Livelihoods after resettlement: women's work, collective care, and solidarity in disaster recovery

Abstract

This paper critically examines livelihood recovery in the peripheral resettlement sites of Chennai, India, following the 2004 Indian Ocean tsunami. Despite the implementation of various gender‐inclusive livelihood programmes as a part of disaster recovery, our longitudinal empirical research reveals a persistent gap between formal recovery measures and the everyday conditions required for livelihoods to become gender‐inclusive and sustainable over time. Drawing on in‐depth fieldwork, we explore how women respond in the long term to these limitations: not simply through individual coping strategies but also by developing and maintaining collective, care‐centred practices, including adapting domestic spaces to work, sharing childcare responsibilities, and building networks of mutual support. These practices are often invisible in official recovery frameworks, yet they serve as critical choices for women both to sustain their livelihoods and perform their care roles. We build on Joan C. Tronto and Berenice Fisher's (1990) theory of caring to conceptualise these acts as forms of ‘caring with’, where care becomes a collective, political, and relational response to systemic exclusion. In doing so, we reposition disaster recovery not merely as a return to normalcy or economic productivity but also as a deeply social and gendered process that is shaped by care and solidarity.

1. INTRODUCTION

Disasters can have profound and lasting impacts on livelihoods. Beyond the immediate destruction of housing and infrastructure, they fracture employment networks, interrupt access to markets and services, and unsettle the social relations upon which everyday survival is dependent (Oliver‐Smith, 1991; Cernea, 1997; Davis and Alexander, 2016). Across the Global South, disaster recovery is increasingly mediated through large‐scale resettlement programmes which relocate vulnerable populations to peripheral areas (Oliver‐Smith, 1991; Balachandran, Olshansky, and Johnson, 2022; Asgary et al., 2024). While these interventions are commonly presented as humanitarian and risk reduction strategies, in the long term, they frequently reorganise socio‐spatial relations and hamper access to urban resources.

Peripheral resettlement is not merely spatial relocation; it also effectively reduces the viability of urban life (Parpiani et al., 2026). Bhan (2020) argues that relocating low‐income populations to the peripheries often severs links with their livelihood and creates unequal access to infrastructure, services, and citizenship. Peripheral resettlement sites are typically located far from employment centres, social networks, and essential infrastructure (Coelho, 2016; Nikuze et al., 2019; Williams et al., 2022). These spatial dislocations impose additional financial and temporal burdens on already vulnerable households. Increased commuting distances, higher transport costs, and decreased access to informal earning opportunities often destabilise fragile income strategies (Perera, Weerasoori, and Karunarathne, 2012; Kondo and Karatani, 2016; Espina and Canoy, 2021). Simultaneously, displacement unsettles childcare arrangements, neighbourhood reciprocity, and the community support networks that sustain everyday livelihood practices (Bradshaw, 2002; Nygren and Wayessa, 2018). Thus, livelihood recovery becomes more uneven and disproportionate for women as compared to men.

This paper focuses on the resettlement sites of Semmancheri and Perambakkam located on the southern periphery of Chennai, India. These settlements were built in the aftermath of the 2004 Indian Ocean tsunami and were subsequently expanded following recurrent floods and riverbank evictions. They are located 20–25 kilometres away from the urban centre and on ecologically vulnerable floodplains.

Existing research demonstrates that women often experience disproportionate vulnerabilities in disaster contexts, shaped by pre‐existing gender inequalities, limited access to and control over assets, mobility constraints, and increased burdens of unpaid care work (Bradshaw, 2013; Okai, 2022; Roy Chowdhury and Agarwal Goel, 2023). These vulnerabilities can become particularly pronounced following relocation to peripheral resettlement sites.

This study draws on the longitudinal lived experiences of women who have resided in these locations since the 2004 tsunami. In the years since, these women have experienced floods (in 2015, 2019, and 2023), as well as the COVID‐19 (coronavirus disease 2019) pandemic. Their experiences of disasters also help with drawing attention to processes of ‘unrecovery’ that remain obscured when recovery is understood primarily as a short‐term transition from disruption to restoration (Parpiani et al., 2026). The paper illustrates that livelihood recovery is neither a linear return to normalcy nor a discrete process following each disaster; rather, it is a cyclical, layered, and anticipatory process shaped by spatial marginalisation and structural inequality.

1.1. Conceptualising sustainable livelihood recovery

Livelihood restoration is widely recognised as a central component of disaster recovery. Chambers and Conway (1991) define a livelihood as comprising the capabilities, assets, and activities required for a sustainable means of living and to cope with and recover from stresses and shocks. Building on this conceptualisation, the Sustainable Livelihoods Framework (SLF), developed by the United Kingdom's Department for International Development (DFID, 2001), highlights five interrelated forms of capital (human, social, natural, financial, and physical) and their interaction within contexts of vulnerability. The SLF's central contribution is shifting attention beyond income and employment towards a multidimensional understanding of the resources and capabilities that people utilise to construct and sustain their livelihoods. It also emphasises that livelihood outcomes emerge from interaction among assets, strategies, institutions, and broader vulnerability contexts rather than from any single form of capital. The SLF is an analytical tool as opposed to a fixed model, emphasising that livelihoods are dynamic and that the framework should be adapted to different settings (Scoones, 2009).

Disaster‐induced resettlements, however, often fail to account for these local pathways, which are inclusive of social relations, cultural values, and the everyday practices of the communities (Faas et al., 2020; McDonnell, 2021; Goswami, 2024). In practice, many recovery programmes invoke the language of participation; in reality, though, they only offer inadequate and tokenistic participation (Danar and Pushpalal, 2014; Yee, 2020; McDonnell, 2021). Livelihood recovery interventions also fail owing to a lack of critical engagement with underlying power relations, gender norms, political economy, and structural inequalities (Scoones, 2009; Piggott‐McKellar et al., 2020; Daly et al., 2023).

