Vulnerability is a broad term that holds various conceptions from different disciplines and fields of study. In its disaster context, the term details the circumstances that increase the susceptibility of people to the impacts of hazards (Wu et al. 2023). These conditions are determined by physical, social, economic and environmental factors, which interact to define the coping capacity of people or a community (Singh et al. 2014). Bamweyana et al. (2020) assert that these factors are drawn from the physical environment and its socioeconomic features, as well as the political dynamics that characterise a particular geography. Within this understanding, emphasis is adequately placed on community and its capacity to prepare for, mitigate, respond and recover from the impacts of hazards.

Socioeconomic and physical environment attributions of vulnerability imply that the term is a confluence of elements; each of which is associated with features that influence the presence and kind of vulnerability that exists in a community. For instance, Delilah Roque et al. (2020) make note of biophysical vulnerability, which refers to a community’s geographic proximity to a hazard. Biophysical vulnerability confirms the presence and proximity of hazards relative to the physical location of a community, and focuses on this physical characterisation in making communities vulnerable to the impacts of hazards. For this reason, it is regarded as a function of exposure and sensitivity. This notion of vulnerability is mainly applied in research that study natural hazards in changing physical environments (see Blöschl 2022). On the other hand, Blaikie et al. (1994) reference social vulnerability, which is concerned with the socioeconomic and demographical factors that determine the degree of susceptibility a community faces to the effects of hazards. In this regard, vulnerability is explored by the implications of properties such as race, gender, education level, employment status, income and strength of social networks. This implies that population characteristics indicate the degree of vulnerability to hazards because they determine the capacities of communities to cope with the effects of hazards. Furthermore, hazards affect communities differently according to their population characteristics where those on the lower end of the socioeconomic spectrum are most susceptible to the impacts of hazards. In support, Flanagan et al. (2011) found that disaster studies have shown that socially vulnerable groups are ‘… more likely to die in a disaster event and less likely to recover after one’. Social vulnerability is thus a function of resilience.

Various conceptions of disaster have highlighted the need to view vulnerability holistically in terms of the preparation, mitigation, response and recovery of the impacts of hazards. This is on the basis that the impacts of hazards are adverse in vulnerable people – both in the biophysical and social sense. In addition, ‘… [v]ulnerability factors often occur in combination’ (Morrow 1999 in Flanagan et al. 2011:3). The chosen theoretical framework for this research (discussed below) reflects this holistic perspective.

The integrated vulnerability approach (IVA) offers a multidisciplinary perspective to frame vulnerability and its effect on hazards in society. Drawing on the tenets of the Pressure and Release Model (or Disaster Crunch Model), the IVA combines features of social vulnerability and physical vulnerability to understand the factors that subject people to the effect of hazards. The theory was conceptualised by McEntire following the recognition that existing theoretical perspectives on disaster could not encapsulate the situations in which disasters emerge from the early 2000s and in the future (Lagadec 2006). The IVA thus factors human characteristics and activities that render communities susceptible to physical events. In this regard, vulnerability is framed according to the effects of social and economic systems as well as what Boin and Hart (2006:46) refer to as ‘innocent factors’, which include socioeconomic status, gender, ethnicity and race among others. The combination of these factors merges forces that expose people, communities and organisations to the effects of hazards and the subsequent occurrence of disasters, suggesting that the factors that drive vulnerability and disasters are linked to the structural (physical) and social dynamics of a society (Buckle 2005).

According to Poteyeva et al. (2006) in McEntire et al. (2010), the multidisciplinary nature of the IVA facilitates a systematic approach that incorporates the contributions of engineering, physical sciences, social sciences and organisational management. These disciplines conceptualise vulnerability differently in relation to preventing and preparing for disasters. For instance, the physical sciences school expresses that the location of human settlements relative to their proximity to hazards exposes areas to threats of disasters (McEntire et al. 2010). Conversely, the engineering school focuses on creating ‘resilience’ through the construction and reinforcement of the built environments. This school contends that the inadequacy of the quantity, design and quality of physical infrastructure determines the level of vulnerability a settlement is subjected to (McEntire et al. 2010). For example, structures that are built with dubious construction materials and the lack of critical infrastructure face increased vulnerability and subsequent heightened exposure to hazards. The structural school, which is based on social ideals, emphasises the effects of socioeconomic and demographic factors (herein after referred to as population characteristics) in leading to hazard vulnerability (McEntire et al. 2010). Lastly, the organisation school focuses on ‘… the effectiveness of response and recovery operations, and stresses how preparedness, leadership and manage determines the ability to adapt [to] disasters’ (McEntire et al. 2010:57).

