1 Introduction

People across the Majority WorldFootnote 1 are disproportionally affected by climate change and are struggling to adapt and sustain their livelihoods (IPCC 2022). There are existing and emerging limits to climate adaptation, as adaptive responses are unavailable or adaptation may become infeasible (Mechler et al. 2020). Consequently, climate impacts can be 'unavoided' or 'unavoidable' (Mechler and Deubelli 2021). These residual impacts are termed Loss and Damage (L&D) (Boyd et al. 2021). From increasing heatwaves to intense rainfall, various climate-related hazards can result in L&D, which affects various aspects of life that people value. L&D can be 'economic', such as loss of income and assets important to livelihood options, and 'non-economic', such as loss of cultural practises or elements of biodiversity (Boyd et al. 2021). This paper focuses on ‘non-economic’ L&D (NELD), which refers to impacts on aspects of life that are not typically traded in markets (UNFCCC 2013).

An increase in NELD is fundamentally a human rights issue, as people are denied, by climate change impacts or responses to risks, the right to essential aspects of their lives, such as food, health, and culture (McNamara et al. 2023). Therefore, large-scale efforts to address NELD are urgently needed. However, addressing and financing L&D remains contentious. Only recently, increased pressure from Majority World countries contributed to the establishment of a dedicated Fund for responding to Loss and Damage (Wyns 2023)—a potentially significant milestone for mobilising resources to address L&D. Now that progress is being made in the complex political and ethical debate about who should pay for L&D and to whom, it is increasingly clear that we need to improve our understanding of how to address L&D (Warner and Weisberg 2023).

At UNFCCC meetings following the 27th Conference of the Parties, countries repeatedly highlighted a gap in action and knowledge on how to address NELD (UNFCCC 2023a; 2023b). Indeed, the L&D literature primarily focuses on economic L&D, and there is limited understanding of how to address NELD (Jackson et al. 2022). Serdeczny and Lissner (2023, p. 1) argue that “more empirical research addressing how affected communities can respond to non-economic losses is needed to complement economic loss estimates and inform related funding needs”. Improving knowledge on how to address NELD is particularly urgent for people living in Majority World countries, as they “typically have a larger portion of their needs met “non-economically” (Preston 2017, p. 144). This includes Bangladesh, a Least Developed Country that will likely be supported when the Fund for responding to Loss and Damage becomes operational. Climate change has increased the probability, scale, and intensity of climatic hazards, and contributes to new hazards in Bangladesh (IPCC 2022). In response, the Government of Bangladesh (GoB) and non-governmental organisations (NGOs) are increasing their efforts to respond to—predominantly economic—L&D (GoB 2021). Furthermore, historically, coping with floods and cyclones has been a way of life in Bangladesh (Ahmed 1999; Islam and Peterson 2008) and the people of Bangladesh have autonomously developed responses to cope with the impacts of different climate-related hazards (Fenton et al. 2017; Rahman and Hickey 2019; Younus and Harvey 2014).

However, these autonomous efforts are still largely overlooked in L&D discourse. Therefore, we herewith share knowledge on how to address NELD by (1) taking stock of the NELD faced by people in Shyamnagar and Durgapur UpazilaFootnote 2 in Bangladesh; (2) documenting the existing autonomous responses to these impacts; and (3) identifying the limitations in people’s capacity to respond and the barriers to further action. We first present a review of the literature on NELD, autonomous responses and adaptive capacity. We then outline the methodology employed, introduce the study sites, and present the relevant contexts. This is followed by an empirical section in which we explore the NELD experienced by people and the subsequent autonomous responses. The final sections are a discussion that includes policy priorities and avenues for further research, and a conclusion.

2 Understanding responses to NELD

In 2013, a technical report commissioned by the COP on the conceptualisation of 'non-economic loss’Footnote 3 was published (UNFCCC 2013). The dichotomisation of 'economic' and 'non-economic' impacts and other conceptualisations made in the report significantly influenced subsequent L&D studies. We first examine these studies to understand better how NELD is currently assessed and can be addressed. Then, we consider what autonomous responses to NELD look like and what determines these responses.

2.1 Assessing NELD

Researchers mostly assess NELD using qualitative approaches, with most studies focusing on South Asia and the Pacific Islands (Bahinipati 2020; Islam et al. 2022; McNamara et al. 2021; Nand et al. 2023; van Schie et al. 2023b; Westoby et al. 2022). The UNFCCC (2013) report identified main types of non-economic loss as: loss of life, health, human mobility, territory, cultural heritage, Indigenous knowledge, biodiversity, and ecosystem services. Studies assessing or further conceptualising NELD often follow this typology or propose pre-defined typologies similar to these ‘main types’. Various assessments have revealed a wide range of NELD, highly dependent on socio-cultural contexts and influenced by intersecting factors such as gender, class, and race (Hayward and Ayeb-Karlsson 2021; Kronsell and Kaijser 2014). Boyd et al. (2021, p. 1336) even suggest that this strong dependence on beliefs and worldviews makes NELD “potentially infinite”. The context-dependant nature of NELD means that standardised and pre-defined types of non-economic loss cannot fully capture the L&D different people experience (Nand et al. 2023). Furthermore, while the ‘main types’ were suggested to show how NELD materialise, "The distinction between non-economic loss and economic loss will sometimes be blurred” (UNFCCC 2013, p. 4). For example, as specific items may hold ‘economic’ and ‘non-economic’ value to people’s lives or as ‘economic’ L&D can cascade into NELD (Pill 2022). Therefore, pre-defined types must be critically examined to avoid misalignment between local experiences and assessments or policies (Nand et al. 2023).

