Fragile against what? Imagining climate change resilience by unmasking governance fragility Authors: Luissa Vahedi, Carmen Logie Published: 2026 ABSTRACT AbstractSince the 1990s, fragility has been used to describe low‐ and middle‐income countries (LMICs) with limited ability to respond to external threats in ways that ensure citizen safety, security, and welfare. Fragility challenges us to engage with governance, including why and how many institutions (including those in high‐income countries) fail to protect citizens during climate change. We ask the question: how can state fragility be applied, critiqued, and expanded when considering climate change, including anthropogenic drivers, extreme weather events, mitigation, and adaptation? We present a globally applicable framework of governance fragility and emphasise the role of institutional‐level gaps in motivation/willingness to act, evidence uptake, and implementation in creating fragilities that inhibit climate resilience. Using the United States as a case study, we apply the framework to identify the political ideologies that inform institutional‐level motivations, shape evidence uptake, and moderate implementation capacity. This expanded notion of governance fragility (beyond conflict‐affected LMICs) offers a globally‐relevant lens to imagine possible solutions, including a research‐to‐practice agenda that contends both with ideology/willingness to act and decentralised forms of climate governance in the face of national‐level fragilities.Keywords: climate change, disasters, equity, fragile state, fragility, governance, institutions INTRODUCTION AND BACKGROUND Strongmen disappear … knowledge that conflicts with their ideologies and goals … All twenty first century authoritarians suppress climate change science (Ben‐Ghiat, 2020, p. 96). Fragility is pervasive and resilient across various systems and subsystems and over time. Addressing it thus requires merging global and local perspectives to unpack variation and facilitate differentiated approaches, thereby reconciling complexity with simplicity (Desai and Forsberg, 2020, p. 10). The world is grappling with multiple intersecting crises that threaten human security (Landin Basterra et al., 2023; World Economic Forum, 2023; Lawrence et al., 2024). The link between responding to external threats and promoting population security and welfare lies in governance institutions, defined as the rules, norms, and strategies adapted by individuals operating within and across organisations to manage a country's affairs. Governance institutions at the national, sub‐national, and local levels contend with internal (authoritarian motivations/ideologies of political actors) and external (climate change) pressures that intersect in complex ways. Linkages between climate change and authoritarian motivations/ideologies extend across borders, posing population‐level consequences regardless of a nation's development status. Since the 1990s, international monetary and development organisations have used the term ‘fragility’ to describe national governments that are unwilling and/or unable to respond to population‐level threats in ways that protect citizen needs like safety and security (Ferreira, 2017; Carment and Samy, 2023). Traditionally, fragility has been applied narrowly to conflict‐affected low‐ and middle‐income countries (LMICs) (Fund for Peace, 2023; OECD, 2025; World Bank Group, 2026a). Yet, climate change, including its anthropogenic drivers (greenhouse gasses), mitigation (long‐term emission capture and reduction), adaptation (coping with and responding to extreme weather events via disaster risk reduction and long‐term climate adaptation), has exposed national governance fragilities in the ability to mobilise policy, science, and disaster response (IPCC, 2023). Thus, climate change has the potential to broaden the relevance of ‘fragility’ beyond conflict‐affected LMICs. In this paper, we engage with the following question: how can state fragility be applied, critiqued, and expanded when considering climate change, including anthropogenic drivers, extreme weather events, mitigation, and adaptation? The concept that a national government can be ‘fragile’ gained momentum following the terrorist attacks in the United States on 11 September 2001 (Grimm, Lemay‐Hébert, and Nay, 2014). State fragility served as a unifying theory and quantified metric (Fund for Peace, 2023) to inform international development policies related to humanitarian operations, health, education, and security (Carment and Samy, 2023; Fund for Peace, 2023). Organisations such as the World Bank (World Bank Group, 2026a), the Organisation for Economic Co‐operation and Development (OECD, 2025), and the Fund for Peace (2023) assign fragility scores and rank countries accordingly. Depending on the measurement tool and year, 50–60 conflict‐affected low‐income countries are labelled as fragile (Fund for Peace, 2023; OECD, 2025; World Bank Group, 2026a). For example, according to the OECD's multidimensional fragility index, 86 per cent of sub‐Saharan Africa's total population or almost 9 out of 10 people are living in conditions of moderate to extreme fragility (Desai and Forsberg, 2020). In the Middle East and North Africa region, 7 out of 21 countries are fragile, which represents 40 per cent of the regional population (Desai and Forsberg, 2020). In contrast, no high‐income country (HIC) is currently identified as fragile by the Fund for Peace, OECD, or World Bank. Fragility as a concept solely applied to LMICs experiencing armed conflict is being increasingly questioned by scholars and practitioners (Fund for Peace, 2023; Haken and Fiertz, 2023). The Fund for Peace (2023, pp. 9–10) notes: For all its strengths and weaknesses as a unified theory of international relations, in 2022 the focus on state fragility began to shift … this year's events undermined the facile notion that fragility is a threat that only spreads from poor countries. Instead, contagion went the other way round as war in Europe led to inflation, fuel riots, and food insecurity in vulnerable countries around the world. Further, we learned that the Great Powers, whether that be China, Russia, or Western democracies, may be more fragile than we think. Haken and Fiertz (2023) draw attention to the unwarranted exceptionalism concerning fragility in so‐called ‘stable’ nations. In HICs, fragility is thought of as a problem that must be halted abroad, before it spreads (Helman and Ratner, 1992; Hill, 2005; Saeed, 2023). Yet, HICs like the US exhibit different forms of fragility. Institutional and structural weaknesses in settings of masked fragility are hidden by national‐level wealth and power, rendering some HICs unable to withstand resiliently future social, economic, or political threats (Haken and Fiertz, 2023). This current understanding of fragility in HICs has yet to connect fragility to climate change. In line with other scholars, we argue that fragility should not be viewed as a concept applicable only to LMICs and conflict‐affected regions (Saeed, 2020; Haken and Fiertz, 2023). Rather, fragility can be situated as a norm on the international landscape, particularly when considering the threat of climate change drivers and extreme weather events as well as mitigation and adaptation responses (Saeed, 2020). We advance existing critiques (Grimm, Lemay‐Hébert, and Nay, 2014; Saeed, 2020; Haken and Fiertz, 2023) by conceptualising fragility as a governance endpoint wherein institutions facing threats (like climate change) cannot absorb population‐level harms while also promoting equitable outcomes among structurally disadvantaged groups. We draw attention to three interrelated mechanisms involving institutional‐level constraints that have the potential to create governance fragility in response to a threat: (i) failure to act; (ii) failure to consider evidence; and (iii) failure to implement. In our work, fragility is a globally applicable process situated at the core of whether and with what degree of success governance structures ameliorate the harm posed by threats like climate change, considering