In the context of climate change, the concept of vulnerability has been applied to both natural and social systems. The Fifth Assessment Report (AR5) of the Intergovernmental Panel on Climate Change (IPCC) defined vulnerability broadly as “the propensity or predisposition to be adversely affected” by climate events (2014, p. 128). This definition includes three components of vulnerability: exposure to climate hazards, sensitivity of the system, and capacity to adapt.

Many social vulnerability models are concerned with the relationship between these three components, which ultimately determines the vulnerability of a social or a socio-ecological system (e.g., a household, community, institution, region, or country) (Himes-Cornell & Kaspersky, 2015; Maiti et al., 2017; Yenneti et al., 2016). While exposure and sensitivity cause vulnerability, adaptive capacity reduces it; thus “to be vulnerable to the effects of global change, human-environment systems must not only be exposed and sensitive to the effects, but also must exhibit limited ability to adapt” (Polsky et al., 2007, p. 477).

Three general trends can be found in the literature on climate vulnerability and social systems. First, components of sensitivity and/or adaptive capacity are often drawn from pre-existing literature or models, then are applied deductively to analyze the system. For example, such approaches measure how access to resources or capitals (human, social, economic, technological, and natural) affects sensitivity or adaptive capacity (Cutter et al., 2003; Luers, 2005), or how climate change will affect specific economic variables (Akram & Hamid, 2015).

Second, and relatedly, social aspects of sensitivity/adaptation are often understood through observable and measurable factors, such as agricultural management practices and levels of poverty, health, or education (Henríquez et al., 2016; Luers, 2005; Weis et al., 2016). Less attention has been paid to social and economic factors that are more difficult to trace and quantify, or which are sometimes seen as peripheral to the climate stressors —factors Wisner et al. described as “remote root causes” of vulnerability (2004, p. 7). Similarly, Ribot (2014; 2017) emphasized the importance of identifying the generative social, historical, and political-economic structural factors that cause social precariousness in a determined context. For example, beyond poverty as a cause of vulnerability, we must also examine how poverty itself is produced (Ribot 2014; 2017).

A third trend in the existing vulnerability literature is that a vulnerability level is often attributed to an entire system (often a community or region), which has a homogenizing effect inside it. Less attention is paid to power differentials that increase vulnerability in certain people or groups within a community. Vulnerability in social systems is almost always unequally distributed. Social differentiation and inequality on the basis of gender, race, socioeconomic class, and age, among other factors, can create different experiences of a climate event (Cutter et al., 2003; Enarson & Morrow, 1997; Fletcher, 2018; Yadav & Lal, 2018). These inequalities are linked to broader social, economic, and political structures (Fletcher, 2018; IPCC, 2014; Smit & Wandel, 2006; Turner et al., 2003) existent in the system. All these social factors determine how a system may be affected, but the way these effects actually manifest depends on various contextual conditions (Handmer et al., 1999).

Vulnerability assessment is highly context-specific; thus, applying pre-determined vulnerability criteria in a deductive fashion may result in a somewhat circular process of “discovery”. Similarly, focusing only on measurable aspects of vulnerability or adaptation may miss out important cultural, social, political, or economic determinants seen as less relevant to climate change (Ulloa, 2011; Correa, 2012). Even researchers using only a composite vulnerability index to express the aggregation of biophysical exposure, sensitivity, and adaptive capacity recognize the limitations of their method. For example, Zhang et al. (2020) in a climate change vulnerability assessment of the Shenzhen coast in China, highlight that the index is calculated based on a single-time period of datasets although vulnerability is not static; they emphasize that societies respond to hazards in a nonlinear and complex ways.

Other difficulties that some quantitative studies present for the study of vulnerability are of scale: they are usually available for regional or national scales that do not necessarily reveal local vulnerabilities (Krishnamurthy et al., 2014). Another problem is that the weighting of the most relevant components of vulnerability relative to each case may be disregarded (Ashan & Warner, 2014). Previous studies have found that the indicators representing the most methodological difficulties for the study of vulnerability in the cases analyzed herein are those related to the economy, access to water, and infrastructure, technical assistance, associationism, and educational level (Mussetta et al., 2017).

Climate change vulnerability assessment requires methodologies that reveal the differentiated vulnerability patterns of communities and, therefore, help inform appropriate coping and response strategies (Mwangi et al., 2020). Qualitative research methods (which include techniques such as social cartography, oral history, semi-structured interviews, participant observation, and focus groups) allow us to recognize vernacular explanations about climate change, identify local conceptions of space, capture representations about disasters, understand the social, cultural, and environmental context of spontaneous adaptation strategies and understand the reasons for resistance to intervention measures designed by institutions loaded with authority and power (Salinas et al., 2020). Hybrid methodologies combine top-down and bottom-up approaches: they are based on easily quantifiable available data but they also obtain stakeholders’ opinions and values through qualitative methods. A challenge for these hybrid methods is the integration of multiple metrics from dissimilar sources in the final assessment compensated, however, with benefits such as the possibility to reach a fuller community spectrum, the production of more spatially detailed assessments, and potential community empowerment in the planning of adaptive strategies (van Aalst et al., 2008; Cains & Henshel, 2019; Espíndola & Valderrama, 2016). While such hybrid or mixed methods are an important area for future research, this paper aims to help develop the qualitative component of vulnerability analysis, which we suggest has been underemphasized to date. We demonstrate the use of qualitative methods in a community-based research project to help reveal structural aspects of vulnerability in rural communities facing climate extremes.