2.2 Ways to address NELD

To some extent, policies and interventions relevant to addressing aspects of NELD already exist. Humanitarian aid includes the provision of post-disaster health care and psychosocial support. But environmental and cultural impacts are typically not addressed and humanitarian aid generally only provides short-term relief while NELD will continue to occur as climate change intensifies (Knox Clarke and Hillier 2023). Climate change adaptation (CCA) and disaster risk reduction (DRR) efforts can focus more on longer-term needs. CCA and DRR literature have explored aspects of NELD, such as cultural heritage and physical and (mental) health (Aitsi-Selmi et al. 2015). Yet, these studies typically treat NELD from a sectoral perspective and take a siloed approach (e.g. mental health or cultural heritage) that fails to adequately capture the inherent interlinkages and cascading nature of NELD (Aitsi-Selmi et al. 2015; Pearson et al. 2023). Furthermore, while ‘non-economic’ aspects of life are accounted for in these discourses, they can be treated as a potential source of risk reduction and less as a value that is at risk of being lost from climatic and environmental change (Pearson et al. 2023). As such, NELD helps researchers and policymakers coalesce around a wide range of values at risk of being lost or damaged in ways that DRR and CCA have yet to do.

The limitation of extant measures that alleviate or help recovery from often intangible and subjective consequences of climate change has led to explorations on how to address NELD. Some scholars argue that acknowledging sustained losses and responsibility for them—as well as apologies from those responsible for them—can provide recognition to those affected (Serdeczny et al. 2016; Morrissey and Oliver-Smith 2013). However, this does not directly restore livelihoods. McNamara et al. (2021) analysed literature concerning Indigenous groups, disasters, and peace studies to understand the usefulness of healing and coping strategies (e.g. education and documenting traditional knowledge) for addressing NELD in the Pacific Islands. Other specific recommendations from studies include land reforms (Nand et al. 2023), increasing psychosocial support (Hayward and Ayeb-Karlsson 2021), and introducing locally-led projects to improve social awareness, cultural preservation, and social cohesion (Chandra et al. 2023). However, most studies refrain from prescribing tangible interventions to address NELD. Instead, they argue that more research is needed to understand NELD (Chiba et al. 2017; Hirsch et al. 2017), that it should be integrated into existing response strategies (Islam et al. 2022; Pill 2022), and that new response strategies or policies should be developed (Andrei et al. 2014; Thomas and Benjamin 2020).

2.3 Autonomous responses

In addition to external interventions, people also respond autonomously to different climate stressors. Despite some evidence on autonomous responses to NELD, such as addressing health issues (Sen 2023) or restoring ecosystems (Pecl et al. 2019), accounts of economic livelihood restoration predominate (Fenton et al. 2017; Forsyth and Evans 2013; Malik et al. 2010; Rahman and Hickey 2017, Warner et al. 2012). Autonomous responses are not always explicit strategies, they can also be gradual or spontaneous changes that people make in their daily lives (IPCC 2022). Autonomous responses can and should inform decision-making to enhance the efficacy of external interventions and reduce the risk of maladaptation (Malik et al. 2010; Rahman and Hickey 2019). Indeed, external interventions, such as emergency food aid or externally defined development, can often undermine existing capabilities and cultural responses (Audet 2015; Jackson 2020).

However, autonomous responses are seen as falling short in minimising or addressing current L&D, which is set to increase exponentially (Younus and Harvey 2014). In the context of L&D, Warner et al. (2012) make a helpful distinction between coping and adapting, with coping defined as “short-term responses to the impacts of sudden events”, and adapting as “longer-term responses to more gradual changes” (Warner et al. 2012, p. 23). In addition to only incrementally mitigating adverse impacts in the short term, coping responses can further erode livelihoods by undermining the resources of people who already have little capacity to respond (van der Geest and Dietz 2004). Furthermore, evidence shows how socio-economic challenges, such as poverty or social inequality, can constrain people’s ability to respond (Wangui and Smucker 2018). Consequently, “existing class and ethnic hierarchies that trap the poor, powerless and displaced” can be deepened (Sovacool 2018, p. 184). Scholars argue that well-planned and executed external interventions can strengthen people’s adaptive capacity to enhance the efficacy of autonomous responses and prevent erosive coping (Malik et al. 2010; Rahman and Hickey 2019; Rahman et al. 2023; Warner et al. 2012).

2.4 Capacities

The IPCC (2022, p. 2899) defined adaptive capacity as “The ability of systems, institutions, humans and other organisms to adjust to potential damage, to take advantage of opportunities or to respond to consequences”. A critique of adaptive capacity stems from the concept being more focused on people’s individual and collective dispositions instead of on structural determinants of vulnerability. Indeed, Grove (2014, p. 244) observed that a neoliberal assumption is that:

Adaptive capacity is no longer something limited by structural constraints such as race, class, or gender inequalities, it now depends on individuals’ psychological dispositions and the wider cultural belief systems that affect their perceptions of self-efficacy.

