Human dimensions of climate change adaptation: Gaps and knowledge frontiers

Abstract

Adaptation science has grown exponentially in the past two decades and has progressed conceptually. However, these advances remain siloed from adaptation practice. In this commentary, I make three arguments: (1) adaptation theory has grown but is often unfit-for-practical-purpose; (2) disciplines and methodologies to study adaptation remain conservative, and poorly understand adaptation limits and residual risks; and (3) behaviour change as a critical lever of individual and societal adaptation remains understudied. By charting the landscape of knowledge gaps, I suggest frontiers for adaptation researchers to focus on.

Introduction: Is adaptation science up to the challenge?

In a 2012 editorial in Global Environmental Change, titled ‘Cool heads for a hot world – Social sciences under a changing sky’, Arun Agrawal, Maria Carmen Lemos, Ben Orlove, and Jesse Ribot delineated three spaces for the social sciences to contribute to climate change research and action (Agrawal et al., 2012). They argued that the social sciences must theorise (vulnerability, adaptation, adaptive capacity, and resilience); problematise (the causal structures of vulnerability and examine the forms, causes, and effects of adaptation); and understand and inform (adaptation policy and actors).

More than a decade later, we are at a critical juncture in adaptation science. ¹ The Paris Agreement (2015), a momentous juncture in global climate governance, finally recognised that adapting to escalating climate change impacts and risks is inevitable, and the capacities to adapt are unequally spread across countries and societies. The Paris Agreement also titillated and confused with the Global Goal on Adaptation (GGA), which aimed at ‘enhancing adaptive capacity, strengthening resilience and reducing vulnerability to climate change, with a view to contributing to sustainable development and ensuring an adequate response in the context of the temperature goal’ (United Nations, 2015, Article 7.1).

In practice, the GGA has been exasperatingly difficult to define, partly because the GGA was created to be in symmetry with the target on mitigation, ² and partly because of the political tensions that animate adaptation negotiations. In its wake, multiple metrics to assess adaptation progress and effectiveness have emerged, from the normative to the operational (Adaptation Committee, 2021; Craft and Fisher, 2018; Fisher, 2024; Magnan et al., 2023; Owen, 2020; Singh et al., 2021).

As one looks ahead to the crystallisation of the GGA and looks back to the decades of adaptation scholarship, it seems timely to revisit the promise of the climate change adaptation enterprise. Did the adaptation research community take on any of the provocations it laid out? What types of questions have the adaptation community answered? Whose adaptation got picked up and whose were silenced? In this commentary, I draw on decades of adaptation scholarship to argue three things. First, adaptation theory has grown but also ossified and is somewhat unfit-for-present-purpose. Second, methodologies (and disciplines) to study adaptation remain conservative, we have begun diagnosing adaptation failures but poorly understand the limits to adaptation and subsequent residual risks. Third, adaptation science leans too heavily on certain disciplines and certain scholars, which starves the field of innovation and critical levers (I illustrate this through a focus on behaviour change). I make these arguments in conversation with quotes from the ‘Cool heads for a hot world’ editorial, laying out the current landscape of adaptation research, and identifying knowledge gaps.

Adaptation science has a big data, disparate theory problem

we need to build databases, produce case studies, design robust quantitative and qualitative analytical approaches, and compare across the rich library of studies available in the literature focusing on local adaptation so as to build theoretical generalizations that are useful across geographies, cultures, and political systems and also relevant to more specific studies interested in individual contexts. (Agrawal et al., 2012: 329)

Adaptation research has grown substantially (Berrang-Ford et al., 2021; Nalau and Verrall, 2021; Vincent and Cundill, 2021), with scientific articles more than doubling from 2008 to 2011 (Bassett and Fogelman, 2013) and growing by 150% from 2011 to 2014 (Webber, 2016). Wang et al. (2018) estimated that since 2006, the publication rate on climate change adaptation has been 29% per annum. This ‘big literature’ has meant that only 12 years since the quote above, the knowledge base on adaptation has grown substantially, with a dominance of numerous place-based, in-depth, qualitative, single- or small-n adaptation case studies (Berrang-Ford et al., 2021; Biesbroek et al., 2018). While this growth has helped develop an ‘in-depth understanding on adaptation processes and decision making in specific locations’ (Biesbroek et al., 2018: 1), synthesising this evidence to identify what works, under what conditions, and for whom, has been a recurring challenge (Berrang-Ford et al., 2011).

