1. Introduction
Social vulnerability assessment (SVA) is important for addressing rising concerns about the impacts of natural hazards and climate change (Adger, Reference Adger1999; Cutter, Reference Cutter1996; Kelly & Adger, Reference Kelly and Adger2000). Social vulnerability refers to the degree of susceptibility of human systems (including individuals, communities and institutions) to adversity (Adger, Reference Adger1999; Bevacqua et al., Reference Bevacqua, Yu and Zhang2018; Cutter et al., Reference Cutter, Boruff and Shirley2003). Since its emergence, the concept has evolved, gained prominence in a multi-disciplinary context, and multiple methodologies have been developed to assess social vulnerability. These methodologies largely focus on (i) deductive approaches to identify easy to measure indicators (i.e. variables that reflect past or present conditions within a limited temporal range, and at a single spatial scale) and (ii) developing quantitative methods for aggregation and evaluation of these indicators (Aroca-Jimenez et al., Reference Aroca-Jimenez, Bodoque, Garcia and Diez-Herrero2017; Bjarnadottir et al., Reference Bjarnadottir, Li and Stewart2011; Fernandez et al., Reference Fernandez, Bucaram and Renteria2017; Mason et al., Reference Mason, Lindberg, Haenfling, Schori, Thomas, Popovich, Faulkner, Beban, Gunnel, Marsters, Read and Borman2019; Spielman et al., Reference Spielman, Tuccillo, Folch, Schweikert, Davies, Wood and Tate2020; Yoon, Reference Yoon2012).
Conventional quantitative SVA methodologies based on place-based indices (Cutter et al., Reference Cutter, Boruff and Shirley2003; Flanagan et al., Reference Flanagan, Gregory, Hallisey, Heitgerd and Lewis2011, Reference Flanagan, Hallisey, Adams and Lavery2018) can provide an overview of the social vulnerability landscape. However, when facing a constantly changing environment, or multiple plausible futures, these methodologies are insufficient to capture or reflect the complexities of near- and long-term social vulnerability. These limitations reduce the usefulness of quantitative SVA for decision-making, as a critical shortcoming in SVA is the limited ability to address the complex dynamics of linked systems of humans and nature (Adger et al., Reference Adger, Brown and Surminski2018; Chuang et al., Reference Chuang, Garmestani, Eason, Spanbauer, Fried-Petersen, Roberts, Sundstrom, Burnett, Angeler, Chaffin, Gunderson, Twidwell and Allen2018). In particular, traditional methods fail to adequately account for the qualitative and hard to measure aspects of social systems (Fawcett et al., Reference Fawcett, Pearce, Ford and Archer2017), the uncertainty of future scenarios of change (Adger et al., Reference Adger, Brown and Surminski2018), and system connectivity and interactions across multiple temporal and spatial scales (Chuang et al., Reference Chuang, Garmestani, Eason, Spanbauer, Fried-Petersen, Roberts, Sundstrom, Burnett, Angeler, Chaffin, Gunderson, Twidwell and Allen2018; Fawcett et al., Reference Fawcett, Pearce, Ford and Archer2017). For example, Chuang et al. (Reference Chuang, Garmestani, Eason, Spanbauer, Fried-Petersen, Roberts, Sundstrom, Burnett, Angeler, Chaffin, Gunderson, Twidwell and Allen2018) are critical of static social vulnerability analyses based on snapshots in time and space. Static analyses do not consider that linked systems of humans and nature are dynamic social-ecological systems (SES), which have multiple scales, can exist in different configurations (e.g. coral dominated or algae-dominated coral reefs), and interact at various levels and may respond to an impact or a change on a different scale (Gunderson & Holling, Reference Gunderson and Holling2002). Further, many SESs are non-stationary, as baseline conditions measured previously have changed, negatively affecting responses unless this non-stationarity is considered (e.g. Nalau et al. Reference Nalau, Torabi, Edwards, Howes and Morgan2021). Some scholars have highlighted the limitations of these SVA methodologies and discussed problems such as the need for robust conceptual underpinnings in defining the ‘socioeconomic’ vulnerability profile that does not lead to subjective vulnerability assessment (Hufschmidt, Reference Hufschmidt2011), monitoring and evaluating the effectiveness of adaptation responses (Fawcett et al., Reference Fawcett, Pearce, Ford and Archer2017; Hinkel, Reference Hinkel2011), incorporating the complex dynamics of SESs in future scenarios of change (Adger, Reference Adger2000; Fawcett et al., Reference Fawcett, Pearce, Ford and Archer2017), and inclusion of cascading and cumulative impacts of drivers at different locations and spatial scales (Adger et al., Reference Adger, Brown and Surminski2018; Lawrence et al., Reference Lawrence, Blackett and Cradock-Henry2020a; Thomas et al., Reference Thomas, Hardy, Lazrus, Mendez, Orlove, Rivera-Collazo, Roberts, Rockman, Warner and Winthrop2019).