Disaster recovery research has consequently used the SLF to examine how households mobilise different forms of capital, develop livelihood strategies, and negotiate changing vulnerability contexts. (Nakagawa and Shaw, 2004; Aldrich, 2012; Ganapati, 2012; Joshi and Aoki, 2014). The framework is criticised for assessing livelihood conditions in terms of the availability or accumulation of the aforementioned five forms of capital (Scoones, 2009). Such applications risk reducing its relational and dynamic conception of livelihoods to a checklist of assets and outcomes. Scoones (2009) asserts that the problem lies partly in the mechanistic application of livelihood frameworks, where the complexity of livelihood processes is simplified into predetermined categories while insufficient attention is given to political and institutional processes. More recent scholarship similarly identifies persistent problems concerning the lack of engagement with scale, temporality, politics, and the epistemological assumptions of livelihood approaches (Natarajan et al., 2022).

Scoones (2009) therefore calls for greater attention to institutions, power, markets, politics, and historical processes in livelihood analysis. This critique is particularly significant in disaster recovery, where recovery is not simply a process through which affected households rebuild their assets, but one mediated by states, development agencies, financial institutions, markets, and other actors with unequal capacities to shape outcomes.

By deprioritising political and institutional factors, the SLF treats a household as a relatively coherent livelihood unit, which in turn obscures important differences between its members (Natarajan et al., 2022). Resources, decision‐making power, paid employment, unpaid labour, mobility, and access to social networks are often distributed unequally within households, particularly along gendered lines (Moser, 1993; Kabeer, 1994). Hence, an intervention focused narrowly on the household may simultaneously overlook macro‐level structures and power and micro‐level differences within households. This is particularly significant in relation to recovery, where state‐led resettlement decisions are shaped by institutional and political actors, while the consequences of those decisions are experienced differently by women, men, older people, and other household members.

A related critique concerns SLF's relatively static and asset‐based framing of vulnerability and resilience (Karki, Matthewman, and Grayman, 2022; Natarajan et al., 2022). Blaikie et al. (1994) assert that vulnerability is not inherent but is produced through unequal entitlements and constrained access to and control over resources. Within disaster studies, alongside vulnerability, social capital is seen as a relational asset (Nakagawa and Shaw, 2004; Aldrich, 2012). It is defined as the networks, relationships, and norms of trust that individuals and groups mobilise for collective or individual goals (Aldrich, 2012; Ganapati, 2012). Thus, social capital is an explanatory variable for differential recovery outcomes, with dense social ties and high trust repeatedly associated with communities’ capacity to ‘bounce back’ after a disaster (Nakagawa and Shaw, 2004; Aldrich, 2012; Joshi and Aoki, 2014; Uekusa, Matthewman, and Lorenz, 2022).

This framing risks shifting the burden of recovery onto already vulnerable populations by placing the onus on them to ‘bounce back’ without sufficiently interrogating the structural and political‐economic conditions that produce and reproduce harm to individuals and their social ties and networks (Joseph, 2013; Cheek and Chmutina, 2022). Therefore, the capacity to ‘bounce back’ should not be understood as an intrinsic community trait but as a differentiated outcome shaped by the interaction between structurally produced vulnerability and access to livelihood assets, including social capital.

These critiques become particularly relevant in disaster‐induced resettlement, where displacement from established residential and livelihood spaces can fundamentally reshape the conditions under which livelihoods are sustained. Relocation can sever or weaken access to employment, markets, social networks, schools, healthcare, and transport, thereby altering the conditions under which livelihood assets are accessed and converted into livelihood opportunities (Oliver‐Smith, 1991; Coelho, Venkat, and Chandrika, 2012; Goswami, 2024). In this context, spatial marginality or peripherality is understood not simply as geographical distance from the urban core, but as a relational and multidimensional condition produced through unequal access to employment, services, markets, mobility, and social and economic resources (Testori et al., 2021). These processes of ‘unrecovery’ usually unfold gradually and/or accumulate over time, making them difficult to capture in short‐term assessments of asset loss or recovery (Parpiani et al., 2026).

Disaster recovery scholarship increasingly recognises the importance of longitudinal approaches in capturing the ‘unrecovery’ (Rubin, 2009; Rouhanizadeh, Kermanshachi, and Nipa, 2020). Yet, livelihood assessments do not always adequately capture how strategies, assets, vulnerabilities, and institutional relationships evolve over extended periods (Scoones, 2009). In this study we foreground the social, relational, and spatial conditions under which livelihoods are sustained and reproduced over time. A temporal perspective is crucial because apparent signs of recovery may coexist with processes of longer‐term impoverishment. A household may receive a permanent house, regain employment, or access financial assistance while simultaneously experiencing declining income stability, depletion of productive assets, growing reliance on debt, intensified unpaid care work, or weakened social support in the long term. During prolonged periods of economic adjustment and recovery, marginalised groups, particularly women, often bear a disproportionate share of livelihood losses and unpaid care responsibilities (Himmelweit and Plomien, 2014; UN Women, 2014). Attending to these temporal dynamics therefore requires particular attention to care work, gendered divisions of labour, and social reproduction in shaping livelihood possibilities. This shift allows for a more grounded account of how livelihoods are negotiated and transformed in the aftermath of resettlements.

1.2. Gender, care work, and recovery

Feminist disaster scholarship has consistently demonstrated that disasters are not gender‐neutral (Enarson, 1998; Bradshaw, 2002; Enarson, Fothergill, and Peek, 2007). In post‐disaster contexts, women are frequently at the forefront of grassroots mobilisation, mutual aid, and informal recovery efforts (Bradshaw, 2002; Aldrich, 2012; Alam and Rahman, 2019). At the same time, they continue to shoulder increased childcare, eldercare, and household maintenance responsibilities (Fothergill, 1996; Erickson and Ritter, 2001). Despite this dual role, the centrality of everyday care work and labour of social reproduction both in routine life and in shaping long‐term recovery processes remains insufficiently theorised and underemphasised within existing disaster literature (Ariyabandu, 2009).

Existing studies on women‐centred livelihood interventions further highlight the limitations of externally‐driven recovery models. Approaches centred on skill development, asset transfers, or non‐governmental organisation (NGO)‐led diversification frequently fail to produce sustainable outcomes, as they are often misaligned with local economies, gender norms, and ecological conditions (Goulden et al., 2013; Roy, 2020; Karki, Matthewman, and Grayman, 2022). Hanrahan (2015) critiques conventional livelihood frameworks for privileging material outcomes and individual autonomy and argues for an understanding of livelihoods as embedded in cycles of dependency and care.