Collectively, these case studies build a formidable database of ‘specific studies interested in individual contexts’, as Agarwal et al. (2012) identified the need for, but the adaptation research community has somewhat failed ³ to make sense of them to ‘build theoretical generalizations that are useful across geographies, cultures, and political systems’. For example, adaptation theory (and practice) continue to debate the links and boundaries of climate change adaptation and development (as an indication, follow Smit et al., 2000; to McGray et al., 2007; Eakin et al., 2014 onto; Schipper et al., 2021). Ben Orlove, in his commanding review on adaptation, rues ‘the difficulty of providing a succinct definition of adaptation that would support measurement of adaptation and assessment of effectiveness; instead, definitions tend to be taxonomic, providing lists of distinct types of adaptation that are not directly comparable” (Orlove, 2022: 572). This definitional haziness undermines adaptation practice (Leiter et al., 2019) and has often led to unfit-for-purpose adaptation projects, inadequate adaptation implementation, and ineffective adaptation funding. At worst, this has increased vulnerability in certain settings or increased the risk of maladaptation ⁴ (Atteridge and Remling, 2018; Reckien et al., 2023).

More recently, the Intergovernmental Panel on Climate Change's Sixth Assessment Report, (IPCC AR 6) assessed the literature on Climate Resilient Development, which identifies pathways and solutions that meet adaptation, mitigation, and sustainable development goals together (Schipper et al., 2022). As the need for adaptation becomes more pressing, the arenas in which adaptation theory and practice engage must be articulated clearly (Rickards et al., 2023). Current questions the adaptation community are dealing with include how we should delineate adaptation as beyond business-as-usual development? Do we mainstream adaptation at the cost of diluting its focus on risk reduction or look for somewhat illusionary ‘triple wins’ and ‘co-benefits’? Do we prioritise adaptive development (Agrawal and Lemos, 2015) or sustainable adaptation (Eriksen and Brown, 2011)? Does climate-compatible development adequately capture the multiple trade-offs and decisions individuals, households, communities, and countries face as they try to cope and adapt (Ficklin et al., 2018), or do Climate Resilient Development pathways better capture the inequities and path dependencies we must negotiate? For now, the answers to these questions differ widely between practitioners and researchers, as well as within these groups, points to the continuing gap between adaptation theory and practice.

Building on Hailie Eakin et al.'s framing of differentiating between generic and specific capacities to adapt (Eakin et al., 2014), I have argued, with colleagues, that in places where development delivery remains a gap, building a bedrock of ‘good’ development is essential to nurture adaptive capacity but insufficient to deliver adequate adaptation to climatic risks (Singh et al., 2018). Thus, sustainable development is a necessary but insufficient condition for effective adaptation to escalating climate risks.

But based on disciplinary training and geographical location, we are all partial to particular frames and indicators (Singh et al., 2021), and in the multiplicity of case studies and ‘recommendations’, adaptation implementers, funders, and researchers find themselves at sea. Theoretical plurality is a sign of debate, of a growing field but it can also paralyse and confuse; and ‘the apparent lack of critical reflection upon the robustness of these (adaptation) heuristics for diverse contexts may contribute to potential cognitive bias with respect to the framing of adaptation by both researchers and practitioners’ (Preston et al., 2015: 467).