We reject such interpretations and recognise the agential and structural factors can inhibit subjects’ personal and collective capacities to respond to climate change. In a literature review, Siders (2019) identified 158 determinants for assessing adaptive capacity. This wide range of determinants points to the ambiguity of the term and how people's ability to respond to climate-related hazards depends on virtually all aspects of life. Beyond highly individual factors (e.g. age or the number of children), many of the determinants uncovered by Siders (2019) are fundamentally linked to the wider political economy, such as access to healthcare or law. Therefore, policies strengthening adaptive capacities must go beyond individual local-level efforts and dispositional factors to make a difference. Instead, as vulnerability is driven by the confluence of multi-scale political-economic, cultural, and environmental factors (e.g. uneven development and colonial capitalism), scholars argue that multi-scalar transformations of societies addressing the root causes of vulnerability are needed (Dorkenoo et al. 2022; Jackson et al. 2023). Increasing people’s adaptive capacities means challenging processes that lead to vulnerability, marginalisation, and, fundamentally, a lack of choices (Kelman 2020).

3 Materials and methods

This research was conducted in two unionsFootnote 4 in Shyamnagar Upazila and three unions in Durgapur Upazila (Fig. 1). The studies were carried out independently but share methodological features, making synthesising them useful.

Fig. 1
figure 1

A map displaying the location of the assessed unions and upazilas within Bangladesh

Full size image

3.1 Methods

In both upazilas, the methodology involved (1) establishing typologies for assessing NELD, (2) identifying the most prominent climate-related hazards, (3) assessing NELD from these hazards using these typologies, and (4) assessing the responses formulated by affected individuals or groups in response to NELD (see Fig. 2).

Fig. 2
figure 2

Schematic representation of the methods used in Shyamnagar and Durgapur Upazila (FGDs = Focus Group Discussions; SSIs = Semi-Structured Interviews)

Full size image

In Shyamnagar, semi-structured interviews (SSI) were held with 16 participants. Participants were asked to rate the impact of seven pre-identified climate-related hazards on a five-point scale. The participants were asked about the L&D resulting from these hazards using ten pre-defined NELD types: ecosystem services and biodiversity, human life, physical health, mental and emotional well-being, territory, culture and practices, Indigenous and local knowledge, social fabric, and education. We derived these from existing NELD literature (see van Schie et al. 2022). We then asked participants about their responses to this L&D. Finally, we organised four focus group discussions (FGDs) to validate and expand earlier findings.

In Durgapur, climate-related hazards were identified in six FGDs. After, 101 participants ranked the perceived impacts of the identified hazards on a five-point scale. We took a bottom-up approach in Durgapur, identifying ten 'local values' during SSIs and FGDs to ensure that our assessment considered what people valued most (see van Schie et al. 2023a; 2023b). The ten identified values—nature, health, mental health, serenity, culture, religion, society, family, education, and development—mostly reflected aspects of life commonly classified as ‘non-economic’. These values guided 32 SSIs, in which we asked participants how the climate-related hazards they identified impacted each value. In another 32 SSIs, we explored autonomous responses to these impacts.

In both studies, survey and SSI participants were purposively selected to provide depth of understanding and diverse perspectives. Selection criteria included vulnerability to climate-related hazards, demographic representativeness (e.g., religion, gender, and age), and the inclusion of ethnic minorities. The saturation principle was used to determine the sample size (Saunders et al. 2018). However, in Shyamnagar, limited time and resources were a factor. The FGDs were organised with the assistance of community leaders in the different unions who served as gatekeepers and made the research possible. The prior selection criteria were also used here to ensure the inclusion of diverse perspectives. Furthermore, separate discussions were organised according to gender to enable secure discussions of potentially stigmatising issues.

All interviews and discussions were recorded, transcribed, and coded. In Durgapur, the data from the SSIs and FGDs to define the values and climate-related hazards were coded inductively. All other interviews and discussions were coded deductively following the typologies. Although the typologies are different, there is overlap between the themes arising in each study. We have analysed only six themes here that appear in both assessments in order to present a robust, multi-sited case study. These themes are presented below, along with aspects of the lives of participants in Durgapur value that connect to the themes to illustrate what they may signify to Bangladeshi society (Table 1). It should be noted that some of these themes reflect aspects of life that are traditionally perceived as ‘economic’. However, ‘economic’ activities can also have ‘non-economic’ significance (Morrissey and Oliver-Smith 2013). For example, agriculture can bring economic security and be instrumental to culture and traditions.

