Abstract
This article advances theory and methods for integrating sustainable livelihoods approaches (SLAs) with assessments of adaptive capacity to climate change. The livelihoods concept has been inconsistently applied in research on human dimensions of global environmental change, resulting in limited understanding about how development programmes and policies influence adaptive capacity. Encouraging reflection on the conceptual and methodological overlaps of livelihoods and adaptation, I suggest a process-oriented approach to adaptation that centres on how adaptive capacity is unevenly shaped. Livelihoods analytical frameworks can help visualize complex adaptation pathways, illuminating how households and individuals come to differ in their capacities to adapt to climate change.
Introduction
Climate change exacerbates direct and indirect stress on vulnerable people (Adger, 2006; Niang et al., 2014) and adaptation will in many cases be essential (Mearns and Norton, 2009). With negative impacts of increased climatic variability and uncertainty likely to be most pronounced among resource-dependent societies in the Global South (Niang et al., 2014; Thomas and Twyman, 2005), there is further urgency to incorporate climate change adaptation within broader development policies, projects and programmes (Lemos et al., 2007). Yet, the social-ecological complexities underlying development and adaptation have continuously stymied attempts to merge the concepts in research and practice (Agrawal and Lemos, 2015).
Indeed, projects aiming to facilitate adaptation to climate change have frequently proven deficient (Nelson et al., 2016; Shinn, 2016), in some instances even leading to further marginalization of vulnerable groups (Lemos et al., 2016). This relates in part to the fact that households face myriad challenges beyond climate change (Berrang-Ford et al., 2011; Eakin, 2005). Yet, attempts to reconcile this issue by mainstreaming climate adaptation into development have proven similarly inadequate, frequently mired by continued reliance upon standard development concepts and mechanisms (Eakin et al., 2014; Lemos et al., 2013; Tanner and Horn-Phathanothai, 2014). Challenges of tracking adaptation at the national level (Ford et al., 2013) and comparing it across contexts (Dupuis and Biesbroek, 2013) have also played a role in hindering assessments of the impacts of development policy on adaptation.
In light of these concerns, scholars have called for tighter integration of adaptation and development. Bassett and Fogelman (2013) advocate for ‘transformative adaptation’ that actively addresses underlying societal inequities that are seen to be at the core of vulnerability to climate change. Agrawal and Lemos (2015) propose ‘adaptive development’, in which development policies and programmes explicitly take into account the myriad climate-related risks facing households and systems. Others similarly underscore social transformation as integral to effective adaptation (Pelling, 2011; Pelling et al., 2015; Ribot, 2014). However, the conceptual and methodological underpinnings of transformative adaptation and adaptive development have yet to be mapped out. It is here where the present article enters the discussion. Taking these calls for integrating climate and development as a starting point, I draw attention to the promise of livelihoods analytical frameworks for bringing clarity to the assessment of adaptation processes.
Sustainable livelihoods approaches (SLAs), in their several variants, have influenced development studies and practice over the past three decades (Scoones, 2009). SLAs are premised on the concept that, throughout the world, people compose complex and dynamic livelihood portfolios that are attenuated by climatic, political and economic variabilities and uncertainties (Ellis, 2000). Early livelihoods frameworks (e.g., Bebbington, 1999; Chambers and Conway, 1992; Scoones, 1998) implicitly addressed adaptation by highlighting the ingenuity of rural producers, who adopt new strategies to persevere through political-economic and environmental changes. However, bringing livelihoods explicitly to bear on questions of climate adaptation has proven challenging; SLAs have seldom informed adaptation research (Carr, 2014; Eakin et al., 2014). This owes to the inability of SLAs to account for social, political and environmental variability over time (de Haan and Zoomers, 2005; Scoones, 2009) as well as to development’s penchant for a ‘narrowly economic, cultural ecological construction of livelihoods’ (Carr, 2015: 333).
The lack of analytical clarity in livelihoods research is matched by similar ambiguities surrounding climate change adaptation, as highlighted by recent debate (Bassett and Fogelman, 2013; Berrang-Ford et al., 2011; Lorenz et al., 2014). In effort to move forward with assessment of adaptation and development, this article takes to heart Scoones’ (2009: 182) conviction that a ‘central future challenge must be integrating livelihoods thinking and understandings of local contexts and responses with concerns for global environ-mental change’. Specifically, I aim to demonstrate that SLAs, as comprehensive analytical frameworks that bridge the social and the ecological, can ground adaptation research to enable articulation with key concerns of development policy and practice (Lorenz et al., 2014).
