1. Introduction
Recent forecasts from the National Oceanic and Atmospheric Administration and European Center for Medium-Range Weather Forecasts warn that temperatures over the next two years could reach record highs, with El Niño conditions intensifying drought, flooding, disease prevalence, and food insecurity.1 As with other climate extremes, the WHO and U.N. warn that the impacts will be felt most acutely by young children, older persons, and people with disabilities, while simultaneously disrupting health services, schools, and other essential systems. In many of the regions expected to be hardest hit by El Niño, including sub-Saharan Africa, Southeast Asia, and South America, formal services and infrastructure for care of these populations are already limited, despite their importance for building community resilience, response capacity, and recovery capabilities. Yet formal care services infrastructure remains largely absent from climate adaptation planning, including in National Adaptation Plans (NAPs) and Nationally Determined Contributions (NDCs). As countries prepare for the 31st Conference of the Parties to the U.N. Framework Convention on Climate Change (COP31), integrating care services into adaptation planning can strengthen resilience efforts.
Until the adoption of the Gender Action Plan (GAP) at COP30, the topic of care work has largely been absent from the United Nations Framework Convention on Climate Change (UNFCCC). Even within the GAP, recognition of care work remains limited, confined to brief references in the text. It thus provides little operational guidance on how to address the everyday care needs and responsibilities that structure vulnerability, time use, and resilience to disasters. The omission of care services is also reflected in countries’ National Adaptation Plans and Nationally Determined Contributions, which are the primary vehicles through which climate efforts are translated into investments and activities.
Against this backdrop, this commentary seeks to answer to what extent adaptation plans currently address care services infrastructure as a component of climate resilience. We conducted a qualitative assessment of NAPs and NDCs to determine how well they recognize the care services infrastructure in their planned activities. This analysis was motivated by background research associated with forthcoming work by Grown, Ilkkaracan, and De Henau. The authors argue that effective climate adaptation depends fundamentally on the resilience of a comprehensive care services infrastructure—one that ensures quality, universal access, decent jobs, and affordability—which can strengthen the resilience of families and communities before, during, and after climate shocks. According to Grown et al., a comprehensive care services infrastructure comprises both social infrastructure and physical infrastructure. Social infrastructure includes the organization and labor (paid and unpaid) required to deliver care services [e.g., early childhood education and care (ECEC), long-term care (LTC), healthcare, education]. It also includes elements needed for continuity and quality, such as workforce planning, protocols and coordination mechanisms, caregiver and care recipient training and preparedness, registry tools, and monitoring arrangements. Physical infrastructure refers to the built environments and enabling inputs where care occurs—care facilities (e.g., childcare centers, schools, day care centers) and home settings—plus climate resilient upgrades and supporting systems required for operations [e.g., cooling, flood drainage, water, sanitation, and hygiene (WASH), nature-based solutions, and mobile units to maintain access when fixed facilitates are disrupted].
Across NAPS and NDCs, references to care services infrastructure are fragmented, rather than acknowledged as a coherent, encompassing system explicitly recognized in adaptation planning. This has translated into neglect of ECEC and LTC, with only the Philippines referencing day care centers, St Lucia referencing nursing homes, and St. Lucia, the Philippines, the UAE, and Kiribati planning to train or expand the long-term care workforce. Among the sectors that make up a comprehensive care services infrastructure, education and health receive the strongest attention, but commitments tend to prioritize infrastructure upgrades and disease control over the systems and workforces that sustain care. A small group of plans, particularly from Argentina, Bangladesh, Kiribati, Nepal, and the Philippines, incorporates more elements covering shelters, schools, health systems, and social protection with attention to the population segments that comprise carers and care recipients. Even in these cases, however, plans rarely acknowledge the interdependence of care systems; for example, how unpaid carers routinely compensate when health and education services are disrupted during crises.
The next sections summarize the findings of our analysis, grouping them within six sectors: 1) early childhood education and care; 2) long-term care; 3) education; 4) health; 5) disaster risk reduction (DRR); and 6) housing, paying special attention to activities that center the population segments which comprise care recipients (including children, older persons, and people with disabilities) as well as caregivers and the care workforce. Sector-level trends and illustrative examples and outstanding gaps are highlighted. The commentary concludes with six overarching recommendations to improve the integration of care across adaptation planning.
