Chitra Nagar who were the first to be affected indicated that they had been awake when the water came in. This had to do with the fact that in a lot of houses water was present so people were unable to sleep on the floor. Some respondents also indicated that they were worried about more water coming in. Although people faced these difficulties they did not leave their homes. Some respondents indicated that they did not leave their houses because they were scared to lose their possessions. In the H- and R-block people living on the ground floor locked their doors and went to sleep in places higher up, with neighbours or with relatives close by. The fact that the flood came in so fast and so unexpected made turned the area in a state of crisis.
4.2.2 Crisis
When more water started to come in people in Chitra Nagar were forced to leave. Because of the speed at which the water was entering the inhabitants had little
time to get out of the area. ‘We were not able to take anything. The water came in so fast.
And it was also very dark because there was no power’ one female respondent indicated.
Loss of material has been one concern but more important was the concern about personal safety and that of others. To make sure that everyone would leave their home people in Chitra Nagar went to check up upon each other. ‘We went as a group’ I was told by one female respondent. Leaving the area had been difficult because the people had to walk through the flowing water. People knew from previous floods that there is a park nearby which is higher up. ‘We knew where to go because floods happen more often; we always go there’ one female respondent indicated, pointing to the direction of the park. According to my respondents almost everybody in Chitra Nagar went to the same place to find shelter.
For the people living in the other blocks responses had been different. Residents living in the middle of the Housing Board were in contrast to the people in Chitra Nagar asleep when the water came in. Because most of the homes include three-storey buildings finding shelter had been easier. Especially people living on the ground floor were impacted, including material losses. During the day some households had try to save belongings by putting them in places higher up however the water level was much higher than expected and there was no time to save any of their possessions. In a very short period of time water completely drowned the ground floor homes. Even on the first floor water began to enter. ‘We had never
expected the water to be this high’ almost all respondents mentioned. In some cases the
water would not reach the first floor, this is because within Kotturpuram some areas are positioned higher. People living on the first floor were generally less badly affected and were able to save most of their possessions. Because the terrain in Kotturpuram is not equal some first level apartments were impacted more than others. The homes of people on the second floor were not affected by the water. People living on the ground floor found shelter with neighbours living on the first and second floor. ‘Some were known but a lot of people were unknown’ one male respondent told. This indicates that newly made social capital also played a role in the operation of support.
People living on the river banks in poorly constructed homes faced large damages to their houses. When I visited the area in April only a few houses were there (Figure 8). Many of the houses appeared to have washed away during the floods. However one male respondent mentioned that the houses had been damaged badly but that the government took them down3. Just like the households in Chitra
Nagar people living along the riverbank were awake when the water came around midnight but were not able to leave the area. Due to the their geographical position and the force of the water they were not able to leave the area and had to find shelter elsewhere. ‘A lot of people found shelter in this building’ one male respondent explained, pointing to a large empty building. The houses referred to as the Watertank Colony (WTC) were used for shelter during the flood, according to one female respondent. ‘A lot of people were there, but I don’t know where they came from’.
The high level of the water together with its force turned the area into a place of disaster. Finding shelter has been one of the priorities when the water came in; ensuring personal safety and that of others. However the phase of crisis had not yet past. As a consequence of poor drainage systems water was stagnating. In some parts of Kotturpuram the water went up to as high as four meters. Besides the fact that people were confronted with high levels of water, the conditions of the water were very bad. Since a lot of sewage systems were damaged by the force, the water was mixed with a lot of dirt.
Figure 8: The area along the river bank were the huts used to be. Source: Floor Matla
During the second day the situation had not gotten better, water levels were high due to stagnation and the force remained strong which made impossible to leave the area for food or shelter. People were stuck in their homes and depended on their stock and that of others. Several respondents indicated that food was shared freely: ‘All the food that was present was shared, all neighbours provided what they had to
offer’. One female respondent indicated that she was provided with food by the
people had in stock was generally low because most people live on daily wages’. It was also
said by another respondent that ‘Because floods happened at the end of the month people
had no money left to stock’. In addition, the flood came as a surprise so people did not expect a stock was needed.’ Until evacuations could take place people were depended on each
other’s efforts and that of external sources. The first provisions of food were provided by helicopters. Several respondents showed me videos or pictures they had on their phones from helicopters flying over. ‘Helicopters mainly provided biscuits, bread
and water-packages’. Provision of food had been one form of support but the
provision of emotional support was also needed. There had been not network and no electricity so people were not able to contact relatives and friends which created concerns about their safety.