Most coastal governance and management adopts an engineering, community or disaster resilience mindset to better predict hazards, apply engineering strategies to minimise or mitigate impacts, increase system robustness to maintain system state, and allow for rapid recovery to restore the previous configuration (‘return to normal’) (Adams-Hutcheson et al., Reference Adams-Hutcheson, White, Glavovic and Lawrence2019; Garmestani et al., Reference Garmestani, Craig, Gilissen, McDonald, Soininen, van Doorn-Hoekveld and van Rijswick2019a; Pelling, Reference Pelling2010). Coastal governance based upon engineering, community or disaster resilience has been imported to the governance of coastal SES without adequately acknowledging the fundamental differences between different conceptions of resilience (e.g. engineering resilience, community resilience, disaster resilience VS. social-ecological resilience (sensu Holling, Reference Holling1973)).
These problems are fundamental limitations to SVA, as well as engineering, community or disaster resilience discourses for the governance of SES. Here, we seek to improve SVA for coastal governance by incorporating a SES perspective. A SES is a self-organising, dynamic and complex system of humans and nature, characterised by resilience, adaptability, transformability, and multiple scales (panarchy) (Gunderson & Holling, Reference Gunderson and Holling2002). Accounting for SES dynamics at multiple spatial and temporal scales is necessary to increase the accuracy and application of SVA to social vulnerability in real-life (Garmestani et al., Reference Garmestani, Twidwell, Angeler, Sundstrom, Barichievy, Chaffin, Eason, Graham, Granholm, Gunderson, Knutson, Nash, Nelson, Nystrom, Spanbauer, Stow and Allen2020). Importantly, the definition of resilience adopted for coastal governance has critical ramifications for the types of policies and management applied to coastal zones globally (Allen et al., Reference Allen, Angeler, Chaffin, Twidwell and Garmestani2019).
We adopted the definitions of resilience offered by Allen et al. (Reference Allen, Angeler, Chaffin, Twidwell and Garmestani2019) to bring clarity for coastal governance. Allen et al. (Reference Allen, Angeler, Chaffin, Twidwell and Garmestani2019) discussed three main definitions of resilience for SESs: (i) resilience as a rate; (ii) resilience as a process; and (iii) resilience as an emergent property. Resilience as a rate is shaped by an engineering mindset (engineering resilience or resiliency) (Allen et al., Reference Allen, Angeler, Chaffin, Twidwell and Garmestani2019; Holling et al., Reference Holling, Gunderson, Peterson, Gunderson and Holling2002). In this definition, resiliency is a term frequently used to address a system's robustness, its ability to resist change, and the capacity to prevent a regime shift (Holling, Reference Holling and Schulze1996; Holling et al., Reference Holling, Gunderson, Peterson, Gunderson and Holling2002). Resilience as a process, which is mainly applied in community resilience and disaster resilience, emphasises building, maintaining, or enhancing system resilience (Allen et al., Reference Allen, Angeler, Chaffin, Twidwell and Garmestani2019).
Resilience as an emergent property of SESs (social-ecological resilience) explicitly accounts for the possibility that SESs can exist in different configurations or regimes (Gunderson & Holling, Reference Gunderson and Holling2002; Wilson et al., Reference Wilson, Pearson, Kashima, Lusher and Pearson2013). Social-ecological resilience is an overarching approach to explain the dynamics of SESs, as it encompasses adaptation but also transformation, as well as panarchy (scale and cross-scale interactions) (Folke et al., Reference Folke, Carpenter, Walker, Scheffer, Chapin and Rockström2010; Gunderson & Holling, Reference Gunderson and Holling2002; Holling et al., Reference Holling, Gunderson, Peterson, Gunderson and Holling2002). Panarchy, or ‘nature's rules’, incorporates hierarchy and dynamism and accounts for the interconnectedness of social-ecological resilience and vulnerability on multiple spatial and temporal scales (Allen et al., Reference Allen, Angeler, Garmestani, Gunderson and Holling2014; Garmestani et al., Reference Garmestani, Twidwell, Angeler, Sundstrom, Barichievy, Chaffin, Eason, Graham, Granholm, Gunderson, Knutson, Nash, Nelson, Nystrom, Spanbauer, Stow and Allen2020).
Each definition of resilience could be useful and fit the purpose for a specific context. However, the differences between their core assumptions create framing that hinders interchangeable and generic applications for SES. For example, resilience as a process is useful to address management strategies for enhancing community resilience to some natural hazard events (i.e. 1% Annual Exceedance Probability coastal inundation). But for a coastal community that is frequently exposed to inundation (due to climate change impacts), coastal governance should adopt the social-ecological resilience framing in order to allow for adaptation (e.g. protecting coastal wetlands) or transformation (e.g. managed retreat of a community away from its current coastal location).
Adaptation is the process of adjusting to changing conditions but remaining in the same configuration with a similar set of processes and structures defining that regime. Transformation is the process of shifting a SES with the human agency to a more desirable configuration with a different set of processes and structures than the previous configuration. Transformation is a response to the heightened vulnerability that rarely gets invoked, because community and disaster resilience have focused on the rate or process definitions of resilience (i.e. engineering resilience, community resilience or disaster resilience). The resilience as a rate or process definitions fail to account for non-stationarity in systems, and the presence of critical thresholds or tipping points, that when exceeded, will result in the system of interest shifting to a new regime with different processes and structures. Given that coastal SESs, like other SESs, are non-stationary and do have thresholds, consideration of potential alternatives for coastal governance when such a threshold is reached is critical (Lawrence et al., Reference Lawrence, Blackett and Cradock-Henry2020a). Acknowledging these thresholds, and the potential for SES to fundamentally re-organise, provides an opportunity to change systems with high vulnerability through transformation (Lawrence et al., Reference Lawrence, Blackett and Cradock-Henry2020a).