2. Methodology of NAPs and NDCs analysis
To answer the research question of to what extent care services infrastructure is addressed in climate adaptation plans, the analysis focuses on Non-Annex I Parties under the UNFCCC: a group of 155 predominantly low- and middle-income countries recognized as particularly vulnerable to climate impacts. For each country, one primary adaptation document is identified: either the most recent NAP or the most recent NDC with substantive adaptation content, applying simple decision rules when multiple versions exist. At this stage, we restricted the analysis to documents available in English. Of the 103 English-language adaptation documents, six were not machine-readable. The final sample comprises 97 NAPs and NDCs. As a result, our analysis is limited in terms of geographic generalizability (see Table 1).
Within each document, the unit of analysis is the adaptation activity or action, including narrative descriptions, lists of measures, and sectoral programs in adaptation-relevant sections. This means context sections from planning documents were omitted to focus solely on countries’ intended actions. Using qualitative analysis software (Atlas.ti), keyword searches were run to extract and tag full sentences or paragraphs to relevant themes.
A structured codebook was used to identify key search terms related to social and physical infrastructure, organized into seven thematic clusters: general care terms and in-home care;2 early childhood care; health and hospitals; schooling, education, and training; long-term care and ageing; social protection and response; and general physical infrastructure (such as shelters, community facilities, housing, and informal settlements).
Segments were manually reviewed to remove false positives and reassign ambiguous cases, ensuring that only genuine references to care systems were retained. During this stage, actions were further manually categorized by physical or social infrastructure, following the Grown et al. methodology, and into sub-categories. This approach enabled mapping of how adaptation plans recognize care recipients, carers, and care-related services infrastructure.
3. Results
Overall, 85% of the 97 plans include at least one reference to care recipients, the care workforce, or care-related infrastructure and services. There is considerable variation among plans in the breadth of care-related activities, ranging from 14 plans with no care-related activities to 130 care-related activities in Kiribati’s NAP.3 This pattern suggests that efforts to incorporate care often remain at the level of recognition: many countries acknowledge care in principle, but far too few translate it into sustained, multisector actions.
Findings were further disaggregated among country income groups and regions to elicit high-level patterns. High-income countries have the lowest coverage, with 78% of plans referencing care and the fewest number of activities per plan on average at seven. Reflecting this pattern, it was difficult to identify more than a few illustrative examples of best practices. Even countries known for their expansive caregiving policies have few mentions of care in adaptation plans (e.g., Singapore, with zero activities, and South Korea, with four activities). Lower-middle-income countries perform the best at 90% coverage and 24 activities per plan on average. However, coverage and breadth do not always coincide; for example, in ECA, 91% of plans reference care but have the lowest average number of care activities per plan at seven. South Asian countries average 48 activities per plan. However, notably India’s NDC has zero care references. The map below displays the countries with plans that include care activities, color-coded by the number.
Since caregivers are predominantly women and girls, and care recipients include children, older people, and people with disabilities, these population segments were included in search terms. Among plans, one-third mention specific population segments in their care-related activities, ranging from one activity in 15 plans4 to 65 activities in Bangladesh’s NAP. Population segments are routinely grouped together as “at-risk” and “vulnerable groups.” Of the 398 activities that reference a population segment across all plans, one-third generally note recognizing and including these groups in climate action, but unlinked to a specific care sector (health, education, LTC, ECEC, DRR). Activities linked to specific care sectors, directed to specific care recipients or care workers, are discussed further below.
When population segments are mentioned, they are rarely referred to in the context of needing or providing care. A key finding is that care work is invisible and largely detached from activities that mention children, people with disabilities, and older adults. For example, the term “caregiver” is entirely absent from plans, and “carer” is used once, in Kiribati’s NAP. Some plans contain a few tacit references to women’s “informal roles” or “spheres of influence.” Suriname and Thailand are the only countries to mention unpaid care work. Regardless of the sector, no plan contains a mention of decent pay or working conditions for care workers; their inclusion in planning is focused on what they can provide for others as opposed to what they should receive. Finally, the plans that use the term “gender-responsive” do not unpack what it means, for instance, promoting decent conditions for care workers or targeted financial or other support for unpaid women caregivers in climate-vulnerable areas. We discuss this further in Section 6.