both mitigation and adaptation responses. Understanding governance fragility in this way can inform a new global research agenda at the nexus of (i) climate change, (ii) implementation science, (iii) governance, and (iv) public health. REIMAGINING FRAGILITY AS A GLOBALLY‐APPLICABLE PROCESS Figure 1 depicts our globally‐relevant fragility framework, which we explain and apply in the following subsections. In summary, the model has two levels: (i) institutional‐level mechanisms (defined by motivation, evidence acceptability, and implementation) that shape how institutions respond to a threat; and (ii) governance endpoints that are classified according to population‐level threat protection and equity. Climate change governance fragility framework Source: authors. Level 1: how institutions respond to threats (mechanisms) The mechanistic axis reveals institutional responses following a threat, leading to the observable outcomes defined by the degree of population harm reduction and equity promotion. When faced with a threat, institutions are motivated (or not) to respond. Motivation or willingness to respond can reflect underlying institutional ideologies, norms, and knowledge related to the threat. If willing to act, institutions commission, consider, or adapt different types of evidence to inform a response. Evidence acceptability is another important gap. The availability, dissemination, and acceptability of various forms of evidence—whether pertaining to science, experience, or context—depends on the nature of the threat and the underlying determinants of institutional motivation (ideology, norms, and knowledge). The willingness to act and the consideration of evidence yields a particular policy response which requires execution. Hence, the last gap concerns implementation. A policy may not yield its intended effects because of the implementing institution's capacity (resources and communication and mobilisation networks), legitimacy (degree of trust between citizens and the institution), and authority (ability to enforce and sustain change) (Call, 2011). Level 2: the degree to which populations are protected (observable endpoints) When actors within institutions encounter threats but cannot respond to protect safety/security owing to institutional‐level gaps in motivation, evidence uptake, and implementation, fragility ensues. However, fragility is not the only governance endpoint. In contrast to fragility, our model also includes two forms of resilience: bad resilience and multisystemic resilience. These forms of resilience are distinguished based on whether the response is functionally successful (meaning an institution acts, uses some evidence, and implements a response such that aggregate harms decline but socially disadvantaged groups remain unprotected) but leads to inequitable safety and security between socially advantaged and disadvantaged groups (bad resilience), or whether the threat response equitably promotes population‐level safety and security (multisystemic resilience). Thus, the governance endpoints are differentiated by the resulting level of population protection and equity. Fragility occurs when institutions fail to protect adequately populations from harm, in large part because the threat has not been addressed. Bad resilience occurs when the threat is addressed inequitably, yielding disparities between socially advantaged and disadvantaged groups. Multisystemic resilience occurs when institutions reduce population harms while also addressing inequities and promoting more equitable distributions of safety, security, and opportunity. Defining governance endpoints to understand fragility better Our framework draws attention to three possible endpoints when governance institutions are faced with a climate change threat: fragility; bad resilience; and multisystemic resilience. The difference between them is the degree to which governance institutions protect population safety and security in ways that also advance equity, defined as the ‘absence of unfair, avoidable or remedial differences among groups of people … defined socially, economically, demographically, or geographically’ (WHO, 2026). Beyond the concepts of fairness (just and impartial treatment of individuals) and equality (uniform distribution of resources, opportunities, or rights across groups), equity involves the recognition of historical, social, economic, and structural disadvantages that create differential needs and opportunities, requiring tailored approaches to achieve fairness (WHO, 2026). Our model posits that motivation, evidence, and implementation inform mechanisms through which governance institutions can equitably protect human safety and security, thereby distinguishing between the endpoints of fragility, bad resilience, and multisystemic resilience. When crises strike, governance institutions may allocate resources to reduce or eliminate harmful population impacts. We argue that absorbing population‐level harms (such as economic shocks, morbidity, and mortality) in aggregate (that is, on average) is a partial solution. Threat absorption, isolated from strategies that address pre‐existing inequities, is insufficient. It is equally important to consider how population harms and threats manifest differently in relation to known markers of inequity such as Indigeneity, gender identity, rurality, and physical geographies. When a threat response is focused on reducing aggregate population‐level harms without regarding a society's most vulnerable and marginalised strata, local inequities could be masked by population averages. Equity is a root driver of improved social cohesion, which has a strong impact on improving human development (Seyoum, 2020). Hence, national‐level threat absorption without inequity stratification can erode social cohesion and the ability to mount a resilient response to external crises. Box 1 provides an example of equity erosion within the context of climate change. Box 1. Bangladesh case study: a gender‐inequitable climate response In Bangladesh, disasters triggered by natural hazards (such as floods and cyclones) and chronic climactic shifts in temperature or water salinity can lead to elevated population morbidity and mortality due to the primary impacts (like infrastructure damage or drowning) or secondary impacts (like famine, infectious disease, or forced displacement) (Das et al., 2024). Ameliorating population‐level harms through the establishment of emergency shelters, early warning systems, or climate‐resistant infrastructure can mask inequalities if gendered inequities are not also considered. An overall mortality risk reduction does not mean that all communities benefit from a homogeneous risk reduction or share the same underlying vulnerabilities. For instance, owing to persistent gender inequality, reduced autonomy during forced displacement, and the burden of household and other forms of precarious labour, women and girls from densely populated coastal areas are disproportionately affected by the secondary impacts of climate disasters like floods and cyclones (Reggers, 2019). National‐ and sub‐national‐level risk reduction may not hold for women and girls in climate‐affected areas because their baseline vulnerability may be higher. Furthermore, gender‐blind climate change mitigation policies may sideline specific needs related to whether and how women/girls access information, menstrual hygiene management, economic protection for female‐headed households, and violence protection (Logie et al., 2024; Al‐Mamun et al., 2025). Thus, in the post‐climate disaster context, gender inequalities may persist and may become more deeply entrenched. This Bangladesh case study represents the governance endpoint of bad resilience, as climate threats are reduced at the aggregate level, while gender inequities persist and are magnified in the wake of an extreme weather event. Source: authors. The concept of resilience stands in contrast to fragility, which, as we have discussed, is marked by an unwillingness or inability on the part of governance institutions to promote citizen safety and security when faced