The Western Cape is a dynamic province that is disaster-prone, particularly the vulnerable urban communities in and around its environs. Such communities are more vulnerable to wildfire, flooding, pandemic, natural and human-made hazards because of poverty and, consequently, poor living conditions such as overcrowding and non-understanding of community resilience. The inability of these communities to understand community resilience and withstand adversities affects the sustainability of initiatives to develop them. This study aims to identify the mechanisms influencing the level of understanding of community resilience in a vulnerable community and to contribute to the understanding of community resilience to disaster risk. Fieldwork was conducted in an informal settlement in South Africa. The research study was conducted in two cycles of data collection and analysis. Data in the form of observation notes, document analysis and interviews were analysed using grounded-theory principles. Ten inter-related variables or mechanisms emerged from the analysis. The theoretical model consists of four reinforcing (R) feedback loops (R1, R2, R3 and R4), respectively, which explain how the understanding of community resilience in the informal settlement maps on to the relative achievement systems archetype. Negative reinforcing behaviour would explain the lack of understanding of community resilience, while positive reinforcing behaviour indicates how an understanding of community resilience develops. In addition, the variable with the leverage to improve the mechanisms influencing the understanding of community resilience was found to be the ‘level of public education and awareness’. The theory of how these variables behave in context was represented as a qualitative system dynamics model.
Several reports have declared Cape Town as the most disaster-prone city in South Africa, with the highest rate of disaster-related deaths, which mostly occur in informal settlements (Walls et al.
Disasters could steadily jeopardise South Africa’s populations, economy and the sustainable community development of the nation (Hyogo Framework for Action [HFA]
There have been various approaches to disaster risk reduction (DRR). Gaillard and Mercer (
The bottom-up approach values local knowledge and recognises that communities could play a role in managing DR. An integrative approach that combines bottom-up and top-down approaches is argued for by Gaillard and Mercer (
The extent to which societies and citizens mitigate the risks and consequences of disasters is referred to as resilience (Paton & Johnston
In this study, DR in the Western Cape province is viewed as a complex adaptive system. This view challenges simple cause-and-effect assumptions and recognises that components in a system are connected and interact in ways that cannot be predicted. This study therefore aims to identify the mechanisms influencing the level of understanding of CR and how to intervene with respect to them for the purpose of improving resilience and reducing DR in the vulnerable community. To achieve this purpose, this research study focuses on the case of the Phola Park community in the Western Cape province to construct a theoretical model grounded in the empirical actualities of those involved in DRR in the Phola Park, and intends to answer the following research questions:
What is the underlying mechanism(s) influencing the level of understanding of CR in a vulnerable community?
What is the determinant(s) for improving the mechanism(s)?
This site was chosen as a case study for this research because at the time of this study it was an informal settlement that had experienced disasters and needed some level of understanding of CR. To address the complex problems relating to the challenges of DR and the need for understanding CR in the Western Cape, a system dynamics (SD) approach was used in conjunction with appropriate modelling techniques to identify and understand the variables in a community and how they interact when faced with possible disasters. System dynamics modelling has been employed for development planning and support-related decision-making. Brent, Simelane and Clifford-Holmes (
The benefits of understanding the mechanisms contributing to resilience found in vulnerable communities include: (1) possible cost savings on disaster recovery or emergency relief and (2) improvements in CR and disaster response. Studies have shown that tackling vulnerability using resilience development is cheaper than emergency relief (Venton et al.
System dynamics, a particular system thinking approach, was used in this study to achieve the desired purpose of the study. System dynamics is a computer simulation technique which was firstly introduced at the Massachusetts Institute of Technology in 1950 by Jay W. Forrester. This approach has been applied extensively in the field of finance, healthcare, economics, environmental research, biology and information technology since the 1960s (Sapiri et al.
System dynamics has a main goal to understand how the component in a system interacts with each other. The interactions in the components of a system are done through the feedback loops, which means that a change or increase in a component affects the other components (Sapiri et al.
The research problem is analysed as a complete or comprehensive ‘system’ in the SD modelling. Therefore, according to Sapiri et al (
The feedback loop that comes in the form of the causal loop diagrams (CLDs) derived from a software called the Vensim (Sapiri et al.