3.1 Early childhood education and care
For the ECEC sector, the review specifically looked for adaptation measures that strengthened both physical and social infrastructure supporting children from birth to age five. On physical infrastructure, attention was given to climate-resilient upgrades or expansion of facilities such as crèches, preschools, and daycare centers. On social infrastructure, the analysis sought to identify activities aimed at building preparedness and resilience among young children and their caregivers, including facility staff, family members, and domestic workers. This included DRR planning tailored to ECEC settings, hazard-specific protocols, caregiver training, joint preparedness exercises, and awareness-building initiatives designed to protect children’s safety, health, and development.
ECEC is absent from 96 of 97 plans: Only the Philippines explicitly references day care centers for displaced families. Children are mentioned only in general terms in 29 plans. Half of the plans from South Asia mention children, compared to 15% and 18% in Europe and Central Asia (ECA) and Middle East and North Africa (MENA), respectively. The Nepal and Bangladesh plans have greater breadth, each with 20 and 17 child-related activities, respectively. Only 30% of plans from SSA and 20% of plans from low-income countries mention children, which is alarming given that many of these countries are still in the high-fertility phase of the demographic transition and time spent on unpaid care work is extremely high in these contexts.
In the plans that mention children, age groups are not distinguished among them, despite the needs during early childhood (0-to-3-year-olds) being vastly different from those of 7-to-10-year-olds. This matters for the care workforce. Provisions for children tend to prioritize safeguarding and awareness over the caregiving systems that enable children’s well-being—such as accessible childcare, early childhood development services, and support for caregivers in crises. Future plans could commit to expanding climate-resilient childcare, especially in high-risk areas; integrating childcare into resettlement and shelter plans; subsidizing childcare after climate shocks; and training early childhood workers in disaster preparedness and psychosocial support. Nature-based solutions, such as urban tree cover, community gardens, and green walls, are also recommended for day-care centers and playgrounds as they have both cooling properties as well as positive psychological effects on children. Both ECEC workers and the parents of young children were absent from the plans, but ideally, actions can align their training to ensure there is collective knowledge on evacuation plans and emergency communication protocols.
3.2 Long-term care
For LTC, the assessment focused on identifying adaptations that support older adults and people with chronic disabilities across institutional, community, and home-based settings. Physical infrastructure adaptations of interest included climate-resilient upgrades and expansion of adult daycare centers, residential care facilities, community centers, and nursing homes. In terms of social infrastructure, the review looked for resilience-building activities – such as caregiver and care recipient training, including emergency preparedness planning, coordination mechanisms, and continuity-of-care protocols intended to ensure safe and dignified support during climate-related disruptions- directed at both care recipients and their caregivers (e.g., formal care workers, family members, and in-home aides).
Only four plans explicitly discuss LTC: St. Lucia, the Philippines, the UAE, and Kiribati, all of which refer to training the long-term care workers. Of those, only St. Lucia commits to expanding the LTC workforce, specifically in nursing homes during heat waves. Many activities that feature older people and people with disabilities list them with “at-risk” or “vulnerable” groups. Activities identify them as target populations to receive health monitoring and information, early warning information, and relief and rescue support.
Fifty-nine activities mention older people across 24 plans. Half the plans from South Asia and nearly half the plans from LAC mention older people. Less than 20% of plans from ECA, MENA, and SSA, as well as low-income and high-income countries, mention older people. A third of the activities that mention older people focus on health, while another third are general commitments. Plans that specify why and when these populations requiring LTC need coordinated support are more actionable: Belize notes that older people may be less digitally literate and need assistance accessing new electric-vehicle bus schedules; South Korea notes older people’s vulnerability to heat and cold waves and suggests social protection measures; and the Philippines recommends door-to-door primary care since older people are often confined to the home. Future plans would benefit from explicitly mentioning adaptation initiatives at LTC sites of care, such as passive cooling and ventilation systems as well as self-sustaining emergency preparedness infrastructure, especially given that older people and people with disabilities face acute vulnerability to heat and mobility issues.
Eighteen plans mention activities for people with disabilities, comprising ninety-four activities. Half the plans from South Asia mention people with disabilities, compared to a quarter in LAC. Less than 10% of plans from MENA, ECA, and high-income countries mention people with disabilities. A third of the activities that mention people with disabilities focus on DRR, while another third are general commitments. Activities are dominated by Bangladesh’s NAP (25) and Kiribati’s NAP (19). Plans that advocate for their participation in climate adaptation include Kiribati (awareness-raising activities and DRR plans for people with disabilities) and Niue (involving people with disabilities in national climate decision-making processes and ensuring appropriate support networks are in place for their families and communities).