Table 1 The six themes representing NELD used in this study and the aspects of life people in Durgapur associated to these themes
Full size table

3.2 Background and study site

Poverty rates are high in Bangladesh and education is seen as a way out of intergenerational poverty (Kabeer and Mahmud 2009; UNDP 2022). However, there is a mismatch between education and job availability for graduates. Consequently, many remain unemployed or settle for low-income jobs (Dewan 2021). This is also the case in Shyamnagar and Durgapur, where most people depend on crop cultivation, livestock rearing and freshwater fish farming, which often pays little (Rahman and Akter 2014). In addition, transport and wage employment, such as working in sand mining, agriculture, or shrimp farming, are common livelihood options. Gender inequality is pronounced in Bangladesh, especially in remote areas, due to patriarchal norms, structural inequalities, lack of decision-making power, unequal control of resources, burden of unpaid care work and minimal financial independence (Farole et al. 2017).

Bangladeshi society is predominantly Muslim (91% of the total population) and Hindu (7.9% of the total population) (BBS 2023). Shyamnagar is also home to Mundas, who are typically lower-caste Hindus but have distinct social and religious structures. Durgapur is home to various Adivasi (Indigenous) groups, such as Mandis, a matrilineal society, most of whom have converted to Christianity, and Hajongs, who follow Hinduism and are descended from Kshatriyas, who are second highest in the caste system (Gain 2011). Dalits are ranked lowest in the caste system and, like others from lower castes, are marginalised in Bangladeshi society (Shrivastava and Tanchangya 2015). For example, Dalits in Durgapur reported that they receive minimal state support: “As we are lower-caste Hindu, the chairman and Union Parishad members do not care about us” (Interview 76, Durgapur).

In Shyamnagar, sea-level rise contributes to increased salinity levels in soil and groundwater. Tropical cyclones make landfall in Bangladesh every three years on average; cyclones Sidr (in 2007), Aila (in 2009) and Amphan (in 2020) were particularly devastating (Amin and Shammin 2022). These cyclones can bring storm surges that cause severe inundation. River flooding is common in both upazilas, with a devastating flood affecting Durgapur Upazila in June 2022 (UNICEF 2022). These rapid-onset events are accompanied by slow-onset hazards, as the intensity and prevalence of riverbank erosion, erratic rainfall, and droughts are increasing in both regions (IPCC 2022). Over the last century, Bangladesh has made significant progress in reducing deaths and injuries, especially from rapid-onset climate-related disasters, and the government is presenting itself as a role model for CCA (GoB 2021).

Nonetheless, people in Bangladesh still face significant L&D (Islam et al. 2022; Hayward and Ayeb-Karlsson 2021). In response, different ministries within the GoB are responding to climate impacts within their sector (GoB 2021). NGOs are also active—especially in the coastal areas of Bangladesh—and increasingly focus on climate change in their programmes, presumably as donors have shifted their focus (Dewan 2021). These responses and large-scale development projects can cause further harm to people. For example, the introduction and intensification of saltwater shrimp farming in the southern coastal regions, promoted by the government and international agencies as a livelihood option in the face of increasing salinity levels, have further raised salinity levels (Dewan 2021). However, shrimp farms have turned out to be highly extractive, capitalist ventures that drive economic growth and benefit elites at the expense of the well-being of the majority of local people (Paul and Vogl 2011). Although income earning opportunities have risen, they are not without negative social and environmental issues (Paprocki 2021). In Durgapur, the government facilitates large-scale sand mining from the Someshwari River. Again, this enterprise contributes to Bangladesh's economic growth and enables the construction of roads and bridges in the country. However, local people have observed that sand mining alters river flows and increases bank erosion (van Schie et al. 2023a). In both upazilas, people rely on embankments built by the government to increase protection and enable intensive agriculture (Paprocki 2021). But poor maintenance leads to further waterlogging and creates a false sense of security for the people who depend on them (Dewan 2021; Paprocki 2021).

4 Results

In Shyamnagar, participants reported that cyclones, riverbank erosion and flooding were the highest impact hazards. In a study by Ground Truth Solutions (2023), they asked 283 participants in the upazila to rank the severity of various climate-related hazards, resulting in a similar list (from most to least impactful): riverbank erosion, cyclones, flooding, storm surges, salinity intrusion, erratic rainfall and drought. In Durgapur, participants identified flooding, drought and riverbank erosion as the most impactful hazards. We now examine how these hazards have led to NELD and how affected people responded to these impacts in both upazilas. These are also summarised in Table 2.

Table 2 Summary of the recorded L&D and subsequent autonomous responses in Shyamnagar and Durgapur Upazila
Full size table

4.1 Ecosystem services and biodiversity

The growth of trees, plants, and crops in both upazilas is affected by various hazards. For example, floods destroy plants and disrupt agricultural cycles, droughts cause water stress to crops and vegetation, and increased soil salinity in Shyamnagar further hinders plant growth. These impacts cascade into several other impacts. For example, natural products are essential to various cultural practices, and food security depends on home-grown fruits and vegetables. Therefore, cultural practices and food security are adversely affected if biodiversity or agriculture is impacted.