I argue that SLAs can productively frame research on adaptation and development by anchoring to the assessment of adaptive capacity. Adaptive capacity—the ability to adaptively respond to changing and variable climates—has rapidly gained traction in climate and development circles and details of its assessment are still being hammered out (Engle, 2011). Scholars have recently called for understanding of how adaptive capacity is composed of both specific (climate targeted) and generic components (Eakin et al., 2014; Lemos et al., 2016). I suggest first that SLAs are well positioned to improve the assessment of adaptive capacity by helping to empirically decipher how specific and generic capacities interact. Second, I discuss how livelihoods insight helps to re-focus adaptation and development research onto the social-ecological processes by which adaptive capacity is unevenly shaped. The more common interpretation is of adaptive capacity as a static entity that households have or lack to varying degrees. In contrast, a process-oriented understanding of adaptive capacity permits us to visualize how households (or other units of organization) come to differentiate in their abilities to adaptively manage uncertainty. Importantly, such an approach helps us to appreciate the roles of institutions and development policies in enabling or constraining adaptive capacity.
This article proceeds with review and discussion of how SLAs have evolved over the past three decades. I then review key concepts at the climate adaptation-development inter-face, concentrating on the assessment of adaptive capacity as a crucial area of overlap between livelihoods and adaptation. Finally, suggesting how this overlap may be built upon, I put forth an ‘adaptive livelihoods’ heuristic, discussing methodological and conceptual strategies to garner empirical insight into development and adaptation to climate change.
Livelihoods and Adaptation to Climate Change
Livelihoods and Social-ecological Change
Originating in the 1980s as a response to single-sector, top-down development interventions and assessments, SLAs offer a place-based perspective that concentrates upon the activities that people engage to make a living and the material and social assets that facilitate those activities (Chambers, 1983, 1987). Although SLAs began with an overt emphasis on maximizing household material productivity, livelihoods scholars soon came to advocate for a more thorough analysis of the social networks and institutional power dynamics in which households are enmeshed (Scoones, 1998), asking how livelihoods intersect with resource governance regimes (Leach et al., 1999) and how immaterial as well as material rationales may underlie decisions about making a living in particular places (Bebbington, 2000). The several variants of the livelihoods analytic (prominently: Ashley and Carney, 1999; Bebbington, 1999; Ellis, 2000; Leach et al., 1999; Scoones, 1998) all promised a nuanced and comprehensive approach to considering local impacts of and responses to development programmes and policies. Often via case studies, SLAs illustrated development as differential challenges and opportunities arising largely due to variations in capitals and capabilities among households and social groups (Bebbington, 1999).
Ontologically, livelihoods frameworks centre on human agency. They are premised on the idea that rural producers develop livelihood strategies from various combinations of assets (or capitals, including natural, social, cultural, human and economic) and entitlements (the ability to make use of assets), which are delimited through formal and informal resource-regulation institutions (Bebbington, 1999; Leach et al, 1999; Scoones, 1998). Social differentiation (subjectivities such as gender, class and caste) is further acknowledged to shape access and entitlements to these capitals, as is the notion of ‘political capital’, although there has been some disagreement regarding the latter’s inclusion alongside Bebbington’s (1999) five original capitals (Scoones, 2009). Accordingly, assets are seen to vary across space and over time (both between and within households), and as relational in the sense that the livelihood decisions of some actors may create fluctuations in the assets and entitlements of others (Ellis, 2000). A typical starting point for livelihoods research has hence been investigation of the mutable processes shaping access to assets. Bebbington (1999), for example, considers assets as generating the capability of people to be and to act, including the possibility to change their worlds through re-negotiation of power structures imbued in resource management institutions.
Epistemologically, livelihoods research emphasizes empirical explanations of local contexts and how they may change over time, including consideration of the ramifications of social-ecological changes on capitals and entitlements. Interdisciplinary methodology (qualitative and quantitative methods and concepts from social and natural sciences) is considered essential to depict the range of social and ecological processes that may factor into households’ decision-making surrounding maintenance of and adjustments to livelihoods (Scoones, 1998). In this vein, some early livelihoods scholarship dealt with adaptation as a comprehensive process (e.g., Davies, 1996; Davies and Hossain, 1987).
Over the past two decades, scholars have employed SLAs to advance situated understandings of local roles in trajectories of agrarian change and the impacts thereupon. As one key contribution, Bebbington (1999) argued that studies on agrarian transitions tend to bifurcate, either supporting status quo development programmes with copious economic data or critically deconstructing such initiatives as inherently flawed, calling for detailed material and discursive understanding of the development process through ethnographic and historical methods. Bebbington (2000) views livelihood production as negotiation—between structural constraints of development and the agency of households and individuals—about place. Others have added that place-making activities of livelihood production entail political economic, cultural and ecological components that intertwine through history (McSweeney, 2004), and have emphasized the importance of understanding livelihood diversification over time (Ellis, 2000). Adding further nuance to livelihood production, King (2011) draws attention to the contested making of rural spaces, demonstrating how historically rooted power hierarchies are built into the spatial fibre of livelihoods, which can enable or constrain decision-making and resource access.