Here, we use social-ecological resilience to present a novel framing of social vulnerability and argue that social-ecological resilience is an essential consideration for SESs by addressing the neglected aspects of social vulnerability such as cross-scale interactions (temporal and spatial scales) and transformation (Gunderson & Holling, Reference Gunderson and Holling2002). We focus on the links between social-ecological resilience and vulnerability (Allen & Holling, Reference Allen and Holling2010; Gallopín, Reference Gallopín2006) to discuss: (i) how these concepts have been framed and applied in practice. Framing, as we address in this research, refers to the relationship between understanding (ontological and epistemological aspects) of polysemic concepts or boundary objects such as resilience and vulnerability (Baggio et al., Reference Baggio, Brown and Hellebrandt2015; Strunz, Reference Strunz2012), its perception between users (mindset), and defining its contributing variables (Fairhurst & Sarr, Reference Fairhurst and Sarr1996; Tannen, Reference Tannen1993); (ii) how the definition of resilience affects the framing of vulnerability; (iii) how the framing of social vulnerability could influence its assessment outcomes; and (iv) how the application of these different concepts could impact the implementation of adaptive and transformative coastal governance in New Zealand (and globally), with a qualitative two-stage analytical approach consisting of Critical Discourse Analysis (CDA) and key informant interviews.
2. Methods
A qualitative two-stage analytical approach including a CDA and key informant interviews was adopted (Hajer, Reference Hajer, Howarth and Torfing2005; van den Brink & Metze, Reference van den Brink, Metze, van den Brink and Metze2006; Weiss & Wodak, Reference Weiss and Wodak2007). While key literature, including the existing literature reviews (Davidson et al., Reference Davidson, Jacobson, Lyth, Dedekorkut-Howes, Baldwin, Ellison, Holbrook, Howes, Serrao-Neumann and Singh-Peterson2016; De Sherbinin et al., Reference De Sherbinin, Bukvic, Rohat, Gall, McCusker, Preston, Apotsos, Fish, Kienberger and Muhonda2019; Ferro-Azcona et al., Reference Ferro-Azcona, Espinoza-Tenorio, Calderón-Contreras, Ramenzoni, País and Mesa-Jurado2019; Moser et al., Reference Moser, Meerow, Arnott and Jack-Scott2019; Siders, Reference Siders2019) was considered in the CDA process, conducting a separate ‘systematic literature review’ was outside the scope of this research.
CDA is often used in social science research to examine the relationships and connections between the meaning of a concept, its perception among users, and its application in social and political contexts (Catalano & Waugh, Reference Catalano and Waugh2020; Rogers, Reference Rogers2004). Our CDA was not meant to provide a detailed typological review based on individual articles. Instead, CDA was used to deconstruct discourses: principally natural hazards, disaster management and climate change discourses, and analyse trends in social vulnerability framings. Tannen (Reference Tannen1993) found that while different discourses could create multiple framings of a concept (i.e. social vulnerability), framing, in turn, delivered a valuable basis for shaping discourses and their practical application.
As our primary database for CDA, the Web of Science was searched for articles published on the topic of coastal social vulnerability to climate change. Keywords including ‘social vulnerability’, ‘coast*’, and ‘climate change’ were used in multiple combinations to filter and screen the peer-reviewed publications between 2000 and 2020. Our search identified ninety-three papers (Supplementary Table 1). Also, the Google Scholar database was searched to include significant research that was possibly missed during the Web of Science search (20 more papers were reviewed).
We also conducted a series of semi-structured interviews to acquire in-depth, first-hand and practical information about the framing of vulnerability, and its influence on developing and applying SVA frameworks in New Zealand's coastal governance and management (as the case study for research leading to this article). Interviewees were selected:
(1) from decision-makers, managers, and practitioners involved in the development or application of SVA. Ten interviews were selected: eight from regional and local councils (the government organisation responsible for applying SVA in practice); one from the Ministry for the Environment (the Central Government agency responsible for developing SVA guidance); and one leading researcher engaged with developing SVA in New Zealand (Supplementary Table 2).
(2) representing a variety of councils with a diversity of social and ecological features (i.e. economically privileged and non-privileged municipalities).
Interviewees were asked five to six open-ended questions regarding their perceptions of social vulnerability (and transformation), its application in the current coastal governance and management practice, and the features of a forward-looking SVA framework to improve decision-making and policy development in practice (Supplementary Table 3). To preserve the participants' anonymity during the analysis, each participant was allocated a unique code. Interview transcription files were analysed with NVIVO 12 software.
3. Results
3.1. Framing of resilience in the coastal literature
Social vulnerability was commonly defined as a measure of a social system's susceptibility to adversity due to the lack of coping capacity, adaptability or recoverability (Adger, Reference Adger2000; Cinner et al., Reference Cinner, McClanahan, Graham, Daw, Maina, Stead, Wamukota, Brown and Bodin2012; IPCC, 2014; Shaw et al., Reference Shaw, Scully and Hart2014). In these contexts, quantitative assessments of social vulnerability indicators were the primary method of SVA (González-Baheza & Arizpe, Reference González-Baheza and Arizpe2018; Hagenlocher et al., Reference Hagenlocher, Renaud, Haas and Sebesvari2018). Some coastal SVA research either did not address the term resilience or applied it in a generic fashion without any particular definition or framing (Cinner et al., Reference Cinner, McClanahan, Graham, Daw, Maina, Stead, Wamukota, Brown and Bodin2012; Hardy & Hauer, Reference Hardy and Hauer2018; Kelly & Adger, Reference Kelly and Adger2000). Risk, hazard, disaster, recovery, adaptation, coping capacity and resilience were among the most frequent terms.