[
The Vensim software mentioned above was established by ‘Ventana System’ when they were determined to put a stop to the reconstruction of old software and decided to make simulation languages of their own, which is now referred to as ‘Vensim’ (Sapiri et al.
Holling (
According to Cutter et al. (
Cutter et al. (
After looking at the available definitions of resilience, the researcher has chosen a definition of resilience that is appropriate for this research project. In this research study, resilience is therefore defined as a positive turnaround, outcome, or adaptation of a competent individual or community exposed to risk, difficulties, significant challenges or adversities. This is because people need to know and understand resilience to be able to build it within themselves or their communities in the context of withstanding DR (Cutter et al.
Community resilience has similarities with the individual resilience because a community is made up of individuals. In a report given by the Community and Regional Resilience Institute (CARRI), there were 25 CR definitions given by several researchers (Eachus
The CARRI (
Community resilience has contributing elements or factors. Several researchers have developed factors that contribute to CR. For example, Joerin et al. (
Also, in acknowledging the contributions of current models and their restrictions around resilience and vulnerability, there is a model called the ‘DROP’ model for disaster resilience, which was designed to show the theoretically grounded connection between vulnerability and resilience; and this model can be applied to tackle real challenges in real places (Cutter et al.
Community resilience is seen as a dynamic and fundamental characteristic of DRR, which means that resilience should be a part of the community’s life cycle (CARRY
Disaster risk reduction is the concept of practices used to reduce DR by systematically carrying out a detailed analysis and reduction of the factors causing the disasters (United Nations Office for Disaster Reduction [UNISDR] & [WMO]
Disaster risk reduction is linked to sustainable development, which involves regulation and training around key disciplines that include disaster management, disaster mitigation and disaster preparedness. Everyone’s involvement is necessary to develop and sustain these activities and disciplines in order to reduce DR (UNISDR & WMO
The gravity of a disaster depends on the extent of the effect of hazards on the communities, as the level of the effects depends on the choices made by the people for their lives and communities (UNISDR & WMO
In 2010, shelter cluster indicated that a community or an environment has an important part to play in the DRR and the effect of disaster events, because proper ecosystem management can mitigate the risk of hazards such as landslides, flooding and storm surges. Guion, Scammon and Borders (
[
For more than 50 years, the four-phase traditional model of disaster management – namely, mitigation, preparedness, response and recovery – has been utilised and has informed several research studies on DRR (Mileti
This four-phase traditional model could work better when understanding of CR is part of the model. This is because preparedness, response and recovery without understanding CR in the informal settlements might lead to the recurrence of disasters (Cutter et al.
In the next section, the context for the study is described and the methodological aspects of the research design, including SD modelling, are explained. Then, a detailed explanation of the theory building process is provided to demonstrate how isomorphic analogy is used to support the development of a grounded theory.
This study was conducted in the Phola Park community of the Western Cape province in South Africa. Phola Park is an informal settlement located along an arterial road in an area designated for development and is similar to other informal settlements in Cape Town (Haferburg
The Phola Park community is exposed to flood, wildfire and other DRs. This area consists totally of shacks (houses built by hand with available materials) that are frequently prone to disasters because of inadequate drainage system, the lack of housing and general facilities such as toilets and water taps and overcrowding.
This study takes an inductive, grounded theory approach for understanding CR to DR. This article aims to identify the mechanisms influencing the level of understanding of CR and to use these as a basis for recommendations for reducing DR.
Data were gathered through observations and interviews. Participatory observations were performed over 3 months with weekly site visits as a process of gathering data through involvement (widespread community participation) and exposure to day-to-day activities in the informal settlement. Community activities, such as meetings, were also attended by the researcher. A total of 10 members of the community who had personal experiences of DR participated in a focus group.
Semi-structured interviews were conducted with 30 people, mainly youths from the community and the community leader. These people were identified as the most trusted and ardent community members in Phola Park at that time. They were each interviewed for 20–30 min. The choices made with respect to sampling, settings or individuals to be observed and interviewed were informed by the aim to provide credible data rather than representativeness or the capacity to generalise (Nastasi, Moore & Varjas
The analysis was conducted simultaneously with the data collection, focusing on the interviews and observations concurrently. Categorising strategies (open coding, axial coding, selective coding and thematic analysis) and connecting strategies (including narrative analysis and individual case studies, memo and displays) were applied. The summary of the research process from the data collection stage to the analysis or the grounded theory process is depicted in Online Appendix 1.