Notwithstanding these strengths of SLAs and the abundance of engaged research that they have born over the past three decades, numerous shortcomings have dulled their analytical facility and curtailed earnest adoption into development practice (Carr, 2015; de Haan and Zoomers, 2005). Despite flexibility, interdisciplinarity and comprehensiveness, the livelihoods concept has arguably been only superficially adopted into mainstream development. A chief reason for this is that SLAs have privileged economic aspects of decision-making given the more straight-forward applicability of economic data to quantitative analysis and dissemination of results (Scoones, 2009). Thus, even with ballooning interest in livelihoods among non-governmental organizations (NGOs) and United Nations (UN) organizations, the tendency has been to sideline the ‘complex processes requiring in-depth qualitative understandings of power, politics and institutions’ (Scoones, 2009: 178).
By these accounts, SLAs have clearly strayed from the original aim of accounting for a fuller range of social, cultural and ecological variables than permitted by development economics. SLAs have also deviated in terms of attention to climate coping and adaptation. As Chambers and Conway (1992: 26) defined it:
a livelihood is sustainable which can cope with and recover from stress and shocks, maintain or enhance its capabilities and assets, and provide sustainable livelihood opportunities for the next generation; and which contributes net benefits to other livelihoods at the local and global levels and in the short and long term.
Although these sustainability criteria resonate with contemporary understandings of adaptation, SLAs have fared poorly with such assessments (de Haan and Zoomers, 2005).
Indeed, a major source of continued scepticism surrounding SLAs has been their seeming inability to address issues of global change. The short-term, case-study approach is seen as unable to capture longer-term or multi-scale changes (de Haan and Zoomers, 2005; Scoones, 2009). Some, however, indorse the continued usefulness of SLAs. Lemos et al. (2013) suggest mobilizing Bebbington’s (1999) ‘capitals and capabilities’ framework for assessing adaptive capacity and Carr (2014: 110) maintains that ‘livelihoods approaches remain the only broad framework of analysis that allows for the holistic investigations necessary to address issues of vulnerability and resilience at the heart of contemporary development discourse and practice’. Before returning to the potential of SLAs in such analyses, I first review scholarship on human dimensions of climate change.
Vulnerability, Resilience and Adaptive Capacity
Two principal approaches to assessing climate change impacts are resilience (more system-oriented) and vulnerability (more actor-oriented). The ontological and epistemological discrepancies between these approaches have made it a challenge to bring them together. Adaptation scholarship and practice have more recently converged on adaptive capacity, which some view as capable of bridging the divide between resilience and vulnerability (Engle, 2011). I review each of these three concepts in turn.
Rooted in the ecological sciences, resilience took shape in ecosystem dynamics research. It denotes the degree to which a system can fluctuate without shifting into another state (Berkes et al., 1998; Nelson et al., 2007). Emphasizing relationships between components, the concept has proved most useful at ecosystem levels. On the other hand, more localized social factors receive limited attention in resilience frameworks (Adger, 2006) and scholars call for increased attention to the political economic underpinnings of resilience to make the approach policy-relevant (Folke et al., 2002; Turner, 2014).
By contrast, vulnerability has its origins in natural hazards research, which favours case-study approaches focusing on household heads as rational economic decision-makers. Vulnerability scholars have more recently considered how structural political economic aspects shape ‘social vulnerability’ (Adger, 2006). Key aspects of vulnerability include exposure (likelihood and magnitude of climate events), sensitivity (the degree to which a population may be impacted) and adaptive capacity (the ability to adjust so as to reduce sensitivity and/or exposure) (Gallopin, 2006). Vulnerability has proved most useful at local levels, where case studies can identify vulnerable people or groups. Yet, the concept offers little in the way of analysing relationships between groups or the effects of development and adaptation policies on vulnerability (Hinkel, 2011). And vulnerability analyses have arguably overvalued quantifiable components seen as readily comparable across contexts, with the effect that few assessments actually conceptualize adaptation (Tschakert et al., 2013).
More generally, some argue that both resilience and vulnerability perspectives have privileged technological mechanisms, often neglecting political, economic, social and ecological features that could permit more transformative adaptation policies (Bassett and Fogelman, 2013). These numerous deficiencies of vulnerability and resilience are arguably rooted in the concepts’ disciplinary origins (Engle, 2011). The concepts are seen as inadequate for mainstreaming adaptation within development (Agrawal and Lemos, 2015).
One adaptation concept with potential to help overcome the polarized limitations of vulnerability and resilience is adaptive capacity. Also referred to as adaptability, adaptive capacity denotes the ability to prepare for or respond to stresses (both actual and perceived) as well as the effects of those stresses (Engle, 2011; Parry et al., 2007; Smit and Pilifosova, 2001). Adaptive capacity studies concentrate on adaptation processes to identify what challenges are faced by whom and why (Adger et al., 2009). From these studies, adaptive capacity appears unevenly distributed across households, communities, regions and other analytical levels (Adger et al., 2007; Goldman and Riosmena, 2013). Importantly, adaptive capacity is relevant to both resilience (as the ability of human actors to enhance system resilience or influence transition to another stable state) and vulnerability frameworks (where it refers to the ability of actors to reduce sensitivity and/or exposure to stressors). Some have even argued that adaptive capacity can help bridge resilience and vulnerability (Engle, 2011); others add that it can facilitate more direct engagement with broader development goals (Eakin et al., 2014; Lemos et al., 2013).