After about four days evacuation could take place by boat (figures 9 and 10). In the pictures on the next page an impression is given of how evacuations were managed. One of my first respondents had told me ‘Fishermen from nearby villages
came to with boats and brought people to the nearby police office’. This was also confirmed
by other respondents. Pushpanani explained that the police station is positioned higher and therefore was not affected. She also explained that once people were brought their they could go their own way. Most people went to relatives living in areas that were not badly affected. In the area of Kotturpuram many public spaces were available which provided shelter which included marriage halls and schools. Not everybody left their home. Some people, generally those living in higher levels, choose to stay. In some cases only the wife and children went away, the husband would return to Kotturpuram to take care of the home and to help others. Two female respondents showed their dissatisfaction about their male neighbours leaving the area: ‘These are young men, they could have helped us. Instead they left to native places’. The people that came from Chitra Nagar told me that they had been displaced to ‘the
Anna Gem. Gymhall’, close by. They were able to stay there for an entire month.
People in the staying in the WTC were among the last ones to be taken by boats because the water force had been too strong earlier. Together with the people that choose to stay they were depending on food distributions coming from outside. Boats were also used to distribute food and water but it had been difficult to cover the entire area due to the force of the water. One female respondent who stayed in the WTC indicated that she was not satisfied with the provisions of food: ‘We were
only given biscuits for three days. How are we supposed to survive on biscuits’. After the
third day the water started to slow down and the process of food distribution became easier. A lot of materials were handed out at several distribution points. Many
respondents indicated that there was a distribution point situated on the bridge. Food and water made up the most important relief goods. Although a lot of materials were coming, some people received less than others. Some people mentioned that a lot was provided whereas others mentioned that the provisions had not been satisfying.
4.2.3 Post-crisis
‘After about six days the water had receded’, I was told by all my respondents. People who left the area were able to come back in order to clean their houses. The sewage water had left a thick layer of dirt in the streets and homes of people, especially the homes on the ground-level were impacted. Because the area was now accessible more relief support was able to enter the area. Help was coming from different actors, but primarily from non-governmental organizations (NGOs). Larger and smaller organizations from in and outside Chennai were active in helping out Kotturpuram and other affected areas. The level of voluntary support has also been impressive. Many respondents mentioned the support of volunteers and there were many reports in local media4. The support provided mostly involved cleaning, the distributions of cleaning materials and the distributions of clothes and bedsheets. All respondents indicated to have received support from NGOs and volunteers, some however more than others mostly as a result of a lack in accessibility of the neighbourhood. Some parts were not reachable and therefore received less relief goods.
Although most people returned to Kotturpuram to clean their houses some did not. Most of the people that used to live in the houses along riverbank were relocated to another part of Chennai, called Perumbakkam. Also the people living in Chitra Nagar only returned to live in their homes after one month but would frequently come back to Kotturpuram to clean up their houses. The process of cleaning was one of the main activities during the first two weeks after the flood. Everybody was concerned with the same activity and people that were less badly affected helped their neighbours. Relief work also kept on for about two weeks during which cleaning and health care were the most important issues to look after. The government allocated some financial compensations for the affected households.
4 The Hindu (2015). Retrieved from: http://www.thehindu.com/features/metroplus/society/chennai-residents-lend-a-helping-hand-in-the-rain/article7940986.ece on 23/06/2016,
The Indian Express (2015) .Retrieved from: http://indianexpress.com/article/india/india-news-india/rescue-ops-almost-over-focus-now-on-relief-work/ on 23/06/2016,
Hindustan Times (2015). Retrieved from: http://www.hindustantimes.com/india/ex-gratia-relief-for-tamil-nadu-floods-victims-receding-water-brings-trouble/story-8o9qDsfxZL21760lpu2jfN.html
It is also relevant to note that the flood had a significant impact on the mobile network. Many affected areas had limited to no service. Many respondents in Kotturpuram indiciated that they were without service for about four days and some even longer. This was also due to the fact that there was no current to charge the phone. When the network started working again the level of support increased even further.
Figure 9: Evacuations in Kotturpuram. Source: The Hindu (2015), retrieved from
http://www.thehindu.com/news/national/tamil-nadu/heavy-rains-continue-to-lash-chennai-airport-train-services-hit/article7940421.ece
Figure 10: Evacuations in Kotturpuram. Source: Quartz (2015), retrieved from: http://qz.com/563843/chennai-flooding-photos/
5. Support networks in Kotturpuram
In this chapter a closer look will be given to the operation of support and how it came about. The chapter will focus on the internal organization of support in the community of Kotturpuram but also the influence of external support on the community.
5.1 Drivers of internal organization in the community
5.1.1 Personal networks and AltruismThe first help that was needed was related to ensuring personal safety by finding shelter. Neighbours were the first sources of support when the water came in. As indicated by Hawkins and Maurer (2000) the first help received comes from
immediate or proximal networks which are related to bonding of social capital. Hawkins
and Maurer also highlight the importance of bonding social capital prior to the crisis in the development of a plan of action with family members. In the case of Kotturpuram these plans included staying with relatives on higher levels and preventing loss by storing possessions higher up. Mathbor’s (2007) findings that poor coastal communities are often at risk of being uninformed about their vulnerability can be considered applicable to the case of Kotturpuram. Although there had been warnings little to no action was taken to evacuate the area.