with threats. Holistic and multisystemic resilience centres on the ability of individuals, families, communities, civil society organisations, private sector entities, and sub‐national and national governments to prepare for and adapt to threats (Ungar, 2021). Resilient responses are informed by the type of coping/adaptation mechanism employed (Bollettino et al., 2017). In response to crises, governance institutions—whether national or sub‐national governments, private organisations, social/public services, or civil society organisations—respond in ways that either withstand, adapt, or transform pressures imposed by threats (Bollettino et al., 2017). The governance institution can also withstand, adapt, or be transformed by virtue of responding to a threat (Bollettino et al., 2017). Drawing on multisystemic resilience, which describes resilience as an adaptive, transformative, and dynamic process of interacting systems (Ungar, 2021), a multisystemic‐resilient threat response occurs when governance institutions reduce population harms while also inducing transformational change that centres equity among marginalised groups. In crisis situations, the ability of governance institutions to leverage policies as tools for population health and equity is transformational and dynamic. Resilient responses involve coordination and mobilisation across multiple levels of local and executive power and promote institutional evolution that advances equitable population‐level safety and security in the face of threats. Since institutions can be resilient in ways that yield negative effects on population equity, we are deliberate in defining multisystemic resilience as the combination of threat absorption and equity promotion. Saulnier and Topp (2024, p. 2) coined the term ‘bad resilience’ to describe crisis ‘adaptation or transformation that maintains … system function but produces or perpetuates … inequitable outcomes in the process’ (see Box 2 for an example). Consequently, we are deliberate in defining resilience as threat amelioration and equity promotion, as not all forms of resilience yield favourable population outcomes among all social groups (Saulnier and Topp, 2024). Box 2. Bad resilience in protracted humanitarian settings One example of bad or inequitable resilience is international donor dependency on essential public services (such as nutrition, food security, agriculture, water and sanitation, and health) (Katz et al. 2024). In protracted humanitarian settings such as Haiti, South Sudan, or Yemen, public service financing and delivery are highly dependent on international donors and non‐governmental organisations (Bräutigam, 2000; Efe, 2018; Katz et al., 2024). While useful in maintaining lifesaving operations during crises like disasters triggered by natural hazards, bad resilience can occur when, over time, national health systems are bypassed, short‐term project lifecycles prevent long‐term planning and national public sector development, local capacity and human resources are underdeveloped, and donor exit plans are not established (Bräutigam, 2000; Efe, 2018; Katz et al., 2024). Subsequent crises reinforce donor dependency responses, which, in turn, create fragile governance systems that undermine sovereign responses to crises like climate change and related disasters (Bräutigam, 2000; Efe, 2018; Katz et al., 2024). This example illustrates bad resilience. While donors can establish lifesaving services, they are implemented while donor dependency is further maintained and reinforced. This cycle erodes national sovereignty and the ability of countries to identify their own priorities and finance independent infrastructure. Source: authors. The motivation, evidence, and implementation gaps are not themselves endpoints, nor do they map one‐to‐one on to a particular endpoint. Rather, they are interacting mechanisms that shape whether and how institutions respond to threats. Based on the underlying mechanisms, the resulting governance endpoints are classified along two dimensions: (i) the extent to which population harm is reduced; and (ii) the extent to which the response addresses inequitable distributions of vulnerability, voice, benefits, and burdens. Fragility occurs when institutions fail to reduce population harm adequately; bad resilience occurs when system function or aggregate protection is maintained but inequities persist or worsen; and multisystemic resilience occurs when harm reduction is accompanied by equitable and potentially transformative institutional change. Thus far we have defined governance fragility as occurring when governance institutions cannot or will not respond to protect population safety and security when threats are present. We have contrasted it with the competing endpoint of resilience, where governance institutions respond to a threat and are transformed either closer to or away from equity. We have highlighted the endpoint of bad resilience (where governance institutions respond to a threat in ways that produce inequitable population outcomes) and multisystemic resilience (where governance institutions achieve equitable population‐level safety and security in the face of threats). In the next subsection, we describe three institutional‐level gaps situated as intermediaries between the threats and three endpoints (fragility, bad resilience, and multisystemic resilience). The combination and size of institutional‐level gaps in motivation, evidence uptake, and implementation influence which governance endpoint is achieved when confronted with a threat. The motivation gap Threats like climate change have the potential to strain and destabilise lifesaving infrastructure, leading to population‐level harms, the magnification of existing disparities, and the creation of new inequalities (IPCC, 2022). The first determination is whether governance institutions believe they can and should act in the light of a threat. Governance structures differ in their knowledge, risk perception, and norms with respect to climate change. One important differentiator of motivation is whether the climate change threat represents a slow‐onset event that requires prospective disaster risk management/mitigation given uncertain population harms or an extreme weather event necessitating corrective disaster risk management owing to already present harms. The lack of a policy response in either case is a decision in and of itself, reflecting different ways in which threats are perceived by actors within institutions with ideologies, norms, and knowledge. Importantly, we argue that the underlying institutional ideology influences (i) the motivation of whether to respond to a variety of climate change threats (whether requiring corrective or prospective disaster risk management) and (ii) how likely institutional motivation is to change when presented with evidence. Ideology is a system's underlying and implicit belief in the nature of humanity and the role of institutions relative to this perceived humanity (Prinja, 2010). In our model, the willingness of individual policy actors to respond is moderated by the ideological undercurrent of the institutions they operate within and institutional norms and knowledge relative to the threat (see Box 3 for an example). The motivation gap also includes a critical examination of equity concerns, including whose exposure, safety, and welfare institutions recognise as requiring action. From an equity perspective, climate change threats experienced by socially disadvantaged groups may engender less attention as compared to impacts experienced by populations with political and social capital. Environmental racism illustrates equity issues relevant to the motivation gap (Bullard, 1994). The lack of political representation and accountability among equity‐seeking groups creates situations wherein Black, Indigenous, and people of colour communities are sites of pollution and waste facilities and exposed to environmental agents or exposures (pollutants, rising temperatures, and extreme weather events) that yield adverse health outcomes (Bullard, 1994). Another way that ideologies demotivate