Despite the theoretical promise of adaptive capacity, lack of empirical documentation has meant only sparse understanding of its distribution patterns (Lemos et al., 2013). This owes to the challenges of assessing adaptive capacity, which lie in its latent nature: adaptive capacity can only be identified after it is called into action by extreme weather events or conditions of longer-term climatic stress (Engle and Lemos, 2010). Indeed, adaptation studies have only rarely documented adaptation actions (Berrang-Ford et al., 2011). The disciplinary trappings of resilience and vulnerability frameworks (discussed above and elaborated on in Engle, 2011) also limit adaptive capacity assessments. For instance, from a vulnerability perspective, adaptive capacity studies consider a single spatial scale (generally the community) over a short-time period that corresponds with a specific climatic event and tend to privilege specific components (e.g., actors), with less attention to relationships between component parts. While these snapshot analyses have proved highly tractable—producing place-specific data that are readily categorized, ranked and mapped—for policy and practice, they cannot detect dynamic processes that may unfold across multiple spatial scales over longer time periods (Gallopin, 2006). From resilience perspective, adaptive capacity assessment does consider broader temporal and spatial units that enable visualization of such dynamics, including relationships between components. Yet, this tends to produce data that are more difficult to translate into actionable policy and practice given that the spatial units common to resilience studies (i.e., watershed) seldom correspond with administrative boundaries upon which projects and policies are based (Engle, 2011).
These shortcomings of adaptation research have repercussions for development studies in that we have little empirical or conceptual understanding of the influences of development policies on adaptation trajectories (Lemos et al., 2013). A burgeoning literature is seeking to correct this by paying greater attention to how local actors and institutions shape opportunities and constraints of adaptation (Agrawal, 2010; Agrawal and Lemos, 2015; Lemos et al., 2013; Tschakert, 2007). Such a premise will of course be familiar to livelihoods scholars, who have long recognized that rural households weave management of climatic risks into the very fibre of daily activities that constitute a livelihood. For adaptation research, this represents a new and promising direction. As Eakin et al. (2014, p. 653) recommend, ‘rather than focusing on which specific adaptations were most effective during these periods of gradual and rapid change, it might be more useful to understand what structures, relationships, processes, and other variables allowed for (or blocked) the facilitation of such adaptations (i.e., adaptive capacity)’. I argue that such a view necessarily opens up adaptive capacity assessment to insight from livelihoods research. It is to the juxtaposition of livelihoods and adaptation that I turn in the following section.
Ontological and Epistemological Overlap between Livelihoods and Adaptation
Whether viewed from vulnerability or resilience perspectives, climate change adaptation involves modifications to human-environment systems so as to mitigate or avoid negative impacts and even take advantage of new opportunities. Environment-development scholars have recognized linkages between adaptation and political economies of natural resource access and use, drawing attention to the state-society negotiations about sharing risks and benefits (Adger et al., 2003). Among resource-dependent societies throughout the Global South, household capitals are expected to facilitate the majority of adaptations, yet some suggest that links between development and adaptation may be difficult to track given the plethora of other factors shaping local experiences with development (Eakin, 2005; O’Brien et al., 2004). As such, adaptation research can benefit from a more practical, empirical and participatory approach to understanding local experiences with changing climatic conditions (Smit and Wandel, 2006). Livelihoods approaches allow a clear window into these complexities and, I argue, can help visualize links between local experiences and extra-local processes.
The unique applicability of SLAs for querying adaptation and development lies in the fact that they overlap resilience and vulnerability approaches both epistemologically and ontologically. Figure 1, based on Engle (2011), depicts these overlaps: research methods on the horizontal axis (ranging from more structural and ecological to more agency-oriented and social) juxtaposed to concepts on the vertical axis (ranging from abstract and long timeframes to applied and short timeframes). Attention to adaptive capacity serves as a link between resilience and vulnerability (Engle, 2011). Onto this, I position SLAs, which tend to overlap more with vulnerability approaches, given their orientation towards empirical data on local-level aspects of household agency and also their proclivity towards social and applied research. Like vulnerability research, livelihoods research is generally framed as a snapshot in time.
In spite of these conceptual parallels, livelihoods and adaptation frameworks have been seldom integrated. Where livelihoods-based studies have considered adaptation and adaptive capacity, it has most often been implicitly, with little attention to longer-term or broader-scale social and biophysical dynamics (Scoones, 2009), thereby masking the potential value of SLAs in investigating adaptation and development. Yet, the model of place-based studies that assess development in terms of adaptation, resilience and vulnerability is relatively common in the field of political ecology (Bassett and Fogelman, 2013; Turner, 2014, 2015). As one example, Carr (2008) demonstrates the worth of attending to gendered livelihood practices as a way of questioning what we mean by a successful adaptation by showing that in the face of economic and environmental change in rural Ghana, men gravitate towards livelihood diversification not as an adaptation that enhances household security, but as one that upholds gender roles about use of income.