To address this gap, the present paper situates livelihood recovery within the framework of social reproduction, which comprises the gendered processes and structures that are often unpaid and invisible in sustaining daily lives (Bauhardt and Harcourt, 2019; Mezzadri, 2019; Federici, 2021). We do this by emphasising care as the labour and ethics that support well‐being and commoning as the collective management of shared resources and mutual support systems. According to Tronto and Fisher (1990), care is understood not as a private or individual responsibility but as a collective, political practice grounded in interdependence. Following these authors, we comprehend care as a relational practice that encompasses activities required to maintain, continue, and repair our shared world, extending beyond interpersonal caregiving to broader social conditions that sustain everyday life. Tronto's (2013) notion of ‘caring with’ is particularly pertinent here, as it foregrounds solidarity, mutual recognition, and shared responsibility as central to sustaining life. Therefore, this framework helps in shifting our engagement with care from individual or private responsibility to that of networks of interdependence. This relational understanding of care can be further deepened by Federici's (2019) conceptualisation of commoning, which frames collective practices of care by sharing labour, resources, and knowledge as central to sustaining life under conditions of capitalist dispossession. Extending this perspective, feminist scholars have shown how everyday practices of friendship, mutual aid, and informal support function as forms of care that mould socio‐economic life (Rao, 2014; Phadke and Kanagasabai, 2023). These practices are not peripheral to economic systems; rather, in the long term, they actively shape how people access resources, navigate uncertainty, and reproduce their livelihoods in contexts where formal institutional support is limited or uneven. In this sense, livelihoods are not simply constituted through assets or capital but are relationally produced through networks of care that both enable and constrain economic possibilities. Hanrahan (2015) developed the connection between care ethics and livelihoods by highlighting the relationality of care, where attending to the needs of others can directly influence one's own economic opportunities and constraints.

Building on these strands of literature, this paper departs from approaches that treat livelihood recovery as a technical or time‐bound intervention. Instead, it conceptualises recovery as a long‐term, relational, and gendered process, unfolding through the everyday practices of care and social reproduction.

2. METHODOLOGY

This study adopts a qualitative research design to unpack the long‐term livelihoods of women in a disaster resettlement context. The study draws on in‐depth interviews with 19 women residing in the Semmancheri and Perambakkam resettlement sites. Interviews were conducted in two phases: first from December 2022 to April 2023, and then from December 2023 to March 2024. The participants were resettled in the two sites either because of the 2004 Indian Ocean tsunami or the 2015 Chennai floods. At the time of the research, more than two decades had passed since the 2004 tsunami, and nearly a decade since the 2015 floods. By carrying out the fieldwork well beyond these disaster events, the study captures the complex interplay of livelihood recovery and time.

Respondents were selected using purposive sampling, focusing on women engaged in informal and small‐scale livelihood activities, including small shop owners, door‐to‐door vendors, domestic workers, waste collectors, and construction labourers. Snowball sampling was employed to identify additional respondents in similar occupational categories. The majority were married (n = 14), while others were widowed or separated. Literacy status ranged from no formal schooling to completion of secondary school. In addition to the 19 women mentioned above, 10 interviews were also conducted with government officials and representatives of NGOs involved in housing, welfare provision, and livelihood support in the resettlement sites. These interviews provided institutional perspectives and histories of livelihood recovery that complemented women's narratives.

Data collection involved in‐depth semi‐structured interviews, informal conversations, and participant observation. The first author, Suchismita Goswami, approached the participants at their workplace or met them in their houses during the late afternoon when they had completed their shifts. For women managing shops or kiosks, interviews were often held on their work premises. In these cases, care was taken to avoid disrupting commercial activities; interviews were paused to allow respondents to attend to customers. These interactions also enabled observation of transactions, negotiations, and everyday decision‐making practices embedded within livelihood activities. Interviews explored livelihood trajectories before displacement, experiences of relocation, current income‐generating strategies, spatial mobility, and household responsibilities, vulnerabilities, and aspirations. Attention was paid to changes in labour patterns and risk exposure following relocation, such as means of employment, safety, debts, and care responsibilities. Interviews lasted for approximately 40–65 minutes and were conducted in Tamil. All were audio‐recorded with consent and later transcribed and translated into English for analysis. Follow‐up visits and informal interactions were undertaken to clarify emerging themes and deepen contextual understanding.

The data were analysed using an inductive thematic analysis approach (Flick, 2009). The analysis followed a systematic multi‐stage process. First, all transcripts were read repeatedly to ensure familiarity with the data and to pinpoint preliminary patterns. Second, open coding was performed manually, generating initial codes related to livelihood practices, gendered labour burdens, spatial marginalisation, institutional mediation, vulnerability, and adaptive strategies. Coding was iterative and reflexive, identifying new categories. Third, codes were grouped into broader thematic clusters through constant comparison across interviews. This stage involved spotting relationships among codes and consolidating them into higher‐order themes such as intersecting risks, skill adaptation, livelihood diversification, and gendered negotiations of space. Fourth, themes were refined through cross‐checking among the authors to ensure analytical consistency and conceptual clarity. Discrepancies in interpretation were discussed until consensus was reached. Analytical memos were maintained throughout the process to document emerging insights and reflexive observations.

Reflexivity was integral to fieldwork, analysis, and writing. The first author conducted the fieldwork and has resided in Chennai for nearly a decade, developing familiarity with local practices, climatic conditions, neighbourhood geographies, and women's informal networks. This long‐term engagement facilitated rapport building and contextual understanding. However, despite this proximity, she occupied a position of partial outsiderness shaped by her socio‐economic location in the city centre and her academic affiliation (Lindvall and Rueda, 2014). The researcher's shared gendered identity with women respondents often facilitated open and candid discussions, particularly about experiences of care work, financial precarity, and informal employment. At the same time, differences in class position, educational background, and institutional affiliation created asymmetries that required careful negotiation. The authors remained attentive to these intersecting positionalities. Reflective field notes were maintained throughout the research process to examine critically the assumptions, positionality, and relational dynamics that shaped interactions and analytical decisions. Coding choices and thematic interpretations were discussed collaboratively among the authors to minimise individual bias.