In response to the loss of trees, participants indicated that they cut down fewer trees and people in both upazilas are replanting trees and plants based on local understanding of the cooling effect of green spaces and the need for natural shade: "After [cyclone] Aila, there were no trees here. Whatever you are looking at, we regrew them" (Interview 11, Shyamnagar). Trees are replanted from seeds found along the riverbanks or bought from bazaars. Participants showed us how they use nets to protect plants and trees from animals. However, the success rate of replanting is relatively low due to hazards such as rising salinity and drought. Consequently, if financial resources permit, people buy previously home-grown fruits, flowers, and vegetables at shops. They also use increased amounts of (chemical) fertilisers to enhance tree or plant growth. Moreover, wildlife, particularly bird species, are declining. This is believed to be due to the noise and disturbance from sand dredging in Durgapur and to other factors eroding biodiversity in both upazilas. Some participants reported that they had planted trees and reduced cutting down trees to bring back wildlife, but no widespread response was observed.

4.2 Culture and religion

Riverbank erosion, cyclones and floods have caused the loss of religious buildings in both upazilas. To this, no response was recorded. Religious buildings also become inaccessible when areas are flooded, limiting different people’s ability to practice their religion. Muslims are not always able to wash during periods of flooding or water scarcity, inhibiting them from praying. In this case, performing Tayammum (ablution with clean soil or dust) is possible, but participants rarely mentioned this. Reduced crop yields are problematic for Mandis, who use rice for various celebrations. Mundas raise hens of specific colours for sacrifice but struggle to do so in adverse conditions. Hindus and Hajongs cannot always grow the fruits and flowers used during puja. Again, people buy these items from markets instead of growing and raising them themselves or reduce their use in response.

As agriculture and animal husbandry have cultural and historical significance in Bangladesh, L&D to these aspects of people's lives affects their culture: “We cannot grow crops here [due to the floods]. Talking about agriculture to the new generation is like telling a story” (Interview 16, Shyamnagar). Similarly, a decline in livestock has led to the disappearance of the traditional rakhal (cattle keeper) job as “people had 20 to 25 cows around their houses. Now, everyone has one or two cows” (Interview 55, Durgapur). No response was recorded in response to this loss. Reduced financial security following climate-related hazards inhibits people from organising religious festivals and rituals in the upazilas; Muslims cannot always afford to sacrifice cows during Eid ul-Adha, and Hindus indicated that they did not have enough money to organise Durga Puja: "Our biggest festival is Durga Puja, for which we would give 2000 taka per family. This year we could not arrange it and, instead, held another Puja which only cost 200 Taka” (Interview 74, Durgapur).

In Shyamnagar, participants reported that families were less likely to allow their daughter to marry someone from a disaster-prone village because it meant relocating their child to an unsafe area: "How could I give my daughter to someone from Gabura Union? She could stay on the embankment road for three months," noted one woman, referring to the length of time people spend on the embankments post floods (FGD2, Shyamnagar). Men from Gabura Union, a region particularly prone to climate-related hazards, now relocate during the marriage process or marry someone from a lower income family.

4.3 Social fabric

Participants shared that social ties can strengthen during and after disasters as people repair homes, share resources, and provide shelter: “We help and cooperate with our neighbours. Not everyone’s crops are damaged to the same degree… …those losing fewer crops help the ones who face major losses” (Interview 99, Durgapur)”. However, people also spoke of a decline in social ties. In Shyamnagar, people indicated that widespread poverty made people 'individualistic' and 'heartless': "When poverty comes in the door, love flies out the window" (FGD 4, Shyamnagar). In both upazilas, people are busier with work and have less time to socialise due to the financial insecurity caused by climate-related disasters. In addition, hospitality decreases due to resource constraints, households cannot share vegetables due to reduced crop yields, and conflicts increase. Climate-related disasters can also separate families, communities, and households. For example, people reported how family members moved to the city following climate-related hazards to find work and recover lost income, and a significant part of a Hindu community in Durgapur left the area after their temple was lost to riverbank erosion. Consequently, the remaining Hindu households were a small minority in the village, and their Muslim neighbours gradually took over their fruit trees and land. Few autonomous responses to L&D were observed in terms of social cohesion. However, people are trying to resolve conflicts, inviting each other to their homes and calling those who have moved away, sometimes borrowing a smartphone to make video calls.

4.4 Education

Floods and cyclones can temporarily make schools inaccessible and disrupt education, sometimes for months. School supplies, such as books or stationery, can be destroyed during these hazards. In Durgapur, students reported that extreme temperatures, which are increasing, are affecting their ability to learn. In addition, financial insecurity following climate-related hazards can mean that parents cannot pay school fees for their children. In Shyamnagar, several participants noted that when schools are inaccessible, parents and local teachers tutor students at home, or students’ study independently at home. However, education is not always a priority: "When there is a flood, we cannot move and cook on the bed. There is no space to think about education" (FGD 1, Shyamnagar). In addition, home education is not always possible as parents are often uneducated in modern curricula. School buildings can be reached by boat during minor floods, and participants from Durgapur said they had bought new copies of books they had lost.