action is by discounting certain forms of evidence from altering threat perception and promoting opinion or dogma. Box 3. US case study: ideologies of climate disaster response in Puerto Rico The ideologies of national governance structures can perpetuate policy stagnation in the face of a climate disaster. Consider the case of the self‐governing Caribbean island of Puerto Rico. In 2017, this unincorporated US territory faced the dual crisis of sovereign debt default and Hurricane Maria. The US commitment to fiscal austerity and territorial governance delayed and weakened the federal US response to Maria (Espinosa, 2020; Willison et al., 2019). The Puerto Rico Oversight, Management, and Economic Stability Act of 2016 created a fiscal control board to enforce austerity and debt repayment (Lamba‐Nieves et al., 2021). The Act has been criticised for operating under a neoliberal ideology that prioritises debt servicing over public investment, even during humanitarian emergencies (Cortés, 2018; Lamba‐Nieves et al., 2021). Following Hurricane Maria, the US federal response was slow and under‐resourced, partly due to Puerto Rico's territorial status and lack of political representation in the US Congress (Espinosa, 2020; Willison et al., 2019). The crisis was framed as a budgetary rather than a humanitarian issue; more than 4,500 excess deaths were linked to the disaster and its aftermath (Kishore et al., 2018). The example of Hurricane Maria exemplifies that governance structures can choose not to act resiliently based on an underlying ideology that shapes the policymaking landscape. Source: authors. The evidence gap Policy responses can be shaped by diverse sources of evidence filtered through underlying opinion and ideology. Evidence‐informed policies in contrast are shaped by the best quality evidence that is available at a given time and contextualised/adapted for a given setting or threat (Head, 2016). Research (informed by the natural sciences, engineering, public health, and sociology, among other disciplines), experiences (inclusive of professional, policy, or clinical expertise and lived events), and context (including culture, history, norms, and physical geographies) inform the evidence base of potential policy strategies. However, not all climate change threats have an equally robust evidence base with respect to protecting population‐level harm and promoting equity, not all forms of evidence are perceived as being sufficiently legitimate or rigorous, and not everyone is equally recognised as an expert (Fricker, 2007). Within the evidence gap, equity issues concern whose knowledge is regarded as credible, which population outcomes are measured, and whether available evidence reflects the experiences and contexts of structurally disadvantaged groups. In some cases, institutional ideology influences who and what kinds of evidence are credible. The potential for resilient response improves when there is a strong evidence‐to‐action pipeline that integrates scientific research and policy expertise, lived experiences, and contextual considerations. Instead of engaging with evidence from a hierarchical positioning of credible truth, evidence is increasingly recognised as an ecosystem (WHO, 2021). There are varied forms of evidence: science and research (biomedical, social, political, economic, and impact evaluations); policy expertise; and community histories and lived experiences (Oliver, Lorenc, and Innvær, 2014; WHO, 2021). These forms of evidence may conflict with one another. For example, an impact evaluation of a given policy solution may indicate statistically significant effects; yet, in practice, collective and personal lived experiences may challenge them (Copeland et al., 2022). In these circumstances, adapting policies to contextual realities is key. Using evidence to make decisions about legislation, resource allocation, regulations, or strategies not only involves the creation, collection, and synthesis of evidence, but also the contextualisation of evidence. Evidence needs to be both generated and tailored to the specific implementation context, noting equity considerations like geography, culture, language, and social norms (Copeland et al., 2022). When governance structures are ideologically and motivationally inclined to seek out, evaluate, and contextualise the evidence available or fund the creation of new evidence from equity‐seeking groups, evidence‐informed policies can ensue (Parkhurst, 2017). The mere adoption of an evidence‐informed strategy is not sufficient, though, for achieving resilience: implementation matters too. Evidence‐informed policy adoption refers to the decision to commit or develop a policy based on evidence, whereas implementation is the process of putting the evidence‐informed policy into routine practice. The implementation gap The mere adoption of an evidence‐informed and contextualised strategy by institutions with adequate motivational action is not sufficient to achieve multisystemic resilience. Drawing on the political science literature, we argue that implementation is determined by the implementing institution's capacity, legitimacy and authority (Call, 2011). Institutions leverage their capacity, legitimacy, and authority to translate policy strategies into the routine practices, system transformations, or changed behaviours that can culminate in crisis amelioration and equity promotion (Bauer et al., 2015). However, an evidence‐informed policy response that is otherwise efficacious, or successful under ideal or strictly controlled conditions, can be ineffective, in that it fails to ameliorate harms and promote equity in real world circumstances (Bauer et al., 2015). The implementation gap, consisting of strategies that should work in theory but fail to work in practice, warrants additional examination (Ridde, 2016). We contend that structural implementation gaps (i) operate at a level above the behaviours of individual implementers and (ii) are determined by governance‐level capacity, legitimacy, and authority. Capacity is a relational process involving the potential to mobilise and deploy infrastructure, resources, and knowledge to implement a response (Ferreira, 2017; van Popering‐Verkerk et al., 2022). Capacity encompasses interaction, exchange, and coordination either vertically (international, national, sub‐national, municipal, and civil society) or horizontally (within departments or units at the same level of governance power). Given that state and non‐state institutions can collectively interact, capacity is an attribute of individual institutions and inter‐institutional networks with regard to resource sharing and mobilisation (van Popering‐Verkerk et al., 2022). Various resources inform capacity: natural, physical, and financial (van Popering‐Verkerk et al., 2022). The capacity to exchange vertically and horizontally involves relational information and resource sharing with respect to opportunities and threats (van Popering‐Verkerk et al., 2022). Depleted resources and the lack of horizontal or vertical coordination can lead to capacity weaknesses and implementation issues. Further implementation considerations related to fragility include the degree of trust between institutions and citizens and citizens' level of confidence that the institution can meet public needs (legitimacy) (Zoellick, 2008; Haider and McLoughlin, 2016), as well as the consolidation of power that gives institutions enforcement control via force, the law, or regulatory means (authority) (Zoellick, 2008). Even the most promising evidence‐informed policy strategies can fail to yield their intended outcomes if implementing institutions have weak legitimacy and authority. Legitimacy weakness can promote strong feelings of mistrust, resulting in the perception that institutions responsible for implementation or evidence creation are not acting in the best interests of the people they serve, reflecting deeper historical inequities (Zoellick, 2008; Haider and McLoughlin, 2016). Authority weakness erodes the ability of institutions to codify policy strategies into regulation or law, rendering