With impetus from this political ecology work, I suggest that SLAs can effectively be brought to bear on issues of human dimensions of climate change. I see particular promise in the concept of adaptive capacity. Not only does adaptive capacity act as a bridge between vulnerability and resilience frameworks (Engle, 2011), the concept has arguably always been integral in livelihoods research in the sense of possibility for transformation, whereby individuals can wield assets so as to ‘change the dominant rules and relationships governing the ways in which resources are controlled, distributed and transformed into income streams’ (Bebbington, 1999: 2039); or what Scoones (2009: 173) has called ‘contested patterns of livelihood change’. Moreover, with a conceptual commitment to adaptation as a process situated between structure and agency (Carr, 2008; Carr and McCusker, 2009), livelihoods frameworks are uniquely positioned to garner insight into how adaptive capacity is unevenly shaped. I discuss this in the following section alongside efforts to integrate livelihoods and adaptive capacity, offering suggestions for strengthening these linkages.
Incorporating Livelihoods Frameworks into Adaptation–Development Research
Merging Livelihoods and Adaptation
Several authors have begun mobilizing livelihoods insights towards advancing theory and assessment of adaptive capacity within contexts of development. This work centres largely on vulnerability analysis. For example, Eakin and Bojorquez-Tapia (2008) constructed a model for vulnerability assessment at the household level through decision-analysis drawing on the livelihood capitals. Lin and Chang (2013) insert vulnerability concerns into SLAs and demonstrate how individuals draw on social and cultural capitals to promote adaptation to natural disasters. Lemos et al. (2013: 448) put forth the idea of adaptive development, which they argue should give ‘specific attention to how risk management intersects (positively and negatively) with policies aiming at economic growth, human and sustainable development’ to bolster conceptual ties between adaptation and development. These authors position the five livelihood capitals as relatively straightforward generators of adaptive capacity. With a more comprehensive take on the livelihoods analytic, Carr (2014: 112) suggests that assessment of development policy and programmes should consider livelihoods as ‘intimate governance’ that can help ‘rigorously explain the holistic character of vulnerability and resilience’.

These moves to merge livelihoods ideas into research on adaptation have yielded important insights that can be built upon. For example, Lemos et al. (2013) draw attention to how the determinants of adaptive capacity mirror the five livelihood capitals and suggest the growing need for development to account for livelihoods in contexts of heightened climatic risk. Connecting this with earlier work on climate change and institutions (Agrawal, 2010), it is crucial to understand the role of institutions in negotiating access to capitals (or the means to generate them). It is paramount to apply livelihoods insight with care, making sure not to overlook persistent shortcomings of SLAs (as discussed in Scoones, 2009). Otherwise we risk superficially including livelihoods language in adaptation studies (Nelson et al., 2007). In the following section, I explore how livelihoods thinking can be more thoroughly integrated with adaptation by considering adaptive capacity as a relational process.
Adaptive Capacity as Process
A primary impediment to integrating SLAs and adaptation lies in the under-conceptualization of livelihoods with regard to dynamic spatio-temporal processes. This shortcoming has, for the most part, been transposed into adaptation research, where generic usage of the term ‘livelihoods’ nullifies its analytical potential to understand realities from local and situated perspectives without ‘imposing artificial categories and divides’ (Scoones, 2009: 172). The generic usage of livelihoods in adaptation scholarship has meant a dulling of the process-oriented approach and has obstructed potential elucidation of how particular livelihood pathways tend to respond to climatic changes within a given context. This is surprising given that the concept of adaptive capacity has long been at the core of development research, albeit in terms of economic and political stressors rather than climatic ones (Lemos et al., 2007). In any case, it amounts to a lost opportunity to understand how to build adaptive capacity for the billions of people who depend on natural resources to make a living. Here, I suggest three ways in which livelihoods frameworks can help understand adaptive capacity as a complex and iterative process:
Livelihoods frameworks can bring clarity to the assessment of adaptive capacity. While adaptive capacity is recognized as a ‘latent condition’ that can only be observed and ascertained when called into action through some sort of ‘concrete adaptation’ (Lemos et al., 2007), livelihoods are immanently observable, if complex. This ready observability of livelihoods makes them a more feasible entry point in assessment of adaptive capacity. While scholars have noted challenges in assessing how shifting institutional variables produce ebbs and flows in adaptive capacity over a longer timeframe (Engle, 2011) and patterns of variation between and within households and communities (Goldman and Riosmena, 2013), the tangibility of livelihoods can serve as a crucial anchor. It bears mentioning that certain complexities of human-environment interaction (e.g., intra-household decision-making, gendered norms of environmental governance or complex power relations) are difficult to apprehend even using livelihoods analytics, yet SLAs are arguably better-equipped than either resilience or vulnerability frameworks to make sense of the complexities of human-environment systems undergoing change (Carr, 2014). Livelihoods can help comprehend the complex social-ecological processes that shape adaptive capacity. Working to develop protocol for the measurement and characterization of adaptive capacity remains essential (Engle, 2011). However, protocols that force concentration upon certain quantifiable