In other research, resilience as a process (Mussi et al., Reference Mussi, Bonetti and Sperb2018; Orencio & Fujii, Reference Orencio and Fujii2013; Shao et al., Reference Shao, Jackson, Ha and Winemiller2020) and resilience as a rate (Lam et al., Reference Lam, Reams, Li, Li and Mata2016; Silver et al., Reference Silver, Arkema, Griffin, Lashley, Lemay, Maldonado, Moultrie, Ruckelshaus, Schill and Thomas2019; Wu et al., Reference Wu, Jhan, Ting, Tsai, Lee, Hsu and Liu2016) were the most frequent definitions of resilience. These discourses discussed the resilience and vulnerability of a social system through three perspectives:
(1) Enhancing resiliency or resistance: commonly mentioned in the research influenced by the rate definition. Resilience was associated with an intrinsic or pre-existing capacity to reduce system vulnerability. Avoiding or mitigating hazard exposures and risks were the main strategies for enhancing system resiliency and reducing vulnerability (Colburn et al., Reference Colburn, Jepson, Weng, Seara, Weiss and Hare2016; Ge et al., Reference Ge, Dou and Zhang2017; Martins & Gasalla, Reference Martins and Gasalla2020).
(2) Coping capacity (adaptation) and adaptability (adaptive capacity): more frequent in research that defined resilience as a process. These terms implied enhancing (or building up) resilience as the pre-disturbance capacity to improve a social system's adaptability, maintaining its existing state, and avoiding regime shifts (Gerrity & Phillips, Reference Gerrity and Phillips2020; Smith et al., Reference Smith, Lieske, Keys and Smith2018).
(3) Return and recovery: common in both rate and process discourses, addresses post-disturbance capacity in human systems to bounce back when a system is out of equilibrium (Adger, Reference Adger2000; Kelly &Adger, Reference Kelly and Adger2000). Returning to a system's optimal or equilibrium state was more noticeable in the rate discourse (Shaw et al., Reference Shaw, Scully and Hart2014). In the process definition, recovery mainly indicated the capacity to return to a general pre-disturbed state, or ‘return to normal’ regardless of optimality or desirability of that state (Bennett et al., Reference Bennett, Kadfak and Dearden2016; Chang et al., Reference Chang, Yip, Conger, Oulahen and Marteleira2018; Martins & Gasalla, Reference Martins and Gasalla2020).
Despite the growing application of social-ecological resilience in environmental governance and management research in the last couple of decades (Angeler et al., Reference Angeler, Allen, Garmestani, Pope, Twidwell and Bundschuh2018; Garmestani & Benson, Reference Garmestani and Benson2013; Jozaei et al., Reference Jozaei, Mitchell and Clement2020), this discourse has not been well considered in SVA research. Adger et al. (Reference Adger, Hughes, Folke, Carpenter and Rockström2005) were among the few scholars who adopted social-ecological resilience and argued the role of multi-level governance, with cross-scale connections, to create both adaptive and transformative capacities in coastal SESs. Noting the clear limitations of adaptive governance with respect to scale, cross-scale interactions and inadequate consideration of law, Garmestani and Benson (Reference Garmestani and Benson2013) combined adaptive governance, panarchy and law to create resilience-based governance of SES with applications for any SES, but with clear ramifications for coastal governance (e.g. Florida, USA).
Although several SVA scholars mentioned social-ecological resilience in their research, there was no evidence that the concept was adopted to create a social-ecological resilience framing. For example, Hagenlocher et al. (Reference Hagenlocher, Renaud, Haas and Sebesvari2018, p. 76) addressed the vulnerability of coupled ‘social-ecological systems’, but the term was applied as a more suitable ‘unit of analysis’ in hazard and risk assessment (Hagenlocher et al., Reference Hagenlocher, Renaud, Haas and Sebesvari2018). The authors instead adopted a resilience as a process discourse (‘build resilience’ or ‘enhance resilience’) with no clear link to social-ecological resilience (resilience as an emergent property) of SESs (Hagenlocher et al., Reference Hagenlocher, Renaud, Haas and Sebesvari2018).
Dominated by resilience as a process and resilience as a rate definitions, adaptive capacity was identified as the primary capacity required to reduce social vulnerability. From these perspectives, adaptive capacity refers to the ability of a social system (i.e. a community) to cope with hazards and risk drivers, recover after being disturbed, return to a normal state, and avoid transformation (Delfino et al., Reference Delfino, Dizon, Quimbo and Depositario2019; Freduah et al., Reference Freduah, Fidelman and Smith2018; Joakim et al., Reference Joakim, Mortsch and Oulahen2015).