In isolation, these perspectives show clear limitations by failing to conceptualise vulnerability holistically (McEntire 2005). This supports Lewis and Mioch (2005) who argue that a single framing of vulnerability reduces the meaning of the term. Together, however, the variables expand dynamics that determine factors that: push people to settle in high risk areas (physical science school), prohibit the building of safer human settlement structures (engineering school) and consider the location of human settlements (structural school). McEntire’s IVA presents an integration of these schools by drawing on the links between the foundational conceptions of each, namely coping capacity, resilience and exposure (Manandhar & McEntire 2014; McEntire et al. 2010).

This expression of the IVA model provided the researchers a guideline to scrutinise hazard vulnerability in township informal settlements in the NMBM because it ‘… incorporates insights from the physical world emphasizes the roles of social, economic, and political relations in creating hazardous situations’ (Singh et al. 2014:75). To put differently, this approach focuses on ‘[T]he characteristics of individuals and their relationship with society in wider context, the nature of their relationships and the physical and societal environment they inhabit’ (Singh et al. 2014:75). In his development of the model, McEntire argued that the integration can be achieved by identifying the complex interactions between the types of variables beyond the structural and social causes of disasters (McEntire et al. 2010).

The term township was conceptualised by the Apartheid regime and used to identify marginalised segregated urban areas reserved for non-whites (Hunter & Posel 2012 in Membele et al. 2021). According to Williams-Bruinders and De Wit (2020), these settlements were located on the outskirts of the urban centres that were reserved for white South Africans. Empowered by the Group Areas Act of 1948, the National Party government prohibited land tenure and restricted access to urban centres for non-whites by enforcing Pass laws (colloquially termed Dompass) following the economic boom in the 1970s that necessitated the influx of cheap black migrant labour into urban centres (Lemon 2021; Membele et al. 2021; Webb 2023; Williams-Bruinders & De Wit 2020).

According to Barry and Rüther (2005), the apartheid system during this time allowed for the infux of black migrants in informal settlements across urban centre. However, conflicts arose between local and national authorities as this influx contradicted racial segregation. Therefore, cities were pressured by national authorities to provide housing that would conform to the segregatory national laws and policies that maintained the spatial status quo. This was particularly pertinent in cities such as Port Elizabeth that were mostly composed of white and coloured populations and subsequently expect to receive considerable influxes of black migrant labourers. To respond to the pressures, local governments provided small matchbox-type housing on the outskirts of urban towns and cities (Williams-Bruinders & De Wit 2020). Interestingly, the development of these settlements were proposed to alleviate the socioeconomic and health issues that affected unserviced urban plots that migrants had initially occupied (DeMoss 2012). Among these were better sanitation and larger properties – aspects that served to entice black migrants into voluntarily moving to these areas. However, as Monama et al. (2022) and Modise (2019) highlights, townships were and remain organised as overpopulated clustered dormitory settlements that are poorly serviced by the government. Consequently, these settlements became spaces for social challenges such as health, poverty and unemployment (Moghayedi et al. 2023). These issues have continued despite the democratic dispensation in 1994, with socio-political and socioeconomic dynamics thrusting further ills such as high crime rates, drug abuse, HIV/AIDS, youth unemployment and the propagation of informal settlements. This is in the wake of townships remaining densely populated squalors that subject people to precarious living conditions.

The notion of human settlements is important and serves as a critical variable to the context of townships amid modern-day dynamics of urbanisation that shape the physical and social landscapes of these geographies. Arguably, the most poignant factor is the cheap cost of living in townships and how this attracts migrants and immigrants (Malgas & Zondi 2020; Scheba & Turok 2020). Abrahams and Everratt (2019:256) argue that mass influxes into townships because of its affordability appeal extends capacities of the regulated formal housing market that is often unable to cater for the housing needs of low-skilled and economically deprived segment of work-seekers who constitute the largest number of migrants. In part, formal housing is simply unaffordable for this demographic who are often compelled to seek cheaper and unregulated alternatives. In townships, this comes in the form of informal settlements.

Informal settlements are subject to much debate owing to the interchangeable use of the term with conceptions such as slums, shanty towns, squatter settlements and unplanned towns (Abunyewah et al. 2017:239). Although these interchangeable terms seem to apply to different contexts, their central precept is the illegal occupation of vacant land. Moreover, this land is often outside land-use schemes and without planning permission. Satterthwaite et al. (2020) clarify that informal settlements are located on land that has not been demarcated for human settlements whereby spatial planning and environmental assessments have been conducted to test their suitability and integration into existing infrastructure and socioeconomic systems. As a result, they lack basic infrastructure and municipal services such as water and sanitation, stormwater infrastructure and electricity reticulation among others (Nyashanu et al. 2020; Wiltgen et al. 2021).