4.5 Physical health

Increasingly erratic rainfall and droughts deplete groundwater sources, leading to water shortages. In addition, floods inundate water sources such as ponds or pumps, making them unusable. In response, when possible, people obtain water from other sources, such as other households or the river. Similarly, various hazards affect crop and vegetable production, contributing to food insecurity: "We don't get fresh vegetables these days. Now we have to buy them. They don't grow well anymore" (Interview 57, Durgapur). Access to food can be completely cut off during and after floods. In response, people eat dry food that they have stored, receive meals from others, cook at neighbours' houses, or skip meals. Floods also prevent people from bathing and using latrines, undermining hygiene and cleanliness: “I had to urinate in the stagnant water, as there was no option for a bathroom. The water was up to the knee level, even at the mosque where we took shelter” (Interview 82, Durgapur). This is particularly problematic for women, who face disproportionate harassment and gynaecological problems. People use neighbours' or family members' toilets if possible.

Participants noted an increased incidence of many illnesses and attributed them to various hazards, mainly flooding, salinity intrusion and high temperatures. Commonly mentioned symptoms were rashes, fever, itching, cold, flu, diarrhoea, abdominal pain, headache, fatigue, cough, and increased blood pressure. Besides this, women experience increased gynaecological problems, particularly in Shyamnagar, where exposure to saline water causes discomfort and infections. In response to physical health problems, people treat wounds, buy modern medicines from local pharmacists, visit local doctors and hospitals, use herbal medicines, and visit Kobiraj (traditional herbal practitioners). However, not everyone has access to healthcare due to high costs and limited facilities. In particular, access to female doctors is limited. Women often feel uncomfortable with male doctors, especially for gynaecological issues, which are taboo in rural Bangladesh due to socio-cultural beliefs (Hennegan and Sol 2020). Therefore, these issues often go untreated. In addition, flooding and destroyed roads after floods or cyclones can hinder access to healthcare, making it difficult to find medical help. Pregnant women are particularly affected: “The roads are so bad, that while crossing roads to get to the clinic or hospital both mother and child can lose their lives” (Interview 81, Durgapur). In response, people use their own or borrowed boats, but not everyone has access to boats and the water can be dangerous during storms.

People mentioned increased skin diseases, pigmentation and dark spots related to salinisation in Shyamnagar and rising temperatures in both upazilas. An increase in skin pigmentation can be problematic due to the prevalence of colourism—the preference for lighter skin over darker skin—in Bangladeshi society (Jensen 2020). Skin problems can complicate marriage arrangements for girls, as they are perceived as less attractive according to the convention described. This can lead to early marriage, as the skin is less affected at earlier ages. Girls may also be kept indoors more to preserve their skin. Exposure to saline water in Shyamnagar also causes hair loss, damage, and greasiness. In response, people use henna, shave their heads, or increase the use of shampoo. Conversely, some accept the changes: “We always go through economic trouble, no time to think about hair” (FGD3, Shyamnagar).

4.6 Mental well-being

Participants in both upazilas recalled a plethora of issues concerning mental well-being. Commonly mentioned problems included anxiety, stress, sadness, insomnia, and depression. They can arise from virtually any loss or damage. However, prominent causes are financial insecurity, fear of future hazards, and an uncertainty of what the future will look like for the next generation: “We are suffering due to the weather. I wonder how my child will survive and about his future, whether he will have enough opportunities or not” (Interview 61, Durgapur). In both upazilas, women expressed a wider range of sources of psychological distress, particularly in relation to health. The inundation and destruction of sanitation facilities during cyclones or floods is a particular source of discomfort for women: “My body was completely wet alongside my clothes, and I was menstruating as well. This disgusted me” (Interview 80). Women have less freedom to respond to hazards; during heat waves, "men go to the riverside for wind, we women stay at home" (Interview 44, Durgapur). In addition, participants indicated that there is little privacy after floods due to the lack of gender-segregated cyclone shelters.

Disaster preparedness, such as fortifying levees or homes and checking weather forecasts, helps reduce stress and anxiety. People also turn to religion. However, there is not always the capacity or time: “Everyone prays together and tells each other God will fix everything, but we are in such a state of panic that there rarely is a chance to pray” (Interview 95, Durgapur). People also turn to each other for mental support. There were few accounts of people seeking psychological health services, as mental health awareness and services are limited in rural Bangladesh: “For stress, I think we need to consult with a specialist who can support us, but this is not common in our village; I normally share my stress with friends” (Interview 72, Durgapur). People also manage physical stress symptoms, such as headaches or insomnia, by taking medication or visiting a doctor. Another response to reduce hazard-related mental health problems is to migrate to another, safer area. However, this requires significant economic resources. For example, a Dalit community in Durgapur lives in highly flood-prone areas but does not have the means to move. Finally, participants also noted a degree of resignation: “Before, we were more concerned. We are getting used to it after going through all the cyclones and floods” (Interview 2, Shyamnagar).

5 Discussion

People in Shyamnagar and Durgapur are experiencing significant L&D to which they have formulated different autonomous responses. In this section, we first discuss the observed NELD, the autonomous responses, and the capacities that responses depend upon. We then identify key lessons and examine how NELD can be addressed in an equitable manner. Finally, we conclude by highlighting policy and research priorities for addressing NELD.

5.1 NELD, responses, and capacities

Given the decision to examine only specific themes, this study does not provide an exhaustive overview of NELD experienced in rural Bangladesh. Nonetheless, extensive climate impacts were observed within each theme. This shows that, like in other Majority World countries, people in Bangladesh experience a vast range of NELD that severely disrupt rural societies. Impacts vary across and within groups. For example, women reported more adverse impacts on physical and mental health, different religious groups all experience NELD specific to their beliefs and practices, and low-income households, such as the Dalits in Durgapur, often live in areas highly exposed to floods. Subsequently, these groups all experience disproportionate levels of NELD.