This characterisation of adaptive capacity is very different from research on social-ecological resilience, where adaptive capacity is the potential of a SES to alter resilience in response to change and maintain the current regime, while transformative capacity is the potential of a SES to fundamentally change its processes and structures with human agency and reorganise into a new regime (Garmestani et al., Reference Garmestani, Ruhl, Chaffin, Craig, van Rijswick, Angeler, Folke, Gunderson, Twidwell and Allen2019b). Since social-ecological resilience received little consideration in SVA research, transformative capacity and its role in addressing social system vulnerability was disregarded or ambiguously addressed in coastal SVA research.
Finally, the results showed inadequate consideration of scale and cross-scale spatio-temporal connectedness in addressing social vulnerability. Of the literature reviewed, only a few researchers addressed cross-scale interactions as an aspect of resilience or as a way to reduce vulnerability (Cutter et al., Reference Cutter, Boruff and Shirley2003; Thomas et al., Reference Thomas, Hardy, Lazrus, Mendez, Orlove, Rivera-Collazo, Roberts, Rockman, Warner and Winthrop2019).
3.2. Framing of social vulnerability in New Zealand coastal governance practice
The analysis of interviews demonstrates the predominance of engineering, community, and disaster resilience discourses amongst New Zealand's coastal governance and management practitioners. The key terms frequently observed in responses included: hazard, risk, disaster, coping, and recovery. Resilience as a process was the primary definition used by coastal zone practitioners. As one participant discussed, New Zealand's SVA and risk assessment system has a focus on natural hazard and disaster vulnerability: ‘I think we are reasonably well serviced for the sort of traditional disaster type of event, the major event that causes damage.’
Resilience as a rate was evident in the interviews through addressing resistance or avoiding change, timely recovery and ‘back to stability’. Most importantly, there was only one reference to social-ecological resilience. In general, most participants referred to the IPCC definition of vulnerability as a useful and practical framing. This definition discusses vulnerability as a concept that incorporates a variety of variables such as ‘sensitivity or susceptibility to harm’ and a lack of ‘coping’ and adaptive capacity (IPCC, 2014).
In this context, adaptive capacity was frequently mentioned as a key means for dealing with change and uncertainty. Amongst the interviewees, adaptive capacity was frequently defined as a social system's (or a community's) ability to adapt, retain or maintain its existing state, bounce back, build back (or build back better), return to normal, or recover after turbulence. Again, this definition differs significantly from the definition of adaptive capacity from social-ecological resilience, where adaptive capacity is the potential of a SES to alter resilience in response to change and maintain the current regime (Garmestani et al., Reference Garmestani, Ruhl, Chaffin, Craig, van Rijswick, Angeler, Folke, Gunderson, Twidwell and Allen2019b). Before the questions related to transformative capacity were discussed with participants, only one interviewee directly discussed transformation as an aspect of social-ecological resilience, but argued transformation as an attribute of adaptive capacity:
“…There are probably different ways you could look at it [adaptive capacity]. So, you could say, well, there's the adaptive capacity that allows us to return to the existing status quo, stable state. Ok. [or] you could also think of adaptive capacity in the … and now we're getting back into that resilience as transformation … What resilience is actually is the community thriving in some ways, and it doesn't have to replicate itself as … all of those foundations are evolving over time anyway. So, we might see over time there's been a move from, sheep farming in New Zealand to dairy… because they invented nylon and didn't need to farm sheep for wool anymore. There was a bottom dropped out of the market. So, they transformed the economy.”
Moreover, a few participants indicated the multifaceted and complex nature of concepts such as vulnerability, adaptive capacity, and resilience and argued that various attributes are involved in their framing. One interviewee argued:
“… I think actually, adaptive capacity, we just don't understand what that is. I really don't think anyone understands, and it's just a convenient term. There have probably been some theoretical descriptions of it. However, I don't think in practice it has so many meanings to so many people.”
Another participant added that resilience and adaptive capacity are more challenging to understand and apply in the context of incremental and cumulative drivers such as sea level rise:
“… With ongoing sea level rise, it's kind of a difficult concept, I think, because with the ongoing sea level to rise for several centuries. It's never going to be completely resilient unless we make some major transformational changes, move away or do something radical or move ahead of the curve, so to speak, of sea level.”
In the last part of the interview, transformation and transformative capacity were discussed with the participants. We shared our interpretation of transformation as a system fundamentally changing its key processes and structures with human agency and shifting to a new configuration (see Section 1.). The analogy used to better communicate transformative capacity was a coastal community which is capable of: (i) changing its main functional attributes to, local coastal tourism for example, if farming, as a traditional way of living, is not an option anymore; or (ii) undergoing managed retreat and relocating to another place if adaptation strategies are not feasible.
Although one interviewee argued that transformation is complex or is a ‘leap too far’, others found transformative strategies to be a proactive response to the uncertain impacts of climate change: ‘[transformative strategies] would be more like a proactive vs reactive kind of response.’ Most participants asserted that a coastal society with transformative capacity is less vulnerable than a society without it. In their accounts, transformation is a useful means by which social systems could address the complexity and uncertainty of future scenarios of climate change and reduce vulnerability through adaptive learning and novel experimentation. Therefore, interviewees mentioned that although the current SVA methodologies in New Zealand do not account for transformation, it should be incorporated in a future-oriented SVA methodology:
“… They [community] need to evolve; they need not go back to the baseline but somehow grow and strengthen from that position. So yeah, it [transformative capacity] is fundamental. Transformation is fundamental to improving resilience or reducing vulnerability.”