variables risk obscuring the complex social-ecological interplay from which adaptations spring to life. One of the most compelling reasons for incorporating livelihoods into adaptation research is the possibility for understanding how adaptive capacity is unevenly shaped. This means visualizing mechanisms, categorizing their relationships and identifying feedback patterns (whether positive or negative). Such a move would dovetail with a burgeoning literature seeking to understand how specific and generic adaptive capacities interact (Eakin et al., 2014; Lemos et al., 2013) both synergistically and as trade-offs and poverty traps (Lemos et al., 2016). Empirical exploration of livelihoods (which are at once material entities and socially constructed) can help re-centre adaptive capacity assessment. Framing adaptive capacity as dynamically produced through negotiations between institutions and households about resource entitlements and fundamental aspects of social-ecological difference will also help to understand the links between adaptation and development. This includes identifying whether development policies and programmes should target specific or generic capacities according to which is more likely to optimize both short-term coping and longer-term adaptation (as Lemos et al., 2013 suggest). Interestingly, resilience approaches have long considered both generalized and specific drivers of change in ways that can usefully inform this analysis (Gunderson and Holling, 2001). Empirical attention to livelihood activities helps visualize links between social and biophysical. Scholars employing vulnerability and resilience frameworks often separate social and biophysical components that are dynamically linked (Turner, 2014). SLAs theoretically offer an interdisciplinary platform for investigating complex human-biophysical systems, which institutions of resource access are seen to regulate (Leach et al., 1999; Scoones, 1998). In practice, however, livelihoods scholars often emphasize either social or biophysical (Scoones, 1998). To take into account social-ecological links that shape resilience, vulnerability and adaptive capacity, it is useful to consider how land use strategies (e.g., agricultural intensification or extensification) relate to livelihoods. Exploration of what Turner (2014) calls ‘land use ecology’ can provide important insight as to how households may be pushed or pulled into unsustainable land use practices with a motive of short-term coping or longer-term adaptation. While such a framing is quite common in political ecology studies of environment-development, this approach has not been used to consider adaptation to climate change.
The Uneven Production of Adaptive Capacity
Adaptive Capacity as Relational
It is a common practice in adaptation research to view adaptive capacity as something that households either have or lack to varying (quantifiable) degrees. The above three benefits of instead considering adaptive capacity as something that is produced (both actively and inactively) can be realized through incorporation of SLA insights into adaptation research. Moreover, SLAs can also help us see patterns of how adaptive capacity is unevenly shaped. In other words: why do some come to have adaptive capacity and others not? And why can some individuals transform generic capacities (such as land and social capital) into specific capacities (such as dry-season irrigation technology entitlements) while others cannot? Capitals and entitlements (and related land use patterns) are at the core of why some households are able to leverage generic capacities to build specific capacities to climatic uncertainty and change (leading to sustainable adaptation) while other households must continually draw from generic capacities in order to cope with climatic events, leading to poverty trap scenarios.
Livelihoods are viewed as produced through negotiations at the intersection of structure and agency that can be reworked (Bebbington, 2000; King, 2011; Zimmerer, 1996). Likewise, climate change adaptations can result through (and spur contentious negotiations about) the seemingly pre-ordained viability of certain livelihoods; the essence of such negotiations can be reworking or maintenance of power differentials along lines of gender or class (Carr, 2008). In these ways, vulnerability and adaptive capacity are seen to be relational: increases for some individuals, households or communities often mean corollary decreases for others (Goldman and Riosmena, 2013; Shinn, 2016; Turner, 2015). Along these lines, I suggest that drawing from SLAs can further elucidate the relational pathways by which adaptive capacity may come to differentiate in terms of class, caste, ethnicity, gender or other subjectivities. Since a specific climatic event might necessitate coping and/or adaptation among highly differentiated social actors, the value of one actor’s adaptive capacity must be understood as relational to that of others, who may be adapting to the same (or different) events.
In conceptualizing adaptive capacity as a relational entity, it becomes apparent that even where a household might seem to have high specific adaptive capacity within one area (say, a village), if a climatic event is to affect a larger area (as is often the case with drought, for example), then that household’s adaptive capacity may need to be re-assessed according to the capacity of others in the larger unit of analysis. Recognizing adaptive capacity as relational makes a process-oriented assessment (and livelihoods insights) particularly apt, as they can help document cases where what appeared to be adaptive capacity might in fact prove inadequate if a climate event affects more individuals or a broader area than anticipated. This should provoke reflection on the organizational levels at which adaptive capacity assessment is done (e.g., village, region, etc.). The relational nature of adaptive capacity magnifies the importance of understanding how institutions mediate adaptation, and how institutions are shaped by development policies and programmes. Institutions may either promote collaboration or competition among entities with differential degrees of adaptive capacity. Such institutional analysis is foundational in SLAs.