Ethical protocols were rigorously observed throughout the research process. Informed consent was obtained from all participants before each interview. Respondents were clearly informed about the objectives of the study, the voluntary nature of their participation, their right to withdraw at any stage, and the steps taken to ensure confidentiality. All names used in this paper are pseudonyms and potentially identifying details have been altered where necessary to protect anonymity. Audio recordings and transcripts were securely stored and accessed only by the authors.

3. FINDINGS AND DISCUSSION

The following section explores how women relocated to peripheral resettlement sites were neither able to benefit meaningfully from newly introduced livelihood interventions nor return to pre‐disaster occupations that had been embedded in spatial, social, and economic networks. In this study peripherality is not treated as a single measurable attribute of distance from the city, but as a multidimensional and lived condition. Rather than using Annexure 1 (see the Supplementary Materials) ¹ as a quantitative measure of peripherality, we trace how peripheralisation is experienced through changes in women's mobility, access to employment, markets, and services, and the disruption of spatially embedded livelihood networks. Across respondents, livelihood changes included withdrawal from previous occupations, downward occupational mobility, increased dependence on local and often lower‐paid work, and, in some cases, continued employment that entailed greater travel time and cost.

Although livelihood disruptions were experienced across genders, women's experiences were further shaped by their responsibilities for income generation, care work, and household financial management. The resulting trajectories were therefore uneven and often stalled, characterised by repeated adjustments, occupational shifts, borrowing, and asset liquidation. The livelihood trajectories, household composition, and debt experiences of the 19 in‐depth interview respondents are summarised in Annexure 1 in the Supplementary Materials.

3.1. Limits of state‐led livelihood recovery

After the 2004 Indian Ocean tsunami, gender‐sensitive interventions were incorporated in the resettlement sites. They included allotting housing titles in women's names, skills development programmes, and targeted livelihood schemes. While these interventions provided short‐term support, their long‐term outcomes remained limited. Although these state‐led livelihood interventions were well‐intentioned, they often failed to translate into sustained economic outcomes owing to the interaction of gender norms, spatial constraints, and recurring shocks, as seen in Table 1.

TABLE 1.

Outcomes of livelihood interventions post resettlement.

Type of intervention	Number of respondents who participated	Outcome	Key constraints identified
Vocational training (tailoring, embroidery, and ari work)	14	Limited or no significant increase in income; skills not effectively translated into sustainable livelihoods	Market saturation within resettlement sites, less local demand, a lack of capital, and weak market linkages
Auto rickshaw scheme	4	Largely unsustainable; autos remained unused, transferred to male household members, or sold during financial distress	Gender norms restricting women's mobility, safety concerns, high maintenance costs, and financial shocks (such as the COVID‐19 pandemic and debt burden)
Self‐help groups and microfinance	15	Enabled short‐term financial coping and consumption; however, they contributed to long‐term indebtedness	High interest rates, repeated borrowing cycles, unstable income sources, and limited productive investment opportunities
Solid Waste Management Programme	0	Initially provided local employment and regular income, particularly for women	Institutional discontinuity, withdrawal of funding/support, a lack of long‐term policy commitment, and limited upward mobility

Source: authors.

Following resettlement, the women in Semmancheri were provided with auto rickshaws, driving licences, and training. While several women obtained licences, very few used the vehicles for income generation. Respondents attributed this to safety concerns, restrictive gender norms, and the peripheral location of resettlement sites. This limited viable routes and customer access. As Ramesh, the husband of one respondent (Shalini), explained: ‘Yes, my wife got training and a licence to drive an auto. The government partially paid for the auto and we paid the rest of the amount in instalments every month. Yet my wife never drove the auto as she was not interested, and this area is not safe for women’.

Ramesh, originally from Sengi, migrated to Chennai in the mid‐1990s in response to the declining viability of agrarian livelihoods in his native village. Before the 2004 Indian Ocean tsunami, he resided with his family in central Chennai, where he was engaged in informal street vending, selling boiled peanuts at a major bus junction. As his house was destroyed in the tsunami, Ramesh, along with his family, was relocated to Semmancheri in 2006. Despite this spatial displacement, he continued to commute approximately 20 kilometres to central Chennai to maintain continuity of livelihood. Ramesh started driving the auto rickshaw that his wife was given. Shalini, meanwhile, was unemployed for a long time after resettlement before eventually taking up a job as a domestic worker in the surrounding neighbourhood. However, owing to the financial strain of the COVID‐19 pandemic, Ramesh was compelled to sell the auto rickshaw to repay debts. Since then, he has resumed his earlier occupation of selling peanuts as his primary source of income, commuting back to central Chennai.

Another respondent, Shanti, shared a similar story. She was also allotted an auto rickshaw after resettlement but was forced to sell it following her husband's death. She reinvested the money in building a small shop outside her home and paying off her debts.

These cases demonstrate that asset‐based interventions often fail to account for intra‐household dynamics, gendered mobility constraints, and the economic instability that shapes decision making in resettlement contexts. Rather than enabling women's independent livelihoods, such assets frequently become household coping mechanisms that are liquidated during periods of crisis.

A similar dynamic emerged in relation to vocational training programmes. Many women participated in courses such as tailoring, embroidery, and lace work, regularly facilitated by NGOs. While these programmes enabled skill acquisition, they rarely translated into sustained income. Women reported difficulties in capitalising on their training owing to limited local demand, logistical challenges in accessing materials, and restricted market opportunities within resettlement peripheries characterised by low purchasing power and indebtedness. Of the 19 women interviewees, 14 reported enrolling on tailoring courses provided by the government or NGOs in the resettlement sites. Nithya, for instance, enrolled on a tailoring course as her husband was unable to work after an accident. Although she completed the training, she reported little improvement in earnings, explaining: ‘There are already many women doing tailoring here, so I don't get many orders’. Her experience reflects a broader pattern of market saturation within resettlement sites, where similar skills are widely distributed, but demand remains limited. A recent article indicates that in Perambakkam alone, more than 1,000 certified tailors remain without stable employment opportunities (Radhakrishnan, 2024).