Access to the informal settlement was facilitated by a non-governmental organisation that was working on a project with the Phola Park community.
System dynamics is an approach for modelling complex systems. A context or problem can be conceptualised as a system with interacting variables and SD provides techniques for modelling the behaviour of such a system, including identifying the most impactful elements of the system (Brent et al.
The theory-building process involves the identification of concepts from the grounded theory analytical process. These concepts informed a review of literature to extend on the empirical definitions and contribute to the grounded theory. Then, the theory was further developed using an analogy process as developed by Beer (
In this study, theory-building involves analogy and isomorphism steps that are illustrated in
The process of building a theoretical model: ‘A transformational view of metaphors in organisation science’.
A representation of the structured mapping procedures, explaining the analogical reasoning.
Kieras and Bovair (
The integrated theoretical model of understanding of community resilience based on a community vulnerable to disaster risk.
Concept analysis was performed on the 10-core categories generated in the study, which provided more insights into the core categories. In this step, the analysis of the data derived from the study was used to choose a variable for the categories within the context of the study and paying attention to its importance.
The relationships between variables were tested using an interrelationship diagraph, a tool for exploring for evidence of links or influences between each of the variables identified.
In step 3, a generic archetype was chosen from the Wolstenholme generic archetypes and Braun’s archetypes (Braun
Selection of appropriate Wolstenholme systems archetype as known theories of systems behaviour that maps to the empirical findings.
This step involves the identifying feedback in the causal loops, how they affect each other and the outside environments. The process allows for the documentation of the identified known structural features of the selected archetypes.
This is where the CLD (SD model) is generated, which means that the new findings and understanding derived from the CLD of the research work are explained here. This includes inferring the variables from the CLD into an essential segment of the selected systems archetype in a way that makes it sensible.
As soon as the structural feature of the selected systems archetype is inferred into the interrelationship diagraph, the finalised model can then be constituted as a CLD that corresponds precisely to the concern variable or research problems.
Participation in the study was voluntary; thus, no benefits or cost was determined or given. However, there were refreshments provided for those who needed them. Interviews were also conducted at a time and place convenient for the interviewee and telephonic interviews were conducted if the participants were not available to meet up. Before the interview, the participants were informed that they could stop at any time and consent form was signed by the participants who would allow for the use of audio recording.
This research study received ethics approval from University of Cape Town’s ethics committee: ONYJOH001.
In this section, the findings of the application of the research process are presented. These findings are described below as two cycles of theory building: (1) cycle 1 draws on data from a systematic literature review and (2) cycle 2 draws on the empirical data.
A total of
Research cycle 1 served to focus on the study and provided a foundation for the theory building in cycle 2. The role of literature in cycle 1 was as a source of data rather than to provide a conceptual framework.
Research cycle 2 was built on cycle 1. In this cycle, the interview transcripts from individuals in the case study and journal entries from observations served as data. The cycle 2 concern variable or object of interest was centred on ‘level of understanding of community resilience’. From the data gathered in the research cycle 2, a total of
Core categories framed as core variables (renamed variables).
Core categories | Core variables |
---|---|
Contribution of individual resilience to community resilience | Impact of contribution of individual resilience |
Community engagement (continuous, cohesion and inclusive) | Degree of community engagement |
Positive mindset and readiness of the communities | Effect of community positive mindset |
Public education and public awareness on community resilience | Level of public education and awareness |
Ability to learn from disaster experience for community resilience | Ability to learn from disaster experience |
Transparency and accountability governance of disaster team | Degree of transparent or accountable governance |
Perceived self-efficacy to community resilience | Level of perceived self-efficacy |
Contextual understanding of disaster risk and target audience | Degree of contextual understanding of target audience |
Responsive shared leadership | Level of responsive shared leadership |
Disaster management framework on community resilience | Effectiveness of disaster management framework |
The research milestone, which is documented separately in (Online Appendix 1), shows a brief synopsis on the 60 categories generated.
The understanding gained from the cycle 2 is that in areas exposed to DR, if the resilience from the people living in these areas is understood and contributed to the community, it could contribute to improved CR, thereby attempting to reduce DR.