The literature notes that informal settlements globally are frequently located close to biophysical hazards such as industrial zones and flood plains that, while posing health and safety threats, offer employment opportunities, informal business activities and access to social resources (Williams et al. 2019). For instance, Satterthwaite et al. (2020) postulate that the illegal occupation of land for shack dwellings is driven by economic motives whereby people want to be located in an area that is close to employment opportunities and social infrastructure. Furthermore, informal settlements are dense, clustered and overcrowded settlements that consist of unstable makeshift houses built with unconventional building materials. These characteristics render these settlements vulnerable to hazards. Accordingly, low-income communities experience vulnerability that encompass geographic, socio-economic and attitudinal factors.

The establishment of new informal settlements, expansion of existing freestanding informal settlements and proliferation of informal backyard housing structures in townships indicate that informal settlements have spiralled beyond municipal control. For instance, informal households have grown from 1 170 902 in 1995 to 1 294 904 in 2011 and 1 300 000 in 2016 (Nkonki-Mandleni et al cited in Membele et al. 2021). According to Jeffery (2010), the increase in the number of informal housing has also coincided with more dense informal settlements.

According to Ramunenyiwa (2020), municipal by-laws aim to regulate the development of human settlements in order to promote effective spatial planning and integrated service provision that factor the risks posed on human health, property and the environment. In addition, the Prevention of Illegal Eviction from and Unlawful Occupation of Land Act 1998 seeks to de-escalate the frequency of informal settlements. Regrettably, the growth of informal settlements negatively points to the non-adherence of these by-laws. This is supported by empirical evidence presented by Mabaso (2019:117) who found that communities in the Umlazi Township disregarded spatial planning considerations by illegally settling on vacant lands. This practice was also established by UN Habitat (2015 in Sinharoy et al. 2019), which described informal settlements as residential settlements that have been constructed on land that is occupied illegally.

The negative characterisation of townships is indicative of the vulnerabilities that have rendered informal settlements in these geographies susceptible to floods. According to Hendricks and Van Zandt (2021), the susceptibility to hazards in townships is a consequence of the various social, physical and economic conditions that define the social and physical vulnerabilities of these geographies. Empirical studies have highlighted the consequences of the mentioned factors in causing disasters. For instance, a study by Mawasha and Britz (2021) using GIS methodologies ascertained that the vulnerability to flooding in the Alexandra Township is because of the proximity of housing (mostly informal) to the floodplains of the Juskei River. Napier and Rubin (2002) elucidated that people in Alexandra Township construct their houses on steep slopes that are susceptible to floods and landslides. Moreover, Napier and Rubin (2002:5) posited that the ‘… rising damp, poor indoor air quality and collapsing structures’ exposes these settlements to the disaster risks. In the aforementioned case studies, flood vulnerability arises from the nature of informal settlements themselves.

The social dynamics of urbanisation have influenced the establishment and mushrooming of informal settlements concentrated in the townships. Considering the gentrification of the housing market in the municipality, poor migrants are excluded from participation (Mzileni 2021). This has perpetuated inaccessibility to formal housing for migrants and illegal immigrants who cannot afford formal accommodation. According to Williams-Bruinders and De Wit (2020), most migrants in the municipality survive on one dollar a day and therefore have to make use of alternative housing in the form of informal dwellings on vacant plots usually on the edge of cities and towns around the municipality. This practice contributes towards the ‘… uncontrolled growth of informal settlements flowing from unlawful land invasions’ (NMBM 2023b:201). Thus informal settlements have risen steadily in the municipality from about 76 in 2017 to 156 in 2023 (NMBM 2023b).

Urban informality is however a global phenomenon. The United Nations (2014) in Abunyewah et al. (2017) estimates that nearly one billion people live in informal settlements around the world. With projections estimating that the global population of people residing in cities will increase to 66% by 2050 (United Nations 2014), the number of informal settlements in urban geographies such as townships in the NMBM are likely to increase drastically. The effects of urban migration are derived according to the existing socioeconomic context of geographies. In the NMBM, these effects continue to follow the format of developing societies, whereby the influx of migrants and immigrants limits public service capacity and creates imbalances in the delivery of key social and economic development areas. These imbalances are clear from the enormous service delivery capacity challenges demonstrated by backlogs in the provision of housing, water, sanitation, refuse removal, tarring road and stormwater drainage as well as managing ageing infrastructure – all of which disproportionally affect townships and informal settlements in the municipality (NMBM 2023a).