Trainers working on these programmes also noted that women's participation in formal employment remained difficult to sustain. Even when job placements were arranged, many women withdrew within a few days, citing long travel times, safety concerns, and a lack of family support. As one trainer observed, women often pointed out that ‘it becomes too late to return home’ or that ‘their family does not allow it’. In this sense, the constraints were not limited to skill deficits but also were embedded in everyday conditions associated with the isolated location of their residence as well as societal norms regarding care and mobility.

Lastly, even programmes that initially appeared successful struggled to endure over a longer period of time. The Solid Waste Management Programme in Semmancheri, for example, generated local employment and community participation in its early stages. Eighty women from self‐help groups were trained in waste segregation and composting. Revenue from the sale of vermicomposts was shared between the women and the Tamil Nadu Slum Clearance Board. ² However, its effectiveness declined as administrative responsibilities shifted and political priorities changed. At present, the initiative is privatised and is not attached to local employment or asset creation. None of the 19 women interviewed in this study was employed by the programme. ³ This gap between programme design and actual reach underscores the uneven and limited penetration of such initiatives at the household level.

Taken together, these findings indicate that livelihood recovery in resettlement contexts is not a linear progression from intervention to stability. Instead, it is marked by discontinuities, where initial gains from training or asset provision are frequently undermined by structural constraints, including spatial marginality, weak market linkages, gender norms, and recurring crises such as flooding. As a result, many interventions fail to produce sustainable livelihood outcomes, revealing a critical lacuna between policy design and lived realities.

3.2. Gendered dimensions of spatial marginality and the feminisation of livelihoods

The following subsection shifts the focus to how pre‐disaster lives and livelihoods became inaccessible after relocation. We highlight how spatial displacement to the urban periphery has constrained livelihood reconstruction. Relocation to Semmancheri and Perambakkam did not simply displace households geographically; with limited access to infrastructure, employment, and public services, these areas have remained spatially and functionally disconnected from Chennai's urban core (Arabindoo, 2011; Coelho, Venkat, and Chandrika, 2012). Relocation has therefore continued to reinforce different forms of socio‐economic exclusion, exposure to environmental hazards, and inadequate infrastructure at resettlement sites (Coelho, 2016; Williams et al., 2022). The respondents also consistently reported that distance from the city centre increased commuting costs, and the loss of established job networks made it difficult to sustain pre‐relocation employment. Furthermore, we show how livelihood disruption is mediated by gender, thus producing distinct trajectories for both men and women after resettlement. The non‐uniform spatial marginality across gender is presented in Table 2.

TABLE 2.

Gendered livelihood outcomes post resettlement.

Dimension	Men	Women	Gendered outcome
Pre‐resettlement employment	Better access to stable city‐based jobs and retention than women	Mixture of informal and domestic work	Gendered but spatially integrated
Post‐resettlement employment	Job discontinuity owing to long commute; withdrawal from informal/low status job	Increased informalisation; domestic work; home‐based enterprises; higher visible participation in workforce	Feminisation of distress labour
Response to income shock/disaster	Intermittent withdrawal; alcohol dependency in some cases	Intensified work, debt management, and asset liquidation	Women absorb shock
Asset liquidation	Limited direct asset ownership	Jewellery mortgaging; sale of subsidised assets	Feminised financial risk
Care burden	Minimal change	Increased owing to loss of kin networks and infrastructure gaps	Double burden of livelihood and care duties

Source: authors.

3.2.1. Displacement of work and care

For many households, the immediate impact of relocation was the breakdown of previously viable livelihood arrangements. Meena, a trained home care nurse, formerly lived with her husband in an informal settlement in central Chennai. Her spouse had a stationery shop in the same neighbourhood. In 2016, they were forced to relocate to Perambakkam by the government owing to the risk of flooding. Following their relocation, their household's income configuration was fundamentally altered. After the move, Meena withdrew from paid employment, a decision shaped by the combined pressures of childbirth, childcare responsibilities, and the increased distance from employment centres. Her husband initially attempted to sustain their pre‐relocation livelihood by commuting to the city and running the shop, even renting temporary accommodation near his workplace and travelling back on weekends. However, the financial strain of maintaining two households and the long commute eventually rendered this arrangement unsustainable.

A similar disruption was narrated by Geetha, who was formerly employed by a private company in central Chennai. She also had to quit her job after relocation. When Geetha and her husband were living in the informal settlement in central Chennai, they could go out and work, as her mother, who resided in the same neighbourhood, provided essential childcare. In the wake of the 2015 floods, though, they were relocated under the Chennai Rivers Restoration Trust scheme to Perambakkam, located approximately 28 kilometres from her previous home. Geetha's mother, though, remained near the city centre. Consequently, after moving to the resettlement site, Geetha found it logistically impossible to maintain both wage employment and domestic responsibilities without her mother's support, leading her to resign from her job.

The experiences of Meena and Geetha point to a broader process in which resettlement not only changes the physical space where people live but also reorganises social life, including often invisible gender responsibilities (Mezzadri, 2019; Fernandes et al., 2023). Therefore, the loss of extended kinship networks, combined with increased spatial distances, continues to shift disproportionately the burden of care onto women, constraining their ability to participate in paid work. In this sense, spatial marginality occurs not only because of distance, but also owing to the reconfiguration of care relations.

3.2.2. Infrastructure, time, and labour

Infrastructural constraints further entrench this dynamic. In Semmancheri, the time of water collection was fixed at anywhere between 5 a.m. and 4 p.m. Wealthier households usually invested in private electric pumps, which could store additional water in one's home. The women from lower‐income households, however, had to forego work opportunities to ensure household water security. In Perambakkam, despite in‐house water connections, the supply was limited, and each water tank was shared by three households. Consequently, women had to be present when the water was provided; otherwise, a neighbouring household could acquire their allocated share. This arrangement made water a persistent source of stress and, at times, generated conflict among neighbours. As in Semmancheri, it also hindered women's ability to take up or sustain work outside of the home or beyond the neighbourhood, as being away during water supply hours could directly affect their household's access to water. Coelho (2016) engages with this phenomenon by introducing the concept of ‘occupational domestication’ to describe how women in resettlement colonies such as Kannagi Nagar in Chennai are pushed out of formal employment into low‐paid, neighbourhood‐based, and home‐centred work. This transition mirrors patterns observed in our study, as several respondents experienced downward occupational mobility, moving from skilled or semi‐formal work to domestic labour, vending, or small‐scale informal enterprises.