Other issues raised in the conversations were scale mismatches and inadequate consideration of synergistic relations between different temporal and spatial scales. The changing nature of social vulnerability, through time and across levels (e.g. community, regional or national levels) was argued as a significant component of an effective SVA. Participants discussed that an SVA framework should be forward-looking and deliver a long-term and systems-level (multi-scale) understanding of social vulnerability. For example, some interviewees argued that while some adaptation strategies might decrease community vulnerability in the short term or at a particular location, those strategies could undermine resilience and increase vulnerability in the long term at other locations. In addition, participants discussed that while a community group might have adequate adaptive capacity to cope with climate change impacts, the entire system (including other communities and their governance systems) might still be vulnerable.
Overall, the results of the interview analysis suggested the initial prevalence of resilience as a process, and adaptive capacity as a means of recovery and generally maintaining a system's state amongst New Zealand coastal managers and practitioners. However, when the idea of transformation was discussed, most participants confirmed that transformative capacity plays an important role in reducing social vulnerability. Therefore, a clear understanding of the type of adaptive capacity and identifying components of transformative capacity should be included in developing vulnerability assessment methodologies. Conceptualising system vulnerability, lack of consideration of spatial and temporal scales (cross-scale interactions), absence of forward-looking, non-prescriptive and holistic approaches, and disregarding uncertainty of future scenarios of change in assessing vulnerability were other significant problems of conventional SVA raised in the interviews.
4. Discussion
The results of the CDA demonstrate that coastal social vulnerability research is still dominated by the engineering, community and disaster resilience discourses, which mainly define resilience as a rate or resilience as a process (Gerrity & Phillips, Reference Gerrity and Phillips2020; Smith et al., Reference Smith, Lieske, Keys and Smith2018). The case study interview analysis revealed that this prevalence extends to a smaller scale, namely New Zealand's coastal governance and management. Although resilience as a rate is commonly applied in coastal physical and infrastructure vulnerability assessments, our findings indicate that this definition has been applied in some SVA contexts, both in the literature (Colburn et al., Reference Colburn, Jepson, Weng, Seara, Weiss and Hare2016; Ge et al., Reference Ge, Dou and Zhang2017; Martins & Gasalla, Reference Martins and Gasalla2020) and amongst practitioners. Social-ecological resilience (resilience as an emergent property of SESs) and its core ideas (e.g. multiple regimes or configurations, adaptation and transformation, panarchy) were seldom addressed in coastal research and amongst the interviewees. For example, Adams-Hutcheson et al. (Reference Adams-Hutcheson, White, Glavovic and Lawrence2019) examined the dominance of the disaster resilience discourse in New Zealand (due to the long history of exposure to natural hazards such as tsunamis and earthquakes) and discussed the requirement of ‘fit-for-purpose resilience’ to respond to the dynamics and uncertainty of climate change impacts.
Influenced by the resilience as a process or rate definitions, a framing of vulnerability as an antonym of resilience (i.e. system vulnerability is the flip side of its resilience) is prevalent in coastal research globally (Chen et al., Reference Chen, Liu, Chen and Zhao2020; Yang et al., Reference Yang, Lin, Zhang, Ye, Chen, Jin and Ye2019) and found through interviews with New Zealand coastal governance experts. Both the rate and process definitions created a normative perspective of resilience and vulnerability (high resilience and low vulnerability are desirable), assuming higher system resilience leads to higher stability and lower vulnerability (Harvey & Woodroffe, Reference Harvey and Woodroffe2008; Hufschmidt, Reference Hufschmidt2011), despite the fact that highly resilient systems may be in a very undesirable state (Zellmer & Gunderson, Reference Zellmer and Gunderson2008).
However, some international researchers and interview participants disputed the prevailing disaster resilience discourse, focusing on vulnerability to rapid, episodic and catastrophic drivers. This discourse underpins adaptive capacity for resisting change, system robustness (maintaining system state) and recovery from adversity (O'Connell et al., Reference O'Connell, Walker, Abel and Grigg2015; Reghezza-Zitt et al., Reference Reghezza-Zitt, Rufat, Djament-Tran, Le Blanc and Lhomme2012; Siders, Reference Siders2019). Further, the results of the CDA and interview analysis suggest that the outcomes of conventional SVA based on engineering, community and disaster resilience discourses do not adequately address cumulative and cascading impacts of ongoing climate change, potentially generating inaccurate or skewed results, which could lead to undesirable consequences for coastal SESs (Adams-Hutcheson et al., Reference Adams-Hutcheson, White, Glavovic and Lawrence2019; Fawcett et al., Reference Fawcett, Pearce, Ford and Archer2017; Hinkel, Reference Hinkel2011).
It is clear that given the reality of accelerating environmental change, social-ecological resilience is an essential consideration for vulnerability research, particularly for coastal areas facing ongoing sea-level rise. Social-ecological resilience can address the shortcomings of conventional SVAs based on resilience as a process and resilience as a rate by accounting for scale, cross-scale interactions and the possibility that SESs can exist in different configurations (Folke et al., Reference Folke, Carpenter, Walker, Scheffer, Chapin and Rockström2010; Garmestani et al., Reference Garmestani, Twidwell, Angeler, Sundstrom, Barichievy, Chaffin, Eason, Graham, Granholm, Gunderson, Knutson, Nash, Nelson, Nystrom, Spanbauer, Stow and Allen2020; Holling et al., Reference Holling, Gunderson, Peterson, Gunderson and Holling2002). We call this framing general vulnerability, which, like social-ecological resilience, is a descriptive term and could be either negative or positive.