The socioeconomic context of the municipality highlights that townships and informal settlements, in particular, face vulnerability in various forms. Much of this can be attributed to the legacies of Apartheid that saw townships inherit dilapidating infrastructure and weakened social and economic conditions (Wasserman et al. 2022). Progressively, studies have also accentuated that political factors in the form of disparity in the quality of services delivered in townships (particularly in informal settlements) and those delivered in residents formerly reserved for white people also widen the social vulnerability in townships. As a largely migrant population surviving on less than one US Dollar a day, informal settlement dwellers in the municipality rely significantly on the services of government to provide vulnerability-reducing goods and services. Therefore the failures of government capacity is sure to negatively impact their coping capacities in an environment with clear and evolving vulnerabilities.

In the last 20 years, the municipality has faced invariable climatic conditions that have led to occasional heavy rainfall caused by thunderstorms and coastal weather systems (NMBM 2023b). The frequency of invariable climatic conditions has coincided with the increase in floods (NMBM 2023b) signifying that floods in the municipality are largely initiated by heavy rainfall. Historical disaster indexes of the municipality indicate that the municipality has experienced an increase in floods over the last four decades (NMBM 2023a). In addition, the disaster risk assessment of the municipality ranked floods as the highest disaster risk (SRK Consultants 2010). Drawing on the IVA, the following sections look into the vulnerability factors that advance heavy rainfall into flood risk within informal settlements in the townships.

Interestingly, the exposure to flood hazards is knowingly accepted by people in several informal settlements in the municipality’s townships. For instance, Siyonqwana et al. (2015) found that residents of a Missionvale Informal Settlement are well aware of the risk to floods but prefer the location because of its proximity to employment opportunities, health and educational facilities. This is also the case in an informal settlement in the KwaZakhele township, as found by Mngomezulu (2020). Evidently, people tend to prioritise the socioeconomic advantages presented by their location even when they have adequate knowledge of the danger presented by their occupation of that location. This finding is consistent with Huchzermeyer et al. (2013) and Satterthwaite et al. (2020) who cited that the opportunity to be close to employment opportunities, business activities and education institution is enough for people to live under the risk of floods.

Intriguingly in the present study, the ‘voluntarily hazard exposure’ of informal settlement dwellers amid the risk of floods is also influenced by a ‘sense of place’ or social attachment to location. According to DeMoss (2012:170), some former residents of an informal settlement in the Walmer Township (resettled in Wells Estate by urban authorities) ‘… felt that they have lost a sense of control compared to what they had in their former shacks’. This finding speaks to a social attachment to a location that would explain the reversion to old settlement plots after people have been resettled to safer sites (Lupindo 2015). Interestingly, this place attachment exist despite continued incursions of financial and psychosocial costs caused by floods in the studied areas. This voluntary exposure to flood hazard can also be explained by the fear of losing property by informal settlement dwellers. For example, Van Zyl (2023) and Dangazele (2022) found, in separate instances, that in an informal settlement in the Wells Estate located near the Markman Canal and prone to flooding, residents opted to subject themselves to the risk of floods by opposing their relocation to safer sites by local authorities. For these residents, the fear rested on losing their belonging, homes and plots once they leave for safer sites. On the other hand, in the Missionvale Township, people shared that it was not easy to move from the area because they felt that area is home and aligns with their cultural values (Siyonqwana, Heijne & Tele 2015).

These findings present an interesting prospect in which a ‘sense of place’ (whether derived from sentimentality or social and economic locational advantage) subjugates informal settlement residents to flood vulnerability. Napier and Rubin (2002) also offered a compelling argument by explaining that residing in these locations is sometimes a strategy employed by informal settlement dwellers to access government housing benefits quicker. This claim is, however, not supported in the evidence of this study. However, the notion of ‘sense of place’ remains a pertinent factor of hazard exposure in the literature as it is in the present study. For instance, covering the 2000 floods in the Alexandra Township, Batha (2000) uncovered that people refused to move from the banks of the Juskei River despite floods destroying their property. Their refusal, consistently with the present study’s findings, was because they feared losing their belongings and plots. Reflecting on the Alexandra Township floods, Napier and Rubin (2002) claimed that many of these people returned to the banks of the Juskei River to rebuild their homes after the water levels subsided. In the case where people did not return, they were replaced by other families that moved in.

Engineering perspectives highlight the inefficiencies of the built environment to withstand against the impacts of heavy rain in the municipality’s townships. Reference to the built environment is made on the structural integrity of human settlements and the critical municipal infrastructure meant to hold and network heavy rain during and after storms.