Uzma, around 37 years of age, moved to the Perambakkam resettlement site with her husband and two school‐going sons after the 2015 floods. She was formerly employed as a helper at an elite crèche in the city centre, where she earned a steady income of INR 18,000, but after relocation, owing to care duties, especially stringent water collection timings, she transitioned to running a small ready‐made garment shop within the resettlement site. Reflecting on this tension, she said: ‘I was getting a decent salary in the city centre, but due to unreliable bus schedules, traffic, and frequent workplace delays, I often missed the opportunity to collect water on time. My neighbours used to take away the majority share of the water, and we ended up having to buy it, which became very expensive’. Her experience illustrates how infrastructural inadequacies intersect with gender norms, thus limiting women's mobility and pushing them into localised, informal livelihoods (Small and van der Meulen Rodgers, 2023).

Taken together, the experiences of Meena, Geetha, and Uzma reveal how peripheral resettlement constrained women's ability to remain in or access their former employment. This did not occur through a single pathway. Instead, the disruptions took different forms, including the loss of nearby caregiving support, greater distance from workplaces and family networks, and inflexible infrastructure and transport arrangements.

3.2.3. Gendered divergence in labour participation

Post resettlement, there has been not only widespread livelihood disruption but also gendered reconfiguration of labour. Across the 19 interviews, women were more consistently engaged in some form of income‐generating activity after resettlement, although much of this work was irregular, low paid, and spatially constrained. Respondents’ accounts also drew attention to the visibility of women in the settlement's everyday economy. As Mala, a domestic worker in Semmancheri, noted: ‘If you go from here in the morning to the bus stand, you will find 75 per cent are women and 25 per cent men. Even now you can see that the women have gone to work and will come only in the afternoon, while the men are idling away on the streets’.

Lakshmi's account illustrates this change at the household level. Before relocation, she worked as a skilled X‐ray technician in a major hospital and supplemented her income by providing private acupressure therapy. Relocation made her commute untenable, though, forcing her to shift to low‐wage domestic labour within the neighbourhood, as with Meena, Uzma, and Geetha. Eventually, through this work, she became involved with the Domestic Workers’ Union, where she now holds full‐time employment. Her story stands out, though, with respect to how her spouse experienced a different trajectory in relation to livelihood opportunities after resettlement. Lakshmi's husband remained unemployed for approximately two years after relocation. She remarked:

When we were in the city, our husbands [referring to families displaced in resettlement sites] were working, but after moving here, the distance became a significant barrier. First, the commute is long. Second, they can't find the same type of work they had before. Third, there is simply no money to cover the cost of commuting.

Men's pre‐relocation livelihoods were often tied to specific locations, skills, and networks that were more directly disrupted by resettlement. Several respondents reported that male household members struggled to secure comparable work owing to the distance from employment centres, high commuting costs, and the erosion of labour networks. This aligns with the findings of other studies that show how spatial marginality weakens access to labour markets (Coelho, 2016; Bhide et al., 2019). In contrast, women's participation in the labour force increased in relative terms, but largely within the confines of precarious, low‐return activities located in or near the resettlement sites, such as domestic work, petty vending, and home‐based enterprises (Coelho, Venkat, and Chandrika, 2012). This shift reflects not simply greater opportunity for women versus men, but a complete reconfiguration of household survival strategies under constraint. Women engaged in work which was more adaptable to spatial immobility and care responsibilities, thereby allowing them to participate in income generation without incurring high commuting costs or violating gendered norms around mobility. At the same time, this ‘adaptability’ came at the cost of lower earnings, limited security, and intensified workloads.

Hence, the increased participation of women should be interpreted cautiously. Rather than treating women's greater visibility in paid work as evidence of straightforward economic empowerment, it can also be understood as a response to changing household circumstances. It reflects what Kabeer (2008, p. 28) defines as the ‘distress sale of labour’, where women's increased engagement in paid work is driven by necessity in the context of household economic strain. The higher visibility of women in income‐generating activities, therefore, does not signal improved livelihood security, but rather a maldistribution of economic and care burdens within households under conditions of constrained recovery.

3.2.4. Debt, assets, and gendered financial risk

Lastly, the fragility of livelihoods impacts financial stability. In interviews, households reported repeated cycles of borrowing, often triggered by overlapping crises such as floods, illness, and job losses. During the COVID‐19 pandemic, floods, and other recurrent disasters in resettlement sites, residents got access to food kits from the government and NGOs. However, they were still dependent on external borrowing for children's education, daily expenses, in order to restart livelihoods (buying goods owing to damage caused by floods), and for repaying previous loans, as seen in Table 1. This also indicates that recovery was continually deferred rather than secured.

A striking feature of these coping strategies is the central role of women's assets, particularly jewellery, as a financial buffer. In this study, 8 of the 19 women interviewees reported pawning their jewellery when household income was insufficient to meet expenses or other financial obligations. Most of the women had access to relatively small quantities of gold (like a thin chain or a pair of bangles or earrings, which they usually wore regularly). They reported that pawning gold was a last option, when other sources were unavailable or insufficient. Importantly, jewellery was not permanently relinquished during periods of crisis. Instead, women described regular efforts to redeem their gold through monthly repayment schemes, often at high interest rates. This produced a cyclical process wherein gold jewellery moved between being worn as an everyday personal asset, being pledged to secure short‐term liquidity, and being recovered once the household was able to service the debt.

From an SLF perspective, this phenomenon has an important implication for understanding women's control of assets. Rather than treating women's access to gold as straightforward evidence of economic empowerment, the cyclical pawning and redemption of jewellery reveal a more ambivalent type of asset‐based security for women. This pattern must be comprehended within the broader South Asian context, where gold functions not only as a culturally embedded form of status and wealth but also as a gendered reserve of value, accumulated through dowry and life‐cycle transfers (Mehrotra, 2004). Gold is usually understood as streedhan (translated as women's wealth), an asset culturally and materially associated with women's security, while noting that control over it may in practice extend to husbands and affinal family members (Harriss‐White, 2024). Once married, the jewellery was commonly viewed as available for meeting family needs and could thus be mobilised when the household experienced financial difficulty arising from illness, disaster, or an income shock (Amirthalingam and Lakshman, 2010; Zulfiqar, 2022; Harriss‐White, 2024).