4.1. Specific vulnerability vs general vulnerability
As our findings indicate, social vulnerability primarily indicates the specific vulnerability of a social system, or a particular sector within it (e.g. significant cultural sites, infrastructure, households, farming or housing sector) to a specific (or set of) drivers (e.g. floods, erosion, tsunami, other storm events) (Koks et al., Reference Koks, Jongman, Husby and Botzen2015; Mason et al., Reference Mason, Lindberg, Haenfling, Schori, Thomas, Popovich, Faulkner, Beban, Gunnel, Marsters, Read and Borman2019). As discussed previously (Janssen et al., Reference Janssen, Anderies and Ostrom2007), SVA outcomes based on specific vulnerability mainly inform strategies to reduce the susceptibility of social systems to a specific hazard, without adequately accounting for: (i) broader scale vulnerability or differential vulnerability (Thomas et al., Reference Thomas, Hardy, Lazrus, Mendez, Orlove, Rivera-Collazo, Roberts, Rockman, Warner and Winthrop2019) and the vulnerability of multiple sectors within the social system (or the SES across different spatial scales); (ii) unknown, unknowable, cumulative and cascading drivers (Adger et al., Reference Adger, Brown and Surminski2018; Lawrence et al., Reference Lawrence, Blackett and Cradock-Henry2020a); and (iii) the dynamics of vulnerability and its changing nature across multiple spatio-temporal scales (Chuang et al., Reference Chuang, Garmestani, Eason, Spanbauer, Fried-Petersen, Roberts, Sundstrom, Burnett, Angeler, Chaffin, Gunderson, Twidwell and Allen2018).
With respect to the dynamics of vulnerability over time, some interviewees argued that although hard infrastructure and engineering adaptation measures to sea-level rise (e.g. stop banks, levees, dykes, sea walls) might reduce the vulnerability of local properties to coastal flooding, these strategies can increase social vulnerability at the broader regional scale, and at medium and long-term time horizons. Further, participants maintained that under ongoing sea level rise, focusing on vulnerability strategies (e.g. return to normal, build back, or build back better in the same location) might reduce susceptibility in the short-term, but will likely inhibit preparations for transformation such as managed retreat and increased social vulnerability over the long-term (Lawrence et al., Reference Lawrence, Boston, Bell, Olufson, Kool, Hardcastle and Stroombergen2020b).
Moreover, we found that traditional SVA methods based on specific vulnerability, do not take adequate account of the dynamics of SESs in a changing environment. Disregarding the interconnectedness between human and natural systems could create skewed assessment outcomes. For instance, rural systems have different dependencies to particular ecosystem services or functions, which influence their respective vulnerability. A rural community based on tourism (as the main function of that society) might have a different vulnerability to a specific climate hazard (i.e. drought), compared with an adjacent farming community. However, the traditional SVA indicators (i.e. based on age, income, gender ethnicity brackets) do not capture these functional attributes.
Transformative capacity (Garmestani et al., Reference Garmestani, Ruhl, Chaffin, Craig, van Rijswick, Angeler, Folke, Gunderson, Twidwell and Allen2019b), as a central notion in social-ecological resilience, is a useful concept to address the general vulnerability of a SES. For example, instead of adaptation, if the SES is in an undesirable state coastal governance could intervene and facilitate the transformation of the SES to a more desirable state (Figure 1). Coastal governance could target capacity building and providing incentives to vulnerable coastal communities to shift their main economic focus and transform their SES if the traditional way of living becomes less viable (for instance, from agriculture to locally-based coastal tourism).
The relationship between vulnerability and adaptive capacity for SES. When a SES has low vulnerability and low adaptive capacity, build or foster the resilience of the existing regime (enhance resilience). When vulnerability is low and adaptive capacity is high, governance continues ‘as is’ (business as usual). When vulnerability is high and adaptive capacity is low, governance should encourage a ‘return to normal’ (bounce-back). When vulnerability is high and adaptive capacity is high, governance should look to transform the SES to a new regime. While adaptive capacity is similar to transformative capacity there are differences in these two capacities of SES. In particular, adaptive capacity underpins the resilience of an individual regime, while transformative capacity is the underlying potential of a regime to transform (via human agency) to a new, more desirable regime (see Garmestani et al., Reference Garmestani, Ruhl, Chaffin, Craig, van Rijswick, Angeler, Folke, Gunderson, Twidwell and Allen2019b).
For general vulnerability, rather than prescriptive and deductive approaches to reduce vulnerability to a particular driver, SVA incorporating social-ecological resilience would encourage capacity building in SESs in response to unforeseeable and unknown drivers of change (Carpenter et al., Reference Carpenter, Arrow, Barrett, Biggs, Brock, Crépin, Engström, Folke, Hughes and Kautsky2012, Reference Carpenter, Folke, Scheffer and Westley2019).
Most of the interviewees agreed that a coastal community with more intention to build transformative capacity would be less vulnerable than a locality without it. In their account, transformative capacity is a useful means by which social systems could respond to the complexity and uncertainty of future scenarios of climate change and should be considered in the SVA process:
“… And, what we need to start looking at is a transformation of how we understand … transformability is a really important concept here. And maybe that's more of a better way of looking at it [adaptive capacity] actually is rather than, what are we being resilient to here?”