some as normative rather than enforceable in law (Zoellick, 2008). Altogether, the implementation gap reminds us that even if institutions are otherwise ideologically inclined to act, fragility can manifest when they are unable to implement evidence‐informed policy strategies successfully. Within the implementation gap, equity considerations include whether institutional capacity, legitimacy, and authority are mobilised not only to ensure that protective policies are accessible, responsive, and beneficial within socially advantaged groups, but also to equalise outcomes between socially advantaged and disadvantaged groups. For instance, socially disadvantaged groups can have strong perceptions of mistrust in government institutions, which erodes their implementation capacity. Gaps in capacity, legitimacy, and authority inform the implementation gap, as illustrated in Box 4. Box 4. Case examples of different implementation gaps Capacity: climate disaster response requires vertical and horizontal capacity coordination After acute climate disasters that overwhelm local responses, national‐level emergency management agencies can coordinate with sub‐national governments and local organisations. Federal emergency response agencies are equipped with the humanitarian resources needed to ameliorate harm. Localised knowledge, networks, and organisations are needed to implement disaster response in historically marginalised areas. The vertical coordination of information and resources among federal and local organisations and horizontal coordination of service provision among local organisations is vital to implementing a resilient response (Comfort, 2005). This type of relational capacity gap was present in the 2005 Hurricane Katrina response (Comfort, 2005), which was regarded as a failure of governance as it resulted in about 1,000 deaths directly linked to the disaster (Brunkard, Namulanda, and Ratard, 2008) and 400,000 forcibly displaced residents (Geaghan, 2011). Legitimacy: water use behaviour change requires institutional trust The 2015–18 drought in Cape Town, South Africa led to strict water restrictions, wherein municipal water supplies were shut down (Ziervogel, 2019). Pre‐existing inequities within informal settlements and poor areas fuelled the perception that water restrictions and related messages were political and not due to climate change (Harris, 2020). Some communities believed the crisis was being used to target poorer areas unfairly (Enqvist and Ziervogel, 2019). Apartheid legacies resulted in wealthier and often white areas having more reliable water access and adaptation resources (Thompson and Tollefson, 2025). Poorer communities in informal settlements, which were underserved, faced the brunt of the water restrictions (Enqvist and Ziervogel, 2019). Authority: environmental protection laws need enforcement power Owing to agribusiness agendas, many national governments struggle with enforcing existing environmental laws (Igari, Tambosi, and Pivello, 2009). The illegal deforestation of the Brazilian Amazon being one example of weak legal enforcement of existing environmental laws (Igari et al., 2009). Brazil's comprehensive environmental laws and implementing organisations are undermined by budget cuts, staff reductions, and political interference (Barbosa, Alves, and Grelle, 2021). In some circumstances, inspectors and environmental defenders encounter threats or violence, thereby further deterring enforcement (Rodrigues, Campos, and Santana, 2022). Thus, the state lacks formal action, favouring policy compliance. Source: authors. POTENTIAL USES OF THE FRAGILITY MODEL The model presented in Figure 1 can be used as an analytical heuristic to identify governance fragility and distinguish between responses that are functionally successful but perpetuate inequities (bad resilience) versus responses that resolve threats equitably. Researchers and policymakers can employ the model retrospectively to assess to what degree and among which populations institutions acted in response to a climate change threat, how institutions selected, interpreted, and contextualised evidence (including what evidence was deemed credible), and whether policy responses were implemented with adequate capacity, legitimacy, and authority. These mechanistic considerations can be used to explain observable population patterns with respect to threat amelioration and equity promotion. For example, governance endpoints can be analysed according to whether policy responses reduced population harm and whether benefits, burdens, and protection were distributed equitably across groups. The model can also be used to identify proactively mechanistic gaps that can lead to fragility or bad resilience, thereby serving as an institutional strengthening model. As noted earlier, however, the motivation, evidence, and implementation gaps are not themselves endpoints and nor do they map one‐to‐one on to particular governance endpoints. Rather, they inform a complex and nuanced understanding of the mechanisms that shape whether and how instructions respond to threats that culminate in fragility, bad resilience, and multisystemic resilience endpoints. While the model can inform testable hypotheses and potential causal pathways, it is not meant to be deterministic. Instead, we intend the model to be a heuristic tool to understand better the nature of governance fragility in relation to climate change threats and in HICs. HIGH‐INCOME COUNTRIES Our framework positions ‘fragility’ as a process wherein governance institutions are unwilling or unable to ameliorate threats and promote safety and security due to institutional gaps in motivation (threat perception), evidence uptake (design, adopt, or contextualise evidence‐informed solutions), and implementation (policy response integration informed by institutional capacity, legitimacy, and authority). The size and combination of these institutional gaps also inform two additional governance endpoints, which are distinguished from fragility: (i) bad resilience (where policy responses reproduce or perpetuate inequitable population outcomes); and (ii) multisystemic resilience (where policy responses transform institutional motivation, evidence acceptability, and implementation to achieve equitable population gains). We recognise that governance institutions in many countries struggle with responding to climate threats in a manner that achieves multisystemic resilience, where the threat is absorbed and equity is promoted. As stated by other scholars, fragility is better understood as the norm, rather than as an aberration within the international landscape (Saeed, 2023). In applying this fragility framework, researchers, policymakers, and practitioners can better diagnose governance endpoints and link them to specific threats and institutional‐level gaps. Another benefit is the identification of hidden institutional fragilities in HICs because they are not aligned with the image of a conflict‐affected LMIC (Fund for Peace, 2023; OECD, 2025; World Bank Group, 2026a). We argue that the nature of fragility in HICs has been under‐conceptualised and thus suggest that institutional gaps in motivation, evidence uptake, and implementation are potentially masked by a country's overall development status and economic strength (see Table 1). Institutional gaps dysregulate the mechanism from threat to multisystemic resilience, rendering fragility and bad resilience as suboptimal endpoints. TABLE 1Denoting the three types of fragility gaps.Type of fragility gapDescriptionClimate change applicationMotivation gapThe ideological undercurrent of governance institutions influences whether and to what degree policy action addresses a threat, based on institutional norms and forms of knowledge perceived as legitimate or true. Authoritarianism and nationalism do not position anthropogenic climate change as a priority of global policy. Therefore, appropriate responses are not taken to prevent and respond to slow climactic shifts as these threats are not seen as legitimate. Evidence gapWhether evidence (scientific, experiential, contextual, or a