Adaptive Livelihoods Heuristic
In Figure 2, I present a conceptual heuristic for merging SLA and adaptation insights to consider how adaptive capacity is differentially shaped across multiple levels of social-ecological organization. This heuristic orients social and biophysical components in terms of likely interactions between the macro-level, meso-level(s) and micro-level. Micro-level indicates the smallest unit of analysis (often household or, less commonly, intra-household). Macro-level, in turn, indicates the largest unit of analysis referenced in a study, often ranging from village to nation-state to world region. Meso-levels, which comprise anything in between, have generally received sparse attention yet can be crucial (Faist, 1997). Importantly, the boundaries between levels should be recognized as fluid rather than discrete (Figure 2 represents levels in boxes only for simplicity). The inclusion of meso-level(s) serves to account for boundary fluidity and the important scale-spanning role of institutions. Institutions connect levels of analysis, shaping resource access and adaptation trajectories (Agrawal, 2010; Leach et al., 1999) and thereby mediating adaptive capacity with respect to social and biophysical components of a system. At meso-levels, institutional change or continuity is recursive with social and biophysical change or continuity. In other words, institutions may change due to development policy (e.g., prohibition of subsistence agriculture), to uphold a particular social order in light of biophysical changes (e.g., crop insurance to incentivize commercial agriculture even with drought risk), or to uphold a preferred biophysical characteristic in light of social change (e.g., to maintain extensive land use strategies even with increased population).

While political, economic, social and biophysical changes at the macro-level may directly influence the micro-level, political ecology (Blaikie and Brookfield, 1987; McSweeney, 2004) and development (Faist, 1997) insights suggest that the meso-level(s) may often be more directly relevant to the micro-level. As depicted in Figure 2, the meso-level often corresponds to institutions. Institutional change can lead to shifts in access and entitlements which impact capitals and livelihood decision-making at household and intra-household levels. With capitals in hand (or in mind), households engage in activities that comprise livelihood portfolios which tend to include a combination of land use strategies (broadly either extensive or intensive) as well as non-farm activities. These livelihood options are closely tied to land use ecologies (meaning both how much and how resources are used) (Turner, 2014). In composing livelihood and land use strategies, households invest capitals in ways that may or may not yield returns. These investments arise through capability and entitlements and may result as coping (strategies that maintain or deplete capitals over the longer term), adaptation (adjustments that enhance capitals over the longer term) or under less favourable circumstances could prove to be maladaptations.
It is important to mention that some of the challenges associated with bringing together vulnerability and resilience approaches may also arise in the application of this heuristic, especially given that limited research budgets can necessitate prioritization of methods and limit focus to one or a few levels of analysis. Reflection on these issues is essential for development studies, where ontology and epistemology has been demonstrably opaque (Prowse, 2010). This heuristic is intended primarily to help clarify research design and streamline methodology by spurring reflection on these issues and on hypothetical dynamics of change, including the potential for cross-scale interactions. Place-based/actor-oriented studies as well as larger-scale systems-focused research are essential to furthering robust understanding of adaptive capacity. And, taking insight from resilience literature (e.g., Gunderson and Holling, 2001), it is furthermore crucial to consider recursive links between analytical levels, including upward changes. For instance, household livelihood and land-use decisions can reverberate into the meso-level(s) as social-ecological difference. In other words, adaptive investments have ramifications on social and ecological components of a landscape.
As a few examples, ramifications may include changes to land cover (e.g., decreased forest cover exacerbating erosion risk), institutions (which as Bebbington (1999) and others show can be contested with livelihood decisions), markets (which can be saturated by overproduction, leading to a lower price for all) or sociocultural norms (which may be contested or upheld (Eriksen et al., 2015). Such micro- and meso-level changes can in turn spur adjustments within other meso-level components or even those at the macro-level. Through such cross-scale processes, adaptive capacity is produced, contested and reproduced. In other words, a given institutional, biophysical or social regime may impinge on some households’ adaptive capacity while enhancing capacity for others.
While rather simplified, the conceptual map in Figure 2 can serve to guide empirical analysis of adaptation in a process-oriented understanding deriving from SLAs. Such a conceptual framing can help generate empirical understanding of how adaptive capacity is unevenly shaped. Despite the recognized need for empirical documentation of adaptive responses to observe adaptive capacity, Berrang-Ford et al. (2011) find that few studies of adaptation report on concrete adaptation actions. Livelihoods insights can help re-centre adaptation research on activities that people do and how these activities may be altered in response to (or anticipation of) climate risk or social change. More specifically, a livelihoods perspective on adaptive capacity can productively identify which sets of assets—mediated by which institutional processes—tend to create what type of adaptive capacity for which households. To put this in terms of justice implications: it is essential to understand not only what adaptive capacity is and who has it in a given context but also to ask how that capacity came to be distributed as it is.