The structural mapping explaining the analogical reasoning used in the theory-building process application is presented in
The section below illustrates the application of the process of theory building, which was used to achieve the result of the research study. The intention for this is not for repetition, but to systematically illustrate the processes.
Through the application of grounded theory principles, 10 core categories where developed. The 10 core categories developed, and the variables re-named dynamically, are shown in
The interrelationship diagraph explains the core variables. It was developed by using the variables chosen from the core categories and mapping the direct relationships and indirect relationships (see Online Appendix 4).
The relative achievement archetype is when achievement is realised at the detriment of another sector (Wolstenholme
Wolstenholme (
[
This process involves the interaction of the causal loop feedback and the environment. This is where the structural features of the selected archetypes are recorded. This shows the structural archetype selected for placing and listing (see Online Appendix 2).
Success to successful archetype structure explains that growth in a system leads to a decrease of the next system. Anderson and Kim (
In success to the successful situation, two or more individuals, groups, projects, initiatives, systems are vying for a limited pool of space to gain success, when one of them starts to become more successful (or is historically already more successful) than the other, it tends to garner more spaces, thereby increasing the likelihood of continued success. (p. 87)
This is a process where the variables derived are inferred appropriately into the selected segments of the systems archetype in a way that their relationships are in line with the other variables inferred into the systems archetype. This should also speak to the concern variable in the understanding of the CLD.
Studies have indicated that most CLDs approximate generic CLDs (Kamprath
Archetypes are necessary for getting knowledge into the ‘nature’ of the underlying issues and for giving an initial foundation upon which a model can be further developed and established. The aim is to achieve something at the maximum level or lowest level that can be illustrated by a reinforcing loop. The archetype chosen is to control something, and that led to the decision of the loop applicable to the research. This means that the respective generic CLD was selected because it applies to the research focus.
The research variables were looked at, and the researcher considered how they would fit into the generic CLD. Anderson and Kim (
The unique design of behaviour over time for the success of the successful archetype is more of a diverging curve, where one goes up and the other goes down. This means that if the researcher represented resources A as CR and resources B as DR, it means that the success or increase in the level of CR will lead to the decrease of the DR. It also means that the CR is saying to the DR that the increase leads to your decrease. This is illustrated by the hypothesis in Online Appendix 6.
The theoretical model above explains how the behaviour of the relative achievement archetype relates to the level of understanding of CR. This explains the mechanisms that influence the level of understanding of CR in the Western Cape.
Relocated themes from the relative achievement model, which was used to correspond with the variables from the inter-relationship diagraph, are illustrated in Online Appendix 2, which recognise the themes in the success of successful CLD. These themes were assessed, and insights were given considering their effects to the other themes in the CLD. Interchangeable loops were found through the inter-relationship diagraph derived from the research process, and more insights obtained from the review of the literature were used to equal the variables of the CLD through interchangeable attributes to the generic archetype.
The equality might not be in perfect state, but the loop of the relationship remains in shape (see Online Appendix 2). This directed the researcher to go beyond the simple archetypes by embedding the variables inferred in the archetype in a system that is more complex or selecting the archetype in more depth and range of applications to actualise the theoretical model or research result required for this article (see
The compatible variables were inferred into the relative achievement archetype to develop the CLD, which is related to the concern variable of understanding of CR in the Western Cape.
The theoretical model provides an insight into how similar concepts and core variables interact with each other to develop or influence the understanding of CR. The variables are also the identified mechanisms influencing the level of resilience of the vulnerable community.
To further explain the model, a description is provided below for the four reinforcing feedback loops.
The theoretical model is made up of individual core variables interacting together as reinforcing feedback loops, which indicates the causal relationships existing in the CR model. The nature of the variables makes them difficult to measure quantitatively as they require access to the experiences of people. The dynamics of these relationships are explained next.
The four feedback loops in the model (
In the systemic structure of the CLD (model), there is a need to improve these mechanisms to influence CR; this explains the reinforcing loop 3 (R3) and reinforcing loop 4 (R4) on the model (see
As noticed, all these variables were developed systematically and are to influence the ‘level of understanding of community resilience’, which is the concern variable of this research study. Therefore, the question of the mechanisms influencing the level of understanding of CR in the vulnerable community was answered.