combination thereof) is available, applicable, contextualised, and disseminated to inform and guide policy direction and implementation. Disinvestment in climate science slows progress in green technologies and equitable responses to the climate crisis.The experiences of grassroots community organisations and persons most affected by climate change are ignored in national‐level policymaking. Implementation gapTo what degree governance institutions are able to leverage their capacity, legitimacy, and/or authority for the purpose of sustainably implementing an evidence‐informed solution that ameliorates a crisis and promotes equity. National, sub‐national, and community‐level organisations do not have communication networks to disseminate and tailor early climate change warning system messages.Governance structures like federal governments, humanitarian response entities, and environmental protection agencies are not perceived as trustworthy information sources when climate disasters occur.Normative climate protection frameworks are not enshrined in enforceable law, thereby eroding punitive measures for climate‐disrupting industries. Source: authors. Denoting the three types of fragility gaps. Authoritarianism and nationalism do not position anthropogenic climate change as a priority of global policy. Therefore, appropriate responses are not taken to prevent and respond to slow climactic shifts as these threats are not seen as legitimate. Authoritarianism and nationalism do not position anthropogenic climate change as a priority of global policy. Therefore, appropriate responses are not taken to prevent and respond to slow climactic shifts as these threats are not seen as legitimate. Disinvestment in climate science slows progress in green technologies and equitable responses to the climate crisis. Disinvestment in climate science slows progress in green technologies and equitable responses to the climate crisis. The experiences of grassroots community organisations and persons most affected by climate change are ignored in national‐level policymaking. The experiences of grassroots community organisations and persons most affected by climate change are ignored in national‐level policymaking. National, sub‐national, and community‐level organisations do not have communication networks to disseminate and tailor early climate change warning system messages. National, sub‐national, and community‐level organisations do not have communication networks to disseminate and tailor early climate change warning system messages. Governance structures like federal governments, humanitarian response entities, and environmental protection agencies are not perceived as trustworthy information sources when climate disasters occur. Governance structures like federal governments, humanitarian response entities, and environmental protection agencies are not perceived as trustworthy information sources when climate disasters occur. Normative climate protection frameworks are not enshrined in enforceable law, thereby eroding punitive measures for climate‐disrupting industries. Normative climate protection frameworks are not enshrined in enforceable law, thereby eroding punitive measures for climate‐disrupting industries. Source: authors. These forms of fragility are particularly relevant for transborder threats like climate change. The typology of fragility has applications in research and practice at the interface of governance, climate change, implementation science, and public health. For example, we can consider how fragility manifests as motivational inaction in countries where institutions otherwise have sufficient resources to act (Haken and Fiertz, 2023). However, there is currently a dearth of research and best practices on how to engage across the ideological divides that inform motivation, knowledge, risk perception, and institutional norms (Haken and Fiertz, 2023). Additionally, while there may be evidence available to design policy strategies that ameliorate threats, its acceptability in equity‐seeking groups could be lacking, rendering this a second action area. Evidence acceptability is intimately related to who is seen as a credible source of information and can follow ideological divides. In other areas, more evidence generation and contextualisation are needed to integrate science and experience and context. A third action area concerns pinpointing settings where institutions are motivated to act but cannot achieve a multisystem‐resilient endpoint owing to implementation barriers informed by institutional capacity, legitimacy, and authority. A case study of climate change fragility in HICs Some scholars argue that ‘there is no country that will be safe from climate change’ (Sikorsky, Femia, and Hugh, 2023, p. 208). Government institutions were not designed to withstand the new climate patterns that human activity has created (Carment and Samy, 2023), in part because the borders that mark sovereign nations do not halt climate change and the time horizon of climactic shifts transcends the length of most political terms. The combination of slow climate hazards and rapid shocks leads to interconnected cycles that necessitate a mixture of mitigation and adaptation policy responses that have differently‐sized gaps in motivation, evidence uptake, and implementation (IPCC, 2022). As climate change continues to endanger human life via changes in agriculture, water and air quality, and exposure to hazards, the executive powers of national governments have increasingly endorsed positions that are authoritative and nationalist, reflecting a motivational‐level reality reflected in the proposed fragility framework (Freedom House, 2022). As authoritarian practices grow, political and civil freedoms decline, and fragility endpoints become more common (Freedom House, 2022; OECD, 2025). In this subsection, we apply our fragility framework to the case study of climate change fragility in the US. The US provides a stark example of how the motivation gap (informed by ideology) operates within a developed country that has been classified as an economic powerhouse (Haken and Fiertz, 2023). As other scholars have explained, ‘symptoms of structural weakness can be hidden by a country's wealth or power, while a measurable long‐term worsening in key cohesion indicators and a rise in retrenchment populism and scapegoating create a situation of potential precarity in the event of the next big shock’ (Haken and Fiertz, 2023, pp. 73–74). On the one hand, a country like the US has great potential for climate action in the form of adaptation policies, technological investment, and establishing enforceable international standards (Nnabuife et al., 2024). The US has the world's largest gross domestic product (World Bank Group, 2026b) while also housing some of the world's most prolific technological companies (Stanford University HAI, 2024). However, right‐wing populism, reflected in rising authoritarian and nationalist ideologies, position climate action as an unworthy policy target (affecting motivation), impacts the types of climate change evidence available for policy influence (evidence uptake), and erodes the capacity, legitimacy, and authority of implementing institutions (Fiorino, 2022). Motivation gaps One foundational determinant of climate action is the belief or scepticism that anthropogenic climate change is a real phenomenon that should be acted on by institutions, broadly speaking, to ameliorate population harms (Hornsey et al., 2016). We argue that rising authoritarian and nationalistic ideologies are re‐emerging in the US as root causes of fragility endpoints because they have the power to orient domestic and international institutions in ways that widen motivation, evidence acceptability, and implementation gaps. Authoritarianism and nationalism can engender fragility, regardless of a country's development status, and often begin at the motivation gap. Two key points are notable here. First, authoritarian ideology is fuelled by the consolidation of executive power and the minimisation or elimination of institutions that act as checks and balances on