Moreover, the above heuristic begs consideration not only of responses to a drought or other climatic shock but also the responses to the outcomes of those responses. This will allow for identification of patterns in how responses (broadly either coping or adaptation) inflect further change that shapes adaptive capacity. Differently put, the outcomes of a response—whether success or failure—would be assumed to factor into subsequent adaptive capacities. When considered empirically, it should become evident that this is not a trivial issue. Some responses might end up enhancing adaptive capacities in the longer term while others might not. Such an approach aligns with the suggestion by Wise et al. (2014: 325) to ‘reconceptualize adaptation to climate change as part of pathways of change and response’. This longer timeframe understanding may help scholars and practitioners identify cyclical vulnerabilities and how different social groups gain and lose adaptive capacity with institutional shifts, thereby uncovering synergies and trade-offs in development programmes and policies with regard to building specific and generic adaptive capacities.
Importantly, the determination of ‘successes’ and ‘failures’ of development and adaptation projects is a value-laden enterprise. Various social groups may likely prioritize adaptations differently. A successful adaptation for one group in one location might even be seen by others as a maladaptation. As one example, development policy emphasizing commercial agriculture may be seen by wealthier large-landholder households as commensurate with building adaptive capacity through generating increased income that can be reinvested in land and used to buy crop insurance. Poorer households, on the other hand, may view such a policy as increasing exposure to climate risk while decreasing adaptive capacity by pushing them away from lower-risk subsistence-based systems. As another example, men may make adaptive decisions based on pre-existing gender roles, gravitating towards an adaptation of limiting spouses’ means of production in order to retain more revenue but in the process decreasing adaptive capacity of women and of the household overall (Carr, 2008). These examples suggest that differential perspectives on adaptation make a fruitful juncture to investigate power dynamics underpinning the (often uneven) distribution of adaptive capacity. They further demonstrate the value of empirical research into these issues.
Empirical understanding of the production of adaptive capacity will make it more straightforward to identify adaptive capacity when it is called into action as well as to ask why in some situations adaptive capacity may not have been utilized to its full potential. This question is crucial for investigating the politics of adaptation and one that SLA insights can help address. With a few exceptions, adaptive capacity has been implicitly considered as technical and in accordance with rational action. The logical assumption is that if capacity exists then it will indeed be called into action. However, this rationalized view sidesteps the political nature of adaptive capacity.
Indeed, numerous reasons come to mind as to why adaptive capacity would not be exercised under certain scenarios. Hypothetical rationales fall somewhere on the spectrum from altruistic (e.g., an individual may elect not to overdraw common pool resources because doing so would inhibit others from accessing key resources) to benign (a household may not want to sell their cows due to the cultural importance of cows) to nefarious (with coffee supply low due to drought, a powerful coffee buyer may capitalize on inflated prices but decide not to recalibrate what he pays farmers so as to compensate them for financial losses of a low-production year). In each of these cases, it is more analytically insightful to consider adaptive capacity as imbuing political economic and cultural agendas rather than strictly economic rationales. The above heuristic can help to uncover such insight. This heuristic should certainly not be considered a panacea to the challenges of connecting more actor-oriented and more systems-oriented ontologies, but as a lens that drives constructive reflection on research design.
Conclusion
This article has reflected on the potential for more profound incorporation of SLAs into climate and development research. The empirical scaffolding of livelihoods, which balances on everyday tasks that people do and how those articulate with social, environmental and institutional contexts, holds promise for querying relationships between development programmes/policy and adaptive capacity. A key question emerging from this review is how is adaptive capacity produced?
A process-oriented view of adaptive capacity urges us to ask how generic capacities interact with specific capacities to enable responses to real or perceived climatic variability and/or uncertainty. Such a perspective is necessary to overcome the limitations associated with viewing adaptive capacity as a quantifiable thing that households either have or lack. Attention to the activities and institutions comprising livelihoods can enhance understanding of linkages between social and biophysical aspects underlying the complex and uneven process of climate change adaptation. Concentrating on the processes by which adaptive capacity is unevenly produced will, moreover, help shed light on how various social groups come to differentiate in their abilities to adaptively respond to climatic variability and change. This is essential for investigating development within contexts of increasingly variable and uncertain climate. Shifting focus to the production of adaptive capacity also helps to address calls for ‘adaptive development’ and ‘transformative adaptation’ (Agrawal and Lemos, 2015; Bassett and Fogelman, 2013).
Development policies and programmes play critical—yet often complex—roles in shaping adaptive capacity. Attending to the tangible and intangible aspects of liveli-hoods can help uncover these complexities to better inform policy for incorporating climate change adaptation concerns within broader development trajectories. With clearer vision of the pathways to building adaptive capacity, arising unevenness can more effectively be mitigated.