Adaptation research has a small(er) methodology problem

We need also to make better use of the broad arsenal of social theory and methodological approaches. Indeed, adequately addressing the social complexities of vulnerability and adaptation associated with differences in scales, regions, and sectors requires different kinds of knowledge from and across the social science disciplines, recourse to different analytical frameworks, and even borrowings from the biological and physical sciences. (Agrawal et al., 2012: 330)

The methods used in vulnerability and adaptation research draw on a range of disciplines from ecology to environmental science, geography and development studies, political ecology, history and anthropology, and more recently, media and gender studies. This is a relatively recent change and we only have to look to Mike Hulme's commentary, ‘Meet the Humanities’ to remember how the climate change literature (and hence the IPCC) was ‘heavily dominated by natural science disciplines, especially the Earth sciences, while the minority social science content was heavily dominated by economics. Literature from the humanities was virtually absent’ (Hulme, 2011: 177). This disciplinary conservatism is most pressingly visible in the vast chasms between those studying physical aspects of climate change (atmospheric scientists, hydrologists, meteorologists, etc.), and social scientists (geographers, economists, psychologists, etc.). Adaptation research is also poorly informed by history, with inadequate learning from historical case studies to understand the present moment (Adamson et al., 2018) and insufficient engagement with philosophy/ethics to deepen work on climate justice (beyond the environmental justice framing that has an outsized footprint). These disciplinary gaps, I argue, lead to truncated methodological tools to diagnose and incentivise adaptation.

Current adaptation research is seeing innovative combinations of methodological approaches and conceptual leanings, pluralising analytical frameworks by forefronting issues that sit at inevitable intersections: intersectionality, which discusses how axes of marginalisation such as gender, ethnicity, or caste mediate vulnerability and adaptive capacity (Sultana, 2014; Few et al., 2021), temporality, which spotlights the importance of developing heuristics and methodologies that capture the dynamic nature of climate risks, vulnerability and adaptation (Singh et al., 2019), and translocality, which argues against sedentary biases, calling for a recognition of mobility – of people, resources, ideas – across space (Porst and Sakdapolrak, 2018), to mention a few. Interdisciplinary teams have built global databases (such as the Global Adaptation Mapping Initiative ⁵ ) and transdisciplinary groups have diversified what is measured (e.g. the Intergovernmental Panel on Ecosystem Services and Biodiversity 2023 Assessment Report).

Methodologically, much more is needed. Current renderings of key risks to human systems use ‘burning embers’ to denote higher levels of risk to key systems/sectors, as global warming increases (Zommers et al., 2020). Dependent on expert elicitation, they remain a blunt instrument to inform risk management and inadequately deal with emerging research on complex and systemic risk (Simpson et al., 2021). While the IPCC AR6 innovated by demonstrating how risk changes with and without adaptation, it was unable to develop similar assessments for adaptation limits, a critical piece of the puzzle for decision makers. The adaptation community would do well to borrow from health research to quantify if and how adaptation is enabling people to live (morbidity and mortality), produce (livelihoods and decent work), and access stocks and assets; and predict where limits may emerge. ⁶ Of course, in practice, deploying adaptation-relevant dimensions is not easy; for example, the ‘right’, quantifiable metrics/proxy indicators for ‘safer settlements’ or ‘resilient livelihoods’ can be found but they are inherently reductionist and incomplete. Crucially, without adequately mapping the contours of ‘the existential risk space of climate change’ (Huggel et al., 2022) how can we attempt to chart adaptation options and pathways?