In this sense, the capacity of the household to absorb a livelihood disruption is partly dependent on women's ability to convert their personal or gendered assets into liquidity. The financial burden of recovery is therefore not distributed evenly across household members. In this study, we have seen that women contribute not only their labour and care work but also the assets over which they are culturally recognised as having a dominant claim. Harriss‐White (2024) similarly argues that women often bear a ‘treble burden’ as they are simultaneously managing unpaid care work, engaging in income‐generating activities, and assuming responsibility for any crisis, including the repayment of multiple and overlapping loans. In the context of pawning of gold, it is important to examine critically how asset sustainability is understood within SLF at the household level. Such practices reveal how women mobilise assets to sustain household livelihoods, often by giving up control over scarce resources that they own. Thus, treating women's assets simply as shared household resources can obscure the unequal power relations through which they are accessed, managed, and converted into other forms of livelihood support. In addition, this also overlooks the gendered costs involved in sustaining household livelihoods.

3.3. Care, commoning, and the reconstitution of livelihoods

Building on the earlier discussion of how failed state interventions produce gendered spatial marginality, this subsection appraises how women reconstitute their livelihoods in the long term.

Recovery policies have largely failed to create stable and contextually viable livelihood pathways. At the same time, peripheral resettlement has intensified women's economic precarity while increasing their care responsibilities. However, these conditions did not translate into passive vulnerability, but instead shaped active and adaptive responses in everyday life. Women from the resettlement sites organised their everyday life in a way beyond individual coping, pointing to different means of care and commoning through which resources, risks, and responsibilities are collectively negotiated (Federici, 2019).

3.3.1. Extended spaces for livelihood and sharing care responsibilities

In the resettlement areas of Semmancheri and Perambakkam, Nadaipathai Kadai represents one of the most common and significant livelihood strategies (Goswami, Patil, and Raju, 2026). Many of these pavement shops are built informally by residents themselves, especially women, as vital responses to the scarcity of formal job opportunities. They often take the form of makeshift structures such as raised platforms made from tin and asbestos sheets, mobile carts, extended tin roofs attached to homes, or the temporary conversion of part of a room for work during the day (Goswami and Raju, 2026). These informal shops commonly sell a variety of items, including jewellery, household goods, stationery, groceries, and fresh produce.

Importantly, these extended shops were never included in the official design or planning of the resettlement sites during the recovery phase. They emerged out of sheer necessity as residents sought ways to sustain their livelihoods and ensure economic survival. They represent what Federici (2019) conceptualises as the production of commons: shared, collectively negotiated resources that enable survival outside of formal market and state structures. These self‐made shops are built incrementally by residents, reflecting the process that urban scholars call ‘autoconstruction’ (Holston, 1991; Caldeira, 2017). They require material and financial resources acquired over a period of time. Thus, they constitute a coping mechanism through which women navigate their new socio‐economic reality when formal recovery mechanisms or state interventions have failed to offer viable alternatives.

However, the uses of Nadaipathai Kadai are deeply layered. These self‐constructed livelihood spaces also function as sites of care, revealing the inseparability of income generation and social reproduction through building commons (Federici, 2019). Meena, for example, manages her shop while having her child with her after school hours. On several occasions, the first author observed the child eating, studying, and even falling asleep in the shop while Meena attended to customers and carried out other tasks. Meena also mentioned how other women in the area often leave their children with her when they need to run errands or attend to emergencies. As she explained: ‘I don't mind looking after their children in my shop. It helps build goodwill, and more people come to buy from me’. Similarly, Uzma also constructed an extended shop close to her home as a way to balance her caregiving responsibilities and means of livelihood.

Feminist scholars have emphasised that social reproduction, that is, the labour of maintaining households and communities, is central to economic life, even though it is systematically undervalued (Mitchell, Marston, and Katz, 2003; Federici, 2019). In resettlement contexts, this labour becomes even more visible as traditional infrastructures of support, such as proximity to kin networks, childcare, and reliable services, are disrupted. Extended shops in these settlements therefore serve as an important livelihood strategy, enabling women to combine their caregiving responsibilities with income‐generating activities. Yet, they often do so under difficult and precarious conditions.

3.3.2. Rebuilding social reproduction through kinship and solidarity

Structural precarity within resettlement sites often places a disproportionate burden on women. Over the long term, women have navigated their livelihood and care responsibilities by consolidating resources using digital platforms such as WhatsApp groups and by strengthening kinship ties, particularly with female relatives and friends. These networks function as important channels for sharing information, coordinating collective action, and providing mutual support and solidarity. Such forms of collective organisation extend Federici's (2012) means of collective reorganisation of reproductive labour in contexts of crisis.

Amina, a single mother of four, lost her husband due to complications related to chronic alcoholism. Confronted with the challenges of managing full‐time caregiving alongside paid employment, she arranged for her mother to move in with her in the next tenement. Amina covers her mother's rent and other housing expenses. Reflecting on this decision during an interview, Amina shared:

My parents were in Aminjikarai [in the city centre about 30 kilometres away from the resettlement site]. My father passed away first in 2020. That year, three people in the family died. First my father, then my father‐in‐law, and soon after, my husband. My mother, who was living alone in Aminjikarai and working as a domestic worker, began to suffer from diabetes‐related health issues and had frequent falls and injuries. I couldn't leave her there alone. I also needed her support, and she needed mine, so I brought her here.

This arrangement allows Amina to pursue informal income‐generating activities outside of her house while her mother manages caregiving duties such as childcare, water collection, and daily household chores. Similarly, Chitrakala, a domestic worker, was forced to leave her job after relocation to take care of her two young daughters. To resume paid work, she invited her mother to move in with them. Chitrakala explained that she felt more at ease going to work after her mother moved to take care of her children, as she considered the resettlement site to be ‘unsafe’.

These examples demonstrate that livelihood recovery is not an individual achievement but a relational process, dependent on the reconstitution of care networks. In Tronto's (2013) terms, these practices reflect ongoing processes of ‘caring with’, where collective responsibility is encouraged for sustaining life under precarious conditions.