Affected individuals and groups autonomously formulate responses to most NELD. People in both upazilas seek health care, use boats to upkeep mobility, relocate, start using fertilisers, (re)plant trees, homeschool students, and assist each other in times of need. Moreover, religion is important in addressing NELD, as people turn to God(s) in response to various impacts. However, no responses were recorded to some culture- and religion-related impacts. Nor did we record widespread efforts to restore public goods, such as damaged roads or religious buildings. Moreover, many responses fail to alleviate NELD completely.

The responses we examined were often formulated in the face of immediate stress to protect basic welfare and security in the context of insufficient resources. Participants indicated that they accepted some NELD, prioritising certain aspects of life over others. For example, some specific responses were economically infeasible or other needs were more acute. As a result, and as also found in other geographic contexts (Warner et al. 2012), people in rural Bangladesh could not fully adapt and were forced to cope with L&D. These coping responses frequently erode people’s livelihoods: buying previously home-grown products on markets, using saplings to restore tree populations, or medicine and healthcare further strains societies already living in semi-subsistence conditions (van der Geest and Dietz 2004). Some of the observed autonomous responses are causing harm beyond monetary losses. Increasing chemical fertiliser usage can cause issues such as reproductive abnormalities, cancer, and hormonal disruptions (Hossain et al. 2022) and, therefore, adversely affect people’s health. Marrying off daughters at an earlier age can cause significant mental and physical health issues (Hayward and Ayeb-Karlsson 2021).

Data also reveal various factors people depend on to cope or adapt. Public services, such as local clinics or hospitals, are essential to ensure that people can receive health care when they have mental or physical health problems. Moreover, roads facilitate access to key locations and connect people to family or community members. However, healthcare facilities are scarce in both upazilas, and roads are not always maintained or repaired post-flooding. Many of the responses to NELD require people to use their financial assets. Consequently, NELD can lead to significant economic disadvantage, especially in societies with high levels of poverty. Furthermore, the effectiveness of autonomous responses depends on local ecosystems, which are rapidly degrading due to development and climate-related impacts (Hasnat et al. 2018). Finally, people also depend on social relationships with family or community members who can provide emotional support or food and shelter. While useful, these networks can only provide limited support as family or community members often also must face the same covariate climate-related hazards. All of these factors are expected to be adversely affected by further climate change (Serdeczny et al. 2024). Thus, the barriers and potential limits to autonomous responses outlined above will become more and more apparent.

Not all groups’ capacities to respond to NELD are equal. For example, the absence of adequate public services particularly affects women, who face more mental and physical health impacts. A lack of women doctors prevents them from seeking assistance for particularly culturally sensitive health issues, such as gynaecological problems (Hennegan and Sol 2020). Deteriorating roads make it difficult to access healthcare. The reliance on financial assets to cope with NELD is particularly problematic for economically marginalised households, who do not have the financial means to respond and, if they deem it necessary to spend their savings, their financial security is further reduced (Warner & van der Geest 2013). In addition, social relations vary by person or group, with minorities generally receiving less support. These dynamics show how those who are already vulnerable or marginalised can be disproportionately affected because they live in more exposed areas and have less capacity to respond than privileged or affluent households (Sovacool 2018). Consequently, without external support to those most vulnerable, NELD further exacerbates societal and regional inequalities. This clearly indicates the limits to autonomous responses and where external assistance could be both necessary and appropriate.

5.2 Addressing NELD in a just way

External institutions can learn from the autonomous responses we observed or work with affected societies to strengthen already existing responses (Rahman et al. 2023; Rahman and Hickey 2019). For example, national and international institutions can assist in replanting trees that provide shade and culturally relevant forest products, discuss with affected societies the need to facilitate home schooling during periods when schools are inaccessible, or examine what medicines they most commonly use in response to different climate-related hazards and increase the accessibility of these medicines. In addition, basing interventions on existing local responses helps ensure that interventions fit local contexts (Rahman et al. 2023). However, it should be emphasised that some autonomous responses can impose further harm and should not be replicated. Moreover, the observation that not all NELD was responded to by people themselves indicates that external interventions are needed to address L&D comprehensively. With public goods, such as roads or religious buildings, for which individuals or communities do not have sufficient resources or agency to make changes, interventions are urgently required as these are fundamental baseline needs in communities.

In addition to strengthening specific interventions, people need financial security to address NELD fully. Without sufficient financial resources, people cannot restore culturally relevant objects or access services such as healthcare and education. Financial support can be given through direct monetary compensation, but also by increasing local job and training/re-training opportunities or livelihood diversification, especially for women and girls (van Schie et al. 2023a). The notion that NELD cannot be fully addressed without monetary compensation shows that the strict dichotomy between ‘economic’ and ‘non-economic’ L&D is ambiguous at the local level, as evidenced by our study. It should be emphasised that economic responses cannot fully address the extent of NELD, especially regarding permanent losses. Moreover, the importance of personal financial assets in responding to NELD might differ depending on socio-economic contexts. For example, many Pacific Island nations face significant NELD (Chandra et al. 2023; McNamara et al. 2021), but absolute poverty is often less pronounced than Bangladesh (Connell 2010). As such, more ‘intangible’ ways to address L&D, such as preserving culture or nature, could be perceived as more urgent than financial support (McNamara et al. 2021). Still, mass mobilisation of funds is needed to organise these activities.