This brings me to the second methodological lacuna in adaptation research. Despite the fact that adaptation actions occur over a continuum (from short-term coping to longer-term adaptation to transformational change), we lack theory-informed, usable tools to assess adaptation outcomes over time (Orlove, 2022). The methodological approaches we have either statistically review projects/adaptation options ex post (Reckien et al., 2023; UNEP, 2022) or build forward-looking adaptation pathways and scenarios (Haasnoot et al., 2024; Rickards et al., 2014). Currently, these approaches assess adaptation outcomes incompletely; whether for specific groups or goals (e.g. risk reduced of the most vulnerable) or on temporal aspects of adaptation (i.e. long-term impacts of short-term actions, lock-ins, and trade-offs between actions). Further, such methods rarely draw on historical case studies, which can offer temporal analogues (Ford et al., 2010: 382) and improve ‘our understanding of how human systems experience (i.e. exposure and sensitivity) and respond to stress (i.e. adaptive capacity), not whether past climate stresses are identical to those predicated by climate models’. Instead of lamenting their incompleteness, funders, and implementers need to choose combinations of fit-for-purpose approaches that are retrospective and forward-looking, track long-term and near-term outcomes, and must ascertain what level of ‘incompleteness’ (of approaches) is acceptable without having negative impacts (e.g. missing tracking maladaptive outcomes and only focussing on immediate risks).

And finally, even if we diversify our methods and disciplines, how will this improve the adaptation enterprise? Will more knowledge truly lead to better practice? What is the endeavour of multi/inter/transdisciplinary adaptation research? Ideally, interdisciplinary research should enrich different disciplines it is drawing from, with researchers acting as nodes that take back interdisciplinary threads and conversations to their disciplinary homes. This is happening, albeit slowly. Jon Barnett (2022: 1112), reviewing the political economy literature notes that more is needed: ‘A political economy of transformation would entail a far more profound (re)assembling of environments, technologies, and practices over space and time to sustain human needs and values through a changing climate, enabled by a societal shift in thought about human–nature relations, stronger networks of actors across diverse places united by empathy and solidarity, and renewed social contracts that reaffirm and demand that the state facilitates collective action for the protection of public goods’. This reassembling is critical and currently insufficient in the adaptation research landscape.

Transdisciplinary work, that is, an adaptation that traverses research and practice, is growing but also low. The endeavour of transdisciplinary work is to take learnings back to the transdisciplinary space, that is, to different communities of practice, whether adaptation researchers, practitioners, funders, knowledge brokers, etc. But this is not happening, partially because truly transdisciplinary teams are rare, such work is inherently difficult and time-intensive, and there may not be enough people to take such transdisciplinary lessons back to.

Levers of adaptation action: societal shifts and behaviour change

… social scientists can productively study and inform decision-making and policy processes. Social science efforts form part of a global conversation that includes citizens around the world, activists and social movements, and a variety of institutions and organizations ranging from communities to nations to international organizations. In addition to operating in empirical and theoretical registers, social science scholars of climate change can also undertake concrete engagements, seeking to inform adaptation policies, especially in less-developed regions. (Agrawal et al., 2012: 330)

The social sciences have informed adaptation decision-making and policy in multiple ways: there is a large empirical body of work on governing adaptation at multiple scales, across different sectors, traversing natural resource management for adaptation, urban climate governance, to governing transboundary risk (Bhardwaj and Khosla, 2021; Biesbroek et al., 2015; Vij et al., 2024; Ziervogel et al., 2017). Work on the political economy of adaptation (Huitema et al., 2016) and the politics of adaptation planning and investment (Bulkeley, 2019; Vink et al., 2013) highlight how power and actor agendas shape adaptation constraints, decisions, and outcomes. The expanding literature on adaptive governance draws on social learning and environmental governance to synthesise lessons for governing over time scales and under uncertainty. Related to this is the solutions-focused literature on adaptation pathways (Haasnoot et al., 2024; Singh et al., 2024), which examines sequences of adaptation actions that reduce risk and minimise trade-offs and costs of these decisions.

Besides the methodological challenges to enhance tracking and monitoring adaptation implementation and outcomes described earlier, two other substantial gaps remain. First, is the inadequate attention to the behavioural aspects of adaptation, with the discipline of psychology remaining conscious by its relative absence in the adaptation literature (Swim et al., 2011; van Valkengoed and Steg, 2019 review this literature). Hazard studies have examined individual and collective perceptions of risk in detail (e.g. Blaikie et al., 2004; Grothmann and Patt, 2005), while cultural anthropologists have added that religion, politics, cultural norms, and world views mediate risk perceptions and behaviour (Clayton et al., 2015; Jenkins et al., 2018).