the executive branch (Ben‐Ghiat, 2020; Freedom House, 2022). The ideology of the autocrat determines which crises are acted on and which are ignored (Ben‐Ghiat, 2020; Freedom House, 2022). Second, persons who endorse authoritarian ideologies are less interested in population equity outcomes, nationally or internationally (Ben‐Ghiat, 2020). Nationalistic ideologies are isolationist, associated with an active dismantling of foreign‐facing institutions concerned with equitable outcomes and a reduction in the capacity of national and international institutions to coordinate knowledge and resources (Freedom House, 2022; Prather, 2024). Evidence uptake Under current direction, the US Department of Health and Human Services does not perceive climate and equity‐related topics as aligned with executive branch priorities (Waldman and Lerner, 2025). The executive branch of the US government cancelled federal climate change funding, valued at USD 11 billion, in 2024 (Turnley, 2025). And the executive directive to remove National Institutes of Health funding of grants that address the health effects of climate change, in addition to other equity‐related topics, sharpens the ideological direction of the US administration (Waldman and Lerner, 2025). These funding cancellations are occurring within a wider environment wherein the US Supreme Court sustained the latter grant terminations, valued at USD 2 billion, and related to diversity, equity, and inclusion (Garisto and Kozlov, 2025; Kozlov and Ryan, 2025). With less research funding comes a reduced ability to develop a toolkit of evidence‐informed policy prevention and response strategies that can be deployed. Furthermore, social trust in scientific research conducted in university settings has been declining for the past decade (Kennedy and Tyson, 2023). A 2024 report on institutional trust reported that 67 per cent of people in the US agreed that ‘[s]cience has become politicized in this country’ (higher than the global average of 53 per cent) (Edelman Trust Institute, 2024, p. 17). The COVID‐19 (coronavirus disease 2019) pandemic was a prime example of disinformation being used to downgrade scientific evidence and expert advice; from 2017 to 2021, the executive branch of the US government promoted unproven COVID‐19 treatments and denied the risks of climate change (Turnley, 2025). The lack of trust in scientific research conducted in universities is linked to climate denialism (Alvarez, Debnath, and Ebanks, 2023). In the US, increased distrust of scientists is associated with endorsing climate misinformation (Ejaz et al., 2025). Public distrust in scientific research weakens the science‐to‐society translation and the ability of governments to implement evidence‐informed climate policies (Alvarez, Debnath, and Ebanks, 2023). Other developments include the administration's efforts to repeal research from 2009 that positioned greenhouse gasses as dangerous for public health: landmark scientific evidence underscoring the basis of the US Environmental Protection Agency with regard to regulating greenhouse gases (Thompson and Tollefson, 2025; Tollefson, 2025). A draft report published in July 2025 questioned the evidence underlying global warming, paving the way to replacing or discrediting the original scientific basis of greenhouse gas regulation (Christy et al., 2025). Implementation gaps Following the Second World War (1939–45), the US took the lead in the formation of several multilateral institutions that shaped a democratic world in favour of American security and interests: the North Atlantic Treaty Organization, the World Bank, the International Monetary Fund, and the United Nations and affiliated agencies like the World Health Organization (Rathbun, 2011). Some 80 years later, catalysed by the economic fallout of the COVID‐19 pandemic, the US cancelled its participation in some of the global institutions and instruments that it played a key role in creating (Yamey and Titanji, 2025), notably: the World Health Organization (Yamey and Titanji, 2025), the Paris Agreement (Kenny, 2024), the United Nations Framework Convention on Climate Change (United Nations, 1992), and the Sustainable Development Goals (Tonello, 2025). Rejecting the normative standards set by the Paris Agreement and the Sustainable Development Goals weakens the ability of national government institutions to formalise global targets pertaining to climate change prevention and response by adopting enforceable regulatory amendments or laws (Swain et al., 2025). In other words, normative standards never become enshrined in enforceable laws that shape behaviour through positive reinforcement or punitive measures. The most profound losses are expected to be in the areas of climate change mitigation/prevention, where there is a paucity of existing national protection legislation (Fiorino, 2025). The erosion of vertical capacity mobilisation is illustrated by the ceasing of all communications between the US Centers for Disease Control and Prevention and the World Health Organization in 2025 (Yamey and Titanji, 2025). This communication rupture threatens resource exchange in the form of data and expertise, which is relevant to the early identification of transborder threats such as an emerging infectious disease caused by shifting climate patterns (Yamey and Titanji, 2025). Other foreign policy strategies that engender implementation barriers by eroding capacity include (i) subsuming the United States Agency for International Development (a previously independent agency of the government responsible for foreign aid and development) under the Department of State and reducing its labour force (Burki, 2025) and (ii) the dismantling of the United States Institute of Peace (a nonprofit organisation founded by Congress in 1984 to bolster diplomatic solutions to conflict, including those related to climate change and disasters) (Strand, 2025). As the US case study illustrates, national‐level governance gaps in motivation, evidence uptake, and implementation circumscribe the willingness and ability of individual policy actors working within governance institutions to respond and implement climate action policies. Call to action and conclusions Viewed through the lens of our conceptual model (see Figure 1), the US case illustrates mechanisms towards multiple governance endpoints. National‐level failures to recognise climate threats (failing to act), consider evidence for policy solutions (failing to uptake evidence), and implement solutions (failure to implement) can culminate in fragility, reflected in population harms that are not prevented or absorbed. Movement towards multisystemic resilience requires that institutions narrow all three gaps by recognising climate threats as legitimate grounds for action, incorporating scientific and community knowledge into locally appropriate policy, and mobilising sufficient capacity, legitimacy, and authority to deliver equitable protection. The endpoint therefore depends not merely on whether a response occurs, but also on whom it protects, whose evidence informs it, and whether it transforms the conditions producing unequal vulnerability. In our model, a multisystemic‐resilient climate response is measured by the endpoint of equitable population harm reduction and is achieved mechanistically when actors respond to threats within institutions in ways that minimise gaps in motivation, evidence acceptability, and implementation. Our contribution to conceptualising fragility and resilience in response to threats engages with a recent report published by the Center on International Cooperation, which notes that ‘resilience has undergone a quiet revolution over the last decade’ yet the ‘nascent “fragility to resilience” paradigm … lacks globally recognized, shared, and strategic frameworks and approaches to guide national reformers and international partners to build resilience in an age of crises’ (Papoulidis et al., 2024, pp. 5–6). Although our present work offers a novel framework that contrasts fragility and resilience as governance endpoints influenced by institutional‐level gaps, more work