The different determinants people depend upon in formulating responses to NELD reveal how piecemeal, project-based interventions cannot fully address NELD. Political-economic systems, especially public services and social protection, must be transformed to serve people better in responding to L&D (Jackson et al. 2023). As public services are ultimately the responsibility of the government, external interventions are still needed to repair buildings, schools, and infrastructure and increase healthcare accessibility. However, as Bangladesh is burdened with disproportionate L&D and high absolute poverty levels (UNDP 2022), it cannot be expected that the national government has sufficient capacity to improve and repair these public services and goods fully. The Fund for responding to Loss and Damage, among others, could alleviate this burden and allow the Bangladeshi government to invest in adaptations or public services—especially to alleviate NELD—which is out of scope for autonomous responses. However, we also demonstrated how CCA interventions and developmental activities supported by the government, such as industrial-scale shrimp farming in Shyamnagar or sand mining in Durgapur, can have maladaptive outcomes for human and environmental systems. Similarly, as power often reproduces itself through adaptation (Falzon 2021), there is a risk that increased capacity for NGOs or the government—which is increasingly seen as autocratic (Aase 2021)—will not serve those who need it most and, instead, increases marginalisation and inequalities (Kaijser and Kronsell 2014). Therefore, addressing underlying vulnerabilities while scrutinising the norms guiding externally-funded responses to climate change is needed. Moreover, a capitalist political economy, which both drives emissions through the economic growth imperative and produces and necessitates inequality, must be constrained, if not ended (Jackson et al. 2023; Perry and Sealey-Huggins 2023). Otherwise, people will continue to be marginalised, and their capacities to respond to NELD will stay limited.

5.3 Policy priorities and ways forward for research

We identify four policy priorities accompanied by suggestions for further research to better address NELD. First, recognising existing autonomous responses can enable the development of externally-funded activities that can replicate or strengthen these responses. This will require further assessments in a variety of contexts, particularly in geographies where the academic literature on NELD is scarce, such as many African and South American countries, or marginalised groups in the Global North (Jackson et al. 2023). Moreover, in addition to assessing what autonomous responses exist, this will require deeper insights into how autonomous responses evolve to better understand the mechanics of these processes. As autonomous responses can also cause adverse effects, these assessments should critically analyse the outcomes of each response to avoid strengthening maladaptive responses and perpetuating harm. Second, beyond strengthening individual responses, enhancing people’s adaptive capacity will better enable them to respond to NELD. Third, to enable institutions to strengthen existing responses and people’s adaptive capacities, significant North–South transfers of capital—based on principles of common but differentiated responsibilities and respective capabilities (UNFCCC 2015)—are needed. These resources can help alleviate the burdens of respective Majority World governments, allowing them to address NELD and increase people’s capacities. Finally, marginalisation and inequality can be perpetuated or worsened in certain instances as aid increases. For example, increasing neoliberal or technocratic responses or adaptations, such as large-scale shrimp farms in Bangladesh, can harm local people (Dewan 2021). Thus, climate finance alone will not necessarily address NELD for the most vulnerable. Policies addressing NELD should be designed and implemented by accounting for the structural root causes of vulnerability, differentiated vulnerabilities and emphasise alleviating and recovering NELD for those most marginalised.

6 Conclusion

Our research shows that climate change is causing widespread NELD among people in rural Bangladesh, and that affected populations are active agents in responding to NELD within severe resource constraints. Every day, people in the Majority World bear the burden of climate change, caused primarily by historical emissions from countries and industries in the Global North, and must cope with its impacts. People's autonomous responses to NELD should be recognised and their resilience applauded. However, this cannot distract from the grave injustice that subsistence-dependent people bear a significant burden by working and spending already limited resources to address NELD for which they are not ultimately responsible (Perry and Sealey-Huggins 2023). This burden can be alleviated by initiating external responses that enhance local people's capacities, such as improving their financial situation and the public services and ecosystems on which their livelihoods depend. Identifying, learning from, and supporting existing autonomous responses when designing external responses can ensure that these externally initiated responses are responsive to local needs. A key finding is that addressing NELD often requires economic responses, an observation that is currently not widely recognised in the L&D discourse. Likewise, most ‘economic’ L&D also causes NELD, and vice versa. Importantly, we suggest that simply strengthening existing responses will not fully alleviate NELD, especially as no autonomous responses to some NELD were observed. Some autonomous responses were clearly erosive and caused further harm to people. NELD scholars can take forward our findings and address the remaining gaps and issues to ensure that external responses address the NELD that people cannot address independently. This research is critical as climate change intensifies and international climate governance continues to fail to provide adequate resources to those most affected, such as those in rural Bangladesh.