However, there is less on linking perceptions of risk to adaptation behaviours and potentially transformational societal change. The homo economicus caricature from behavioural economics paints a rational, profit-maximising individual ⁷ which is far removed from the complex and dynamic decision-making households undertake to manage risk (e.g. see Singh et al., 2016). While scholarship on incentivising mitigation behaviours has burgeoned (e.g. taxes or subsidies to promote adoption of electric vehicles), similar empirical evidence is missing in adaptation (Clayton et al., 2015).

Grothmann and Patt (2005) paved the way for greater attention to the ‘importance of psycho-social factors as elements of adaptive capacity’ (Mortreux et al., 2020: 4) but the theoretical milestones in adaptation behaviour have been somewhat few and far between. A transdisciplinary approach to adaptation-as-behaviour change, is most recently visible in the framing of societal systems transitions (Schipper et al., 2022), which highlights that adaptation (and climate resilient development) requires shifts in attitudes, values, consciousness, and behaviour, which are collectively mediated by societal interests, values, and worldviews. Adaptation researchers must engage with the levers of this societal transition more fully (e.g. Adger et al., 2013), focusing on individual and collective action; social change and learning; perceptions (of risk), and perceived capacities to respond (adapt) as crafted by education, religion, politics, media, etc. Such an expanded interrogation of adaptation acknowledges that people, collectives, institutions, and organisations respond to risk in a cultural context, and in moments of crisis, norms and values that underpin behaviour and the social contract, can change, with implications for how individuals and societies adapt (or not).

Conclusion: Envisioning adaptation futures

From the vantage point of today, the adaptation research community, of which I count myself a part of, has made progress on many fronts. We know the outlines of what effective adaptation should look like but the mechanics of it remain elusive, especially in the context of polycrises, competition for limited funding, and short-term fortress world-esque political priorities, ⁸ and a general lack of consensus on long-term desirable futures. We continue to stumble on foundational definitions, for example, adaptive capacity and adaptation are routinely and erroneously used interchangeably and practical delineations between adaptation and development remain. We have too much data on some things (small-n cases on community-based adaptation in specific places) and negligible data on others (empirical evidence of transformation adaptation). Finally, the psychology of adaptation as a critical lever to adapt needs more attention and will shape how we can incentivise and enable adaptation.

The ‘Cool heads for a hot world – Social sciences under a changing sky’ editorial ends on an evocative note, using the familiar yet cautionary tale of Icarus who flies to close to the sun, which melts his wax wings, and he falls to a gruesome death. Agarwal et al. (2012) asked, did Icarus falter because of an overconfidence in his wings (a technical solution as the editorial highlights) or his outsized ambitions (a psychological failure)? Or perhaps it is the moral failure of not listening to his father's caution and inability to map the landscape of risk. They end that perhaps a ‘well-adapted’ solution, a parachute, could have helped Icarus land, singed and perhaps bruised, but alive nonetheless.

I believe, the imagery of Icarus holds, with one important caveat. In today's unequal world, ⁹ there are many Icarus (and perhaps many suns) – some with solar-powered heat-resistant wings, flying towards the sun, and burning the world in their wake; others with parachutes (heat alerts) to soften the blow; a few Icarus have energy bars to keep them going, they may have rudimentary face shields. Some Icarus have ‘relocated’ to the moon, while others have pooled money to build a cheap airship. And then there's a vast sea of would-be Icarus, land-bound and vulnerable, who watch their skies clouded with fliers but are unable to fly themselves. They clean the debris of the fallen solar wings, they sweep away the fallen Icarus, and they continue to live flightless lives.

As adaptation science grows and matures, and adaptation implementation becomes more inevitable and challenging, the gaps in and between knowledge and practice need interlocutors and bridges. This is the knowledge frontier is adaptation science.