The answer shows that the mechanisms that influence the level of understanding of CR are the nine variables in R1 and R2. The variable ‘level of public education and public awareness’ on CR is the determinant one to improve the core mechanisms for better understanding of CR (see
There is a possibility that the theoretical model can be termed adequate regarding the real-world situations. There is also a possibility that the model was developed because of what would like to be seen and not what is happening, these possibilities were handled.
For the enrichment and maintenance of the validity, the research design was iterative. To enrich the validity of the study, the researcher took part in defining the conceptual framework from the start of the research study to provide the trustworthiness criteria of credibility and confirmability, increasing the data collection reliability, using theoretical sampling to check the upcoming ideas, using the literature review and used the different coding processes to get trustworthiness criteria of dependability.
For the transferability of the derived theory, diverse sources of data were used for the enrichment of the validity of the research study. The analogical stance was also used to apply the systems archetype to the related concepts allowing for the alignment of accepted models used to develop the theoretical model, thereby increasing the theoretical validity.
The theory generated is based on data collected in the Western Cape province, that is, in a single context. Therefore, it can serve as a hypothesis that could allow for further study to explore the transferability of the findings.
The researcher recommends that multiple sites or contexts should be used to confirm and further develop the theory. The researcher also recommends that the theoretical model could be used as a basis for understanding possible behaviour during and after disasters. The research result can be applied, as it can be applied in another context, but the audience of the research study are to make judgements of the transferability.
In this study, the modelling of the problematic behaviour was qualitative as the intention was to identify the mechanisms involved in understanding CR as perceived by relevant stakeholders. The SD models can be quantitative by defining measures and collecting appropriate data to allow to quantitative models. The researcher would recommend this as a further course of study.
This research study proposes that communities in informal settlements are affected by disasters and the risks are high. Therefore, people need to know and understand CR to be able to build it within themselves or their communities, in the context of withstanding DR (Cutter et al.
One of the world’s most advanced DR policy frameworks for DRR may be found in South Africa, but implementing them happens to be a huge challenge because DR requires an essential developmental change (SA National Disaster Management Framework
The result of this article showed that the nine variables and the variable ‘public education and awareness’ are important ways for understanding and improving the level of understanding of CR in the vulnerable community in South Africa. The study also gives an insight into the potentials of the Phola Park community towards being a resilient community. However, the community still has some valuable understanding or knowledge that still needs proper guidance and help from external bodies and the government sector.
The participatory approach used for DRR management has assisted in understanding the viewpoints of the Phola Park community members towards understanding and building resilience and providing strategies that can be used to reduce DR in the future.
Furthermore, regarding the use of Internet for data collection that could be biased when trying to validate the research credibility (Metzger, Flanagin & Zwarun
In line with the understanding of the CR model identified, the impact that it has on the nine variables and the variable ‘public education and awareness’ in the informal settlement, has been recognised. It is apparent that through the variables, an individual’s level of understanding of resilience is affected, which impacts the community’s level of developing CR.
Through the application of data credibility, data dependability and data confirmability of the research and its results, the trustworthiness of the research process was established. Ethical approval was also secured and keenly followed to avoid any harm to all participants involved.
The author thanks the University of Cape Town Knowledge CO-OP, the Habitat for Humanity South Africa, the Western Cape Disaster Management Centre, the South African Police Service (Accident Response Team/Disaster Risk Management) and the Phola Park community for their supports and contributions to this research study. The researcher would also like to appreciate the co-authors for their guidance and constant support throughout this study. The researcher wants to also express gratitude to the JAMBA journal and their reviewers for this privilege.
The authors declare that they have no financial or personal relationships that may have inappropriately influenced them in writing this article.
O.J.O. made the draft of the original manuscript; generated, analysed and interpreted the data; and made the revision of the original manuscript and final approval of the published version. T.R. and C.S. made the review of the article before submission and supervised the first author’s master’s research study.
This research study did not receive any particular grant from any funding agency in the public, commercial or not-for-profit sectors.
Data sharing is not applicable to this article as no new data were created or analysed in this study.
The views and opinions expressed in this article are those of the authors and do not necessarily reflect the official policy or position of any affiliated agency of the authors.