is needed to operationalise the path towards multisystemic resilience. Authoritarian and nationalist political ideologies are rising globally and create (by design) institutional‐level gaps in motivation, evidence uptake, and implementation that erode policy action for climate responses like decarbonisation and adaptation. Accordingly, an integrated research‐to‐practice agenda is needed at the nexus of climate change, implementation, governance, and public health. This can promote both theoretical/conceptual debates that spark novel research questions (such as the ones outlined in Box 5) and, from an applied perspective, the design, monitoring, and evaluation of resilient and localised governance systems. Box 5. Research questions informed by our globally applicable model of fragility In what ways do HICs descend into fragility and what are the downstream consequences for climate action? In what ways do HICs descend into fragility and what are the downstream consequences for climate action? How can climate and development actors minimise climate policy action barriers related to (i) motivations and ideologies, (ii) evidence acceptability, and (iii) implementation at the institutional level? How can climate and development actors minimise climate policy action barriers related to (i) motivations and ideologies, (ii) evidence acceptability, and (iii) implementation at the institutional level? What evidence‐informed strategies exist to disrupt the spread of ideologies that contribute to decreased evidence acceptability and implementation gaps related to climate change action? What evidence‐informed strategies exist to disrupt the spread of ideologies that contribute to decreased evidence acceptability and implementation gaps related to climate change action? How do ideologies that promote climate change resilience spread within and between governance structures at the national and sub‐national levels? How do ideologies that promote climate change resilience spread within and between governance structures at the national and sub‐national levels? Source: authors. To support the achievement of multisystemic resilience, an integrated agenda can experiment with models of decentralisation in situations of national‐level climate response fragility. In a context where authoritarianism and nationalism create national‐level governance fragilities, it is necessary to investigate the roles and scope of sub‐national governments, the private sector, civil society, households, and individuals. Decentralisation refers to the transfer of responsibility and authority for government services from the national government to regional and/or local governments (Smoke and Cook, 2022; Quader et al., 2023). Sub‐national and local governance can be sites of innovative devolution models involving vertical and horizontal collaboration with private and non‐governmental actors (Smoke and Cook, 2022). Devolution results in elected sub‐national governments having greater accountability and responsibility, as compared to their national counterparts. Devolution models can be especially useful in contexts where local communities are at high risk of disaster‐induced harm, there is a high degree of localised will to act, evidence needs to be locally applicable/generated, and there is national‐level climate response governance fragility (Smoke and Cook, 2022). Sub‐national and local governance institutions are situated closer to climate change‐affected communities. In addition, local governments control many climate‐related sectors, some of which are sources of greenhouse gases (public transportation, land planning, waste disposal, building energy management and regulations, and water and power utilities), as well as contingent response plans for climate disaster emergencies (Martinez‐Vazquez, 2021). In these decentralised contexts, policy stakeholders may be more willing to contend with the human and economic toll of climate change and associated disasters, in contrast to their national‐level counterparts (reflecting a minimised motivation gap). Beyond a matter of ideological will, in comparison to national governments, sub‐national and local areas may be better positioned to develop policy solutions that are tailored to contextual realities, contextualise existing evidence, and commission locally relevant research like needs assessments (thereby minimising the evidence uptake gap) (Kuriakose et al., 2022). Climate adaptation has typically been a localised activity of incremental gains implemented by individuals/households with low levels of national government representation, reflecting a promise of localised implementation but a potential barrier to scale (Petzold et al., 2023). Sub‐national governments typically play a large role in planning and civil society groups have been generally responsible for coordination (Petzold et al., 2023). Since financing is customarily a core responsibility of national governments, national climate response fragility threatens the viability of fiscal decentralisation (Petzold et al., 2023). In HICs, there is potential for the sub‐national financing of climate adaption through tax revenue (Martinez‐Vazquez, 2021). In LMICs, multilateral governance institutions and the private sector can be key actors in reimaging the financing of climate action decentralisation (Martinez‐Vazquez, 2021). Many sub‐national governance structures have taken autonomous climate action in the absence of initiatives by national governments (Martinez‐Vazquez, 2021). For example, drawing on a United Kingdom case study, researchers developed processes of downscaling Paris Agreement decarbonisation commitments at the sub‐national level (Kuriakose, 2022). In Bangladesh, disaster management involves local disaster management committees, composed of community members, at the sub‐national and local levels (Quader et al., 2023). According to a report by the Data‐Driven EnviroLab and the NewClimate Institute (2020, p. 4), 823 cities and 101 regions have net‐zero targets; these local governments represent more than 846 million people. Building on this existing evidence, an integrated research agenda can design, test, and evaluate governance systems that advance decarbonisation and decentralisation in situations of national climate change response fragility. In conclusion, this article has considered the matter of what makes countries fragile. Two interlinked threats are relevant in HICs like the US: (i) climate change; and (ii) authoritarian ideologies/motivations. In the face of climate change, governance institutions can act resiliently when they are ideologically willing to consider evidence‐informed/adapted policies and leverage their capacity, legitimacy, and authority to implement responses that absorb threats and promote equity. Fragility occurs when there is a breakdown in the mechanism involving motivation, evidence uptake, and implementation, thereby leading to the magnification of population harm and inequities by failing to address the threat. This conceptualisation draws attention to a form of fragility that has received less research attention: countries with sufficient resources and infrastructure that fail to absorb crises and promote equity due to gaps in motivation, evidence acceptability, and implementation. Authoritarian ideologies have the potential to create and widen these gaps, resulting in climate change fragility (as noted in the US case study on national‐level climate change governance fragility). We advocate for a research‐to‐action agenda at the nexus of climate change, implementation, governance, and public health. This could advance theoretical/conceptual arguments concerning the impact of ideology on fragility and applied research focusing on decentralised forms of climate governance that leverage and build sub‐national resilience. DATA AVAILABILITY STATEMENT Data sharing not applicable to this article as no datasets were generated or analysed during the current study. REFERENCES Associated Data Data Availability Statement Data sharing not applicable to this article as no datasets were generated or analysed during the current study.