When the water came in around midnight people were surprised and had no time to make a plan. The pre-crisis very rapidly turned into a crisis phase. Households whose homes were not affected opened their doors to those who were seeking shelter. All respondents either provided shelter or received it. According to the stories of respondents shelter was also provided to people that were unknown. This high level of corporation and solidarity within the community can be explained according to the concept of altruistic or therapeutic community. Kaniasty and Norris (1995) discuss that in an emergency state people are brought closer together as a consequence of a similar fate. Besides shelter food was shared freely among households, known or unknown. Even after the direct threat had disappeared households continued to cooperate.
Because buildings were isolated from each other by water cooperation mostly happened within blocks. Cooperation of blocks was present in the entire area. Many respondents indicated that they were sharing water, food and company with their block. People would come together on the roof to collect goods distributed by the
helicopters but also to cooperate. As a consequence of the interaction and corporation with unknown people. Several respondents indicated that they had made new connections with who some are still in contact. ‘We are still very good friends’ one male respondent commented enthusiastically to newly made contacts. Many female respondents indicated that a large group of women would gather on the roof to cook for the entire block. One female respondent mentioned: ‘There were about hundred
women, all cooking and cleaning’. Men were concerned with the provision of food. They
had helped in distributing food among households. ‘Me and my brother went out of the
area in order to collect food and come back. Were able to provide rice for around 20 households’ one male respondent living in the front side of Kotturpuram explained.
Participation within the altruistic community is as Kaniasty and Norris (1995) explain also dependent on personal characteristics. Reaching other blocks was only possible for the ones with the ability to swim. For people who were not able to reach the distribution point by foot, swimming was the only possibility to get food. People with the ability to swim also collected food for others in the block. The ability to swim is to a certain extent beneficial. One male respondent told me that he saved three people that had been taken by the force of the water and were struggling to save themselves. Since the boy could swim he was recruited to help, even though he did know the victims he risked his own life.
Besides altruism personal-networks have played a role in the facilitation of support. According to Hurlbert, Haines and Beggs (2000) having a lot of strong and homophilous connections increases the access to support. Having direct support of family is beneficial because family members can help with the preparation to the crisis, offer shelter and provide direct emotional support. One woman indicated that a close tie in the area helped her carrying her disabled son upstairs when the water came in. During the first days people were reliant on the support of connections within their community. Reaching out to relatives and friends outside the area had been impossible because Kotturpuram had been geographically and electronically isolated. Up till then people were depending on local connections. People with a lot of relatives within the area faced less concerns about well-being of relatives outside the community. ‘We were worried about my parents, they were also living in an affected
area. After four days I was finally able to make a call’ one female respondent explained
her concerns about family. After about four days mobile services returned enabling contact with distant relatives and friends.
However Aldrich (2008) indicates that reliance on close and proximate ties can also be a disadvantage as these are likely to be in a similar vulnerable positions.
Aldrich concludes that people who are able to reach out of the community are likely to receive more support than those who are not. During the evacuations households with relatives in distant places were able to find shelter there. They went to various areas that had been less badly affected, these included Tambaram, Minjur, Alwarpet, Mandaveli, Triplicane etc. Having family in these places allowed people to keep their children and themselves in safe and familiar environments. Because of all the dirt that the water brought along people were scared of infections, especially for their children. There were also respondents that mentioned their family was living too far away or that they were not in the position to help. ‘They are also affected. They cannot
do anything’ was a response I received from a female respondent in Chitra Nagar.
People that were not able to reach out therefore had more difficulty in finding a a comfortable place to shelter. In some cases family members were living in apartments higher up within Kotturpuram so shelter could be found there. In other cases friends were able to provide shelter. ‘First we went to my brother’s apartment here
in the area. From there two friends of his would come by boat and would take us out of the area. They took us to a hotel for which they would pay’ one female respondent explained
how she found shelter. In this case both local and trans-local ties have played a role. Even when people returned to the area to resettle the support of close-ties remained important. People in the area that were less badly affected would help their neighbours with the cleaning of their houses. Some respondents mentioned that family would come over to Kotturpuram to help with the cleaning. Most of the help with cleaning was however coming from sources outside the community networks. This will be further discussed in chapter 5.2
Whereas altruism operates on community-level personal networks differ among individuals. However the two can also come together when personal networks are used to make contributions on community-level. Murphy (2007) describes that community organizations are important actors in facilitating support from bottom-up.
5.1.2 Contributions of groups
During my research I found two examples of group efforts in which pre-existing connections were used to take action. Stallings and Quantarelli (1985) discuss group efforts in disaster are stimulated by the perception that a problem or issue is not recognized by others. A group of boys in Kotturpuram took action when they recognized that not everybody was receiving equally. ‘We helped with the