3.3 Education
Because many adaptation plans refer to learning systems without specifying a setting, the education category intentionally captured measures applicable to both formal environments (such as schools) and informal platforms (including public education campaigns and youth programs). For social infrastructure, the review looked for activities that build knowledge and capacity of both teachers, aides, parents, and students, such as disaster preparedness education, climate change literacy, climate-health education, and training for climate-related livelihoods. For physical infrastructure, the focus was on identifying adaptation to educational facilities, particularly climate-proofing of schools, as well as alternative education models like tele-education or floating schools that enable learning to continue when traditional education settings are affected by climate impacts.
Two-thirds of plans address education. Each region features over 60% of coverage among its plans. South Asia has the most coverage with 83% of plans including education, followed by SSA with 70%. Lower-middle-income countries perform best in terms of breadth, 10 activities on average, and coverage, with 69% of plans including education activities. Although 67% of high-income countries feature education, they perform the worst, with plans featuring 3 education activities on average.
Many countries lack well-functioning, accessible formal education systems. They also need widespread dissemination of climate risk and adaptation information. Against this backdrop, it is striking that only five countries propose expanding education infrastructure—and none propose expanding the teacher workforce. There are references to primary school, junior high, and college among activities, but notably no mentions of pre-primary education. After-school education is mentioned in the context of clubs to promote climate awareness (Bhutan, Chad, Liberia, Mozambique, and South Sudan) but doesn’t recognize the broader care inherent in after-school programs.
Twelve plans include teacher training, 42 curriculum updates, and 19 school infrastructure upgrades. Interventions include integrating adaptation and DRR into curricula, providing professional development for teachers, and investing in resilient school infrastructure and WASH facilities. A few designate schools as resilience hubs that double as shelters and bases for nutrition, psychosocial support, or child protection.
Unsurprisingly children and youth are prioritized in education initiatives, while older people (two activities) and people with disabilities (nine) are rarely targeted. Kiribati is an exception, embedding “disability-inclusive pedagogy” into teacher professional development. It also specifies that “school retrofit plans should apply gender-responsive and universally accessible design,” operationalized through Red Cross guidelines. Somalia proposes nine activities, advocating for actions such as integrating climate literacy into school curricula and vocational training, and emphasizing gender equity and disability inclusion. Only Georgia’s NDC explicitly notes women’s overrepresentation in teaching and frames them as key agents of behavioral change.
3.4 Health
In the health sector, the review sought to capture a broad range of facilities, services, and personnel involved in care delivery. Physical infrastructure measures included climate-resilient upgrades or expansion of hospitals, community health centers, mobile clinics, and other frontline health care sites. For social infrastructure, the analysis looked for adaptations that strengthen the health workforce, such as doctors, nurses, midwives, community health workers, and volunteers, through measures including workforce expansion and plans for surge forces, specialized training, and disaster and disease preparedness.
Health measures appear in two-thirds of the plans. In every region, at least 60% of plans include health provisions, with coverage exceeding 75% in MENA and LAC. South Asia stands out, with health provisions included in 83% of plans and an average of 21 health-related activities per plan. By contrast, low-income countries show the lowest coverage, with only slightly more than half of plans including health provisions.
As for activities, 35 plans mention training health workers; six aim to expand the health workforce; and 38 propose hospital upgrades. Hospital upgrades include a variety of measures, including safe storage of medical equipment, improving waste management, strengthening WASH systems, and connecting energy systems to solar grids. Only St. Lucia explicitly addresses surge capacity in case of disasters. Future plans could consider reserving buildings to house health care workers in the event that disasters prevent commuting. While many countries rely on the unpaid labor of community health care workers, only Thailand recognizes this as unpaid work, referring to developing the capacity of “volunteer health workers.” Research, surveillance, and early-warning systems appear in 49 plans; disease control in 52, public awareness in 26, and access and engagement strategies in 17. Only Bangladesh’s NAP includes the use of mobile/floating health care units, although more recent plans mention telehealth opportunities to increase service engagement.
Three countries connect household care roles to public health outcomes. Cambodia and Georgia note women’s responsibility over household medical care and influence over behavioral practices that can improve health and hygiene. Suriname similarly highlights the formal and informal roles of women and skills and training for them to advance climate adaptation objectives in their spheres of influence. Future plans can prioritize joint capacity building across the full care continuum, including paid (health) care workers and unpaid in-home carers (see Grown et al., forthcoming).