Women within these resettlement sites are also part of several networks, unions, and NGO‐run support groups. These entities facilitate access to informal job opportunities, vocational training, and other resources. A number of our respondents were members of a prominent network called the Domestic Workers’ Union. Several (including Meena, Lakshmi, and Geetha) leverage their membership to access not only employment opportunities but also capacity‐building programmes. These bodies offer numerous benefits, such as financial literacy training, daycare facilities, and information about job openings within and without the resettlement areas.

WhatsApp has also become an important tool through which women who are part of support groups navigate livelihood opportunities. It enables the rapid circulation of information about domestic work, informal employment, and business opportunities. Many women interviewed for our study are actively engaged in multiple support groups. Shalini, a member of a self‐help group, stated: ‘I get daily messages from about 8 to 10 WhatsApp groups’. WhatsApp groups also facilitate the dissemination of emergency information, such as flood warnings or pandemic safety protocols, allowing women to make timely decisions that directly impact their livelihoods. For instance, Amina, a food vendor (who sells cutlets within the resettlement site from a movable vehicle), belongs to three WhatsApp groups that help her get orders. Furthermore, through one of the WhatsApp groups, she received early disaster warnings about the intensity of the 2023 floods in Chennai. She immediately relocated her stall from the street near her home to a higher elevation area within the resettlement site. This prevented her from incurring significant financial losses.

Such platforms also enable women to share knowledge, offer advice, and support one another, creating an informal but highly effective means of collective survival. According to the respondents, NGOs, trade unions, and self‐help groups disseminate information and communicate with their members through WhatsApp channels.

Women's organisations also extend to addressing safety concerns and infrastructural deficits within resettlement sites. Women's groups share information about unsafe areas, monitor one another's workplaces, and coordinate responses to harassment and resource shortages. The Domestic Workers’ Union and other community actors also share resources and information on workers’ rights to support women engaged in informal and precarious livelihoods. In Perambakkam, a community mapping exercise conducted by the Information and Resource Centre for the Deprived Urban Communities documented residents’ collective knowledge of unsafe areas. This contributed to subsequent government efforts to improve local infrastructure (The Times of India, 2023). Respondents also described everyday practices of mutual support, including warning one another about dangerous areas, watching over each other's stalls, and rotating caregiving responsibilities.

In sum, our data underscore that women in Chennai's resettlement sites have been able to negotiate marginally the structural vulnerability imposed by their spatial and socio‐economic conditions through dynamic processes of collective agency, mutual support, and adaptive innovation. These collective strategies not only facilitate livelihood recovery but also reconfigure care and labour practices in the peripheries.

4. CONCLUSION

Post‐disaster resettlement in Chennai did not simply move people from one place to another; it reorganised the terms on which life itself could be sustained. Communities once embedded within dense urban economies were displaced to peripheral areas where distance, weak infrastructure, and fractured social networks rendered livelihood arrangements untenable.

It is within this terrain that gender‐inclusive recovery programmes have been positioned as progressive. Yet, our study shows that, owing to structural conditions of peripheralisation, those efforts fail in the long term. The challenge is not merely that such programmes fall short of context, but that they are asked to function within spaces where the very possibility of viable livelihoods has already been undermined. Distance from labour markets, the absence of supportive infrastructure, and persistent environmental risk do not just constrain economic activity, they reshape the social organisation of work itself (Coelho, Venkat, and Chandrika, 2012; Desai, Parmar, and Mahadevia, 2019; Parpiani et al., 2026). In these settings, gender does not operate as an isolated axis but is folded into the spatial and political logics of resettlement, intensifying women's burdens even as they are targeted as agents of recovery.

Situated within a longer temporal arc of disaster and resettlement, we moved beyond disaster event‐based understandings of livelihood recovery in the peripheries. Via this study, we emphasise that recovery is continually negotiated through everyday life in resettlement. Women, in particular, stitch together fragile forms of livelihood through practices that exceed formal economic categories by extending homes into sites of production, circulating care across kin and neighbourhood networks, and participating in collective arrangements such as self‐help groups and unions. These are not supplementary activities; they are the ground on which everyday survival is organised. Analysing these practices through Tronto and Fisher's (1990) theory of caring lens, we find that livelihood recovery is not merely assembling income or assets but is an ongoing effort to sustain life through shared childcare, reciprocal labour, and informal reworking of built environments. These forms of commoning (Federici, 2019) constitute a quiet but persistent reorganisation of urban life in the periphery, which the state or the market failed to encompass in these spaces.

Reflecting on other livelihood recovery literature, combined with our data, we find that recovery cannot be grasped in the immediate aftermath of a disaster (Daly et al., 2020; Karki, Matthewman, and Grayman, 2022). It reveals itself over time, in slow and uneven processes through which households attempt to rebuild the conditions of viability. But in peripheral resettlement sites, even these long‐term efforts remain precarious, continually shaped by the very structures that produced displacement (Goswami and Raju, 2026). While disaster scholarship often suggests that resettlement can foster new forms of social capital and collective agency (Nakagawa and Shaw, 2004; Aldrich, 2012), our study shows that such agency is not unbounded; rather, it remains relational and constrained, shaped by spatial marginality, entrenched inequalities, and uneven access to markets and state support. Agency, therefore, is neither absent nor emancipatory; it is relational and negotiated.

Finally, by foregrounding care and commoning, we do not intend to romanticise coping or resilience practices. Rather, the paper advances an alternative analytic of recovery that takes into account the collective labour required to sustain life under conditions of marginality. It also points to the limits of policy frameworks that prioritise material reconstruction while neglecting the social and care infrastructures on which livelihoods depend. If recovery is to be meaningful, it must move beyond the provision of housing and short‐term employment schemes, and instead engage with the relational, spatial, and gendered conditions that make life viable in the long term. Our study suggests that gender‐sensitive recovery must surpass the rhetoric of inclusion and engage more deeply with how women collectively survive, adapt, and develop strategies to sustain their livelihoods through repeated cycles of disaster.

CONFLICT OF INTEREST STATEMENT

No potential conflicts of interest reported by the authors.

Supporting information

Data S1. Annexure 1. Respondent profile.

DATA AVAILABILITY STATEMENT

The data that support the findings of this study are available from the corresponding author upon reasonable request.