The connectedness of resilience and vulnerability across multiple spatio-temporal scales was also addressed as another essential consideration for reform of the SVA process. Panarchy is a nested set of adaptive cycles and has important ramifications for vulnerability (Gunderson & Holling, Reference Gunderson and Holling2002). Panarchy is increasingly being recognised as a strong foundation for delivering more holistic and flexible governance and management interventions under change and uncertainty (Garmestani & Benson, Reference Garmestani and Benson2013; Gunderson et al., Reference Gunderson, Allen and Garmestani2022). Panarchy concerns the cumulative and cascading effects of resilience and vulnerability across spatial and temporal scales (Allen et al., Reference Allen, Angeler, Garmestani, Gunderson and Holling2014; Gunderson & Holling, Reference Gunderson and Holling2002; Gunderson et al., Reference Gunderson, Allen and Garmestani2022). For example, Garmestani et al. (Reference Garmestani, Twidwell, Angeler, Sundstrom, Barichievy, Chaffin, Eason, Graham, Granholm, Gunderson, Knutson, Nash, Nelson, Nystrom, Spanbauer, Stow and Allen2020) argue the utility of panarchy to address ‘emerging vulnerability’ of small and large scale ecosystems under human-induced change, and discuss panarchy as a useful means for better visualisation and communication of scientific data to non-scientific audiences (i.e. practitioners and communities).
Panarchy describes multi-scale SESs and can address case study participants' concerns about a more holistic and visionary approach in vulnerability assessment. As some participants mentioned, although some strategies might decrease sub-system (i.e. a community) vulnerability on one scale, they may increase the vulnerability of the entire system (i.e. a region or a state). Also, some interviewees discussed that some adaptation options increase long-term vulnerability, for example, through intensification of development and increasing human population, and restricts the application of transformative strategies (such as managed retreat to a new location) when they are required (Lawrence et al., Reference Lawrence, Boston, Bell, Olufson, Kool, Hardcastle and Stroombergen2020b)
Incorporating social-ecological resilience could deliver an updated, non-normative framing of vulnerability (general vulnerability) which can help reform governance responses for coastal SES. Coastal governance that integrates social-ecological resilience into the decision-making approach can foment a regime of ‘good governance’ (Graham et al., Reference Graham, Amos and Plumptre2003) that allows for adaptation to ongoing changing conditions in the current regime (Folke et al., Reference Folke, Hahn, Olsson and Norberg2005; Holling, Reference Holling2001; Walker et al., Reference Walker, Holling, Carpenter and Kinzig2004), or transformation when a system's resilience has been eroded (Chaffin et al., Reference Chaffin, Garmestani, Gunderson, Benson, Angeler, Arnold, Cosens, Craig, Ruhl and Allen2016).
5. Conclusion
We highlight the most significant problem with the traditional framing of vulnerability for coastal governance and management: the use of engineering, community and disaster resilience definitions (resilience as a rate or resilience as a process), which do not account for scale and cross-scale interactions and the potential for transformation of SES. This framing results in a focus on recovery and ‘return to normal’ for coastal governance, which will become more difficult or impossible in the future in the face of rapidly accelerating environmental change (Park et al., Reference Park, Marshall, Jakku, Dowd, Howden, Mendham and Fleming2012). These issues were highlighted by our CDA of the relevant literature and our case study of coastal governance experts in New Zealand, but the findings in this work apply to coastal governance of SES around the globe.
In particular, the spectre of climate change, including sea-level rise and larger and more frequent hurricanes (cyclones), highlights the need for a paradigm shift in SVA. These reforms include meaningful dialogue and collaborative engagement with actors and sectors across scales to prepare governance for adaptation and transformation (when necessary), rather than the prevalent disaster risk reduction framing of vulnerability (i.e, ‘return to normal’).
Finally, we recommend that future research is needed for advancing our knowledge and understanding of the characteristics of SES (e.g. adaptive and transformative capacities), their thresholds and tipping points, the process of evaluating general vulnerability, and the types of governance responses necessary for adaptation or transformation.
Supplementary material
The supplementary material for this article can be found at https://doi.org/10.1017/sus.2022.10.
Acknowledgments
The authors would like to thank Drs. Rob Bell, Judy Lawrence, and Paula Blackett for sharing their ideas with us. We also thank the participants who shared their knowledge and experience through the interview process.
Author contributions
J. J. conceived the study, conducted the interviews and analysis, and wrote the paper. W. C., C. R. A. and A. G. conceived the study and wrote the paper.
Financial support
This research was funded in part by a Post-Doctoral research project (J. J.) from the National Institute of Water & Atmospheric Research (NIWA), and New Zealand's Resilience to Nature's Challenges National Science Challenge (Coastal Theme). Some of the material is based upon work supported by the National Science Foundation under Grant Nos. DGE-1735362 and 1920938 (C. R. A.).
Conflicts of interest
Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (C. R. A.). The research was not performed or funded by EPA and was not subject to EPA's quality system requirements (A. G.). The views expressed in this manuscript are those of the authors and do not necessarily represent the views or the policies of the U.S. government (A. G.). J. J. and W. C. declare no conflicts of interest.
Research transparency and reproducibility
We have provided relevant materials in the supplementary files with this manuscript.
Open access