Pregnancy, the postpartum period, and infancy are critical stages in the life course during which care demands intensify. Recognizing these activities in adaptation plans for the health sector is important since sexual and reproductive health are often detrimentally disrupted by crisis. Only 13 countries directly address sexual, reproductive, maternal, neonatal, and infant health. Examples include commitments to maintain sexual and reproductive health services and nutrition for displaced populations, to integrate reproductive health into climate interventions, to improve infant and child feeding programs, or to provide maternity and lactation facilities in shelters. The Seychelles’ NDC includes an activity to develop pre-service midwifery training that incorporates climate risks and impacts.
3.5 Disaster Risk Reduction
Within DRR planning, the review focused on whether adaptation efforts accounted for populations requiring or giving care. The analysis looked for evidence that physical infrastructure, especially emergency shelters and evacuation systems, was designed to be accessible and responsive to the needs of these groups. In addition, the review searched for social infrastructure measures that promote preparedness, such as targeted early warning systems, tailored risk communication, and coordination mechanisms that ensure caregivers and care recipients are adequately supported before, during, and after climate-related crises.
Twenty-one plans address care within DRR activities. Seventeen plans discuss evacuation shelters, with 10 planning to expand the number of shelters and eight planning to upgrade the infrastructure of existing shelters. St. Lucia commits to constructing new shelters on higher ground while Pakistan, Namibia, Myanmar, and the Maldives focus on multi-purpose shelters. Seven plans link shelters to gender and social equity, calling for accessible evacuation centers, safe spaces for women and children, or cooling shelters for older people and low-income households.5 Bangladesh combines new construction and rehabilitation with ‘gender- and disability-friendly’ features, including WASH, renewable energy, lactation spaces, and access roads; similar integrated approaches appear in the Philippines and Nepal NAPs. Nepal also recommends developing and training with a roster of diverse service providers, like community women’s groups, counselors, and health care providers, that can be immediately mobilized.
To strengthen emergency infrastructure as part of the care services infrastructure, future plans would be more effective with minimum care-responsive design standards (e.g. childcare spaces, breastfeeding and privacy spaces, accessible WASH, and assistive technologies); link relocation and upgrading programs to on-site access to health and social services; define shelter coverage targets with high care needs; and budget for maintenance and staffing, including support for community-based carers and flexible teams of trained care workers.
3.6 Housing
Recognizing the home as the primary site of caregiving, the housing category centered on identifying adaptations that strengthen household resilience and support unpaid care workers as a first line of defense against displacement and health emergencies. General urban planning measures were excluded to maintain a focus on the caregiving function of the household. For physical infrastructure, the review looked for upgrades to homes and informal settlements, including improvements to WASH, cooking technologies, and other in-home assets essential to caregiving. For social infrastructure, attention was given to behavioral interventions that promote climate-resilient practices within the home.
Housing is addressed in 160 activities across 39 plans. Unsurprisingly, low-income countries place the greatest emphasis on housing, with about half including housing references. Only a quarter of plans from MENA and ECA mention housing, while LAC and South Asia have the highest coverage, with 62% and 67% of plans mentioning housing activities, respectively.
Most activities focus on physical housing upgrades, risk assessments, or construction guidelines. Upgrade measures include green building designs, WASH improvements, ecosystem-based approaches, and resettlement away from flood zones. Nine plans have activities related to informal settlements or slums.6 Thailand’s NAP outlines 18 priorities under “Guidelines and measures for human settlement and security,” ranging from detailed actions to enhance accessibility for all population groups across warning systems, building codes, land-use planning, green and nature-based infrastructure, among others.
Activities that connect housing interventions to broader care systems, especially health and education, can strengthen NAPs. Two examples are illustrative. Argentina’s NAP notes that heatwaves increase hospitalizations and fatalities among “elderly individuals, children, people with chronic illnesses, persons with disabilities, and those in vulnerable situations – including those experiencing homelessness, living in inadequate environments, or belonging to indigenous communities” and therefore prioritizes public awareness alongside the adaptation and construction of resilient homes for these populations. South Korea’s NDC reaches a similar conclusion and prioritizes insulation programs for the homes of the elderly, low-income, and vulnerable groups.