The fifth IPCC report (2014) briefly recognized the usefulness of qualitative research on climate vulnerability and risk, stating that “useful approaches for managing risk do not necessarily require that risk levels can be accurately quantified. Approaches recognizing diverse qualitative values, goals and priorities, based on ethical, psychological, cultural or social factors, could increase the effectiveness of risk management” (p. 36). This statement reflects the growing acceptance of qualitative, community-based, and participatory approaches to vulnerability research, which offer an important complement to quantitative studies in the literature (Cadag & Gaillard, 2012; Wright et al., 2014). Qualitative and participatory approaches usually focus on stressors and priorities identified by communities, rather than the application of pre-defined criteria by researchers.

Despite this shift, challenges remain for qualitative and community-based approaches. The first challenge is that qualitative approaches to climate research may be viewed by some as “unscientific” (van Aalst et al., 2008, p. 174) and in need of quantification to be taken seriously as “data”. For climate researchers and policymakers operating within a positivist paradigm, community members’ own interpretations and experiences of vulnerability may not be considered a reliable evidentiary base. However, we argue that understanding the meanings people attribute to their experiences can point toward the larger causes of those experiences. Critiques of solely quantitative assessments include the risk of oversimplification due to aggregation (Fernandez et al., 2017) and their inability to reveal the complex relations between actors and resources that shape vulnerability (Montaña & Mussetta, 2016; Retamal et al., 2011). Qualitative approaches, in contrast, have the power to identify deep structural causes of vulnerability that otherwise may not be noticed.

A second and related challenge with qualitative approaches is the dominant “technocratic” approach to adaptation. In this approach, a variety of (usually technological) options are presented as solutions to reduce vulnerability. Infrastructure changes and new crop varieties are often amongst the solutions promoted, regardless of whether these solutions are feasible or acceptable to farmers or communities. Infrastructures designed to reduce community vulnerability to climate change are potentially not reliable for the future if they do not take into account the increase in frequency and intensity of climate extremes, the inner uncertainty of climate change, and the intersections of community, livelihoods, and governance of those infrastructures (McMartin et al., 2018). Without community engagement and input into the research, such technical solutions may ignore communities’ own concerns (Wright et al., 2014). Structural factors may render some “solutions” unusable —or even worse, as maladaptations that exacerbate problems (Barnett & O’Neill, 2010). As our findings demonstrate, for example, technological solutions are not always affordable and over-reliance on these solutions may strengthen social inequalities.

The third challenge is that qualitative community-based studies may be considered contextually limited and only locally useful (Cameron, 2012). Existing literature has drawn attention to structural causes of climate vulnerability in particular areas, for example, HIV/AIDS in Uganda (Niyibizi et al., 2013), colonization in Canada (Cameron, 2012), and poverty in Latin America (Manzanal, 2016). In these cases, vulnerability is shaped by deeper factors that go beyond the climate extreme itself. Although vulnerability will manifest differently in different places, its root causes can be found in broader structures —such as political economy (Ribot, 2014), past and present forms of colonization (Whyte, 2017), and gender inequality (MacGregor, 2010) —that are relatively persistent across time and space.

Broad political, economic, and social conditions that determine vulnerability may be a priori unknown to researchers (Smit & Wandel, 2006), but may become visible in participants’ everyday lives and problems (van Aalst et al., 2008). In the sections below, we present the findings from an intercontinental study using qualitative community vulnerability assessments (CVAs) with over 300 participants in Argentina, Canada, and Colombia. After providing an overview of the conceptual framework and methods of the project, we present key findings from each of the three countries. Beyond simply identifying the features of vulnerability within each region, we identify structural causes of vulnerability that operate across these diverse contexts. Specifically, we find that the most significant vulnerabilities experienced by farmers in each country are linked to unequal access to resources, which is exacerbated by neoliberal macro-level political-economic structures affecting agriculture (Mussetta & Hurlbert, 2020).

Our analysis of structural causes of vulnerability is informed by a particular philosophy of science, Critical Realism (CR) (Archer, 1995; Bhaskar, 2015; Collier, 1994; Danermark et al., 2002). CR suggests that social events and experiences are driven by structurally rooted mechanisms that may not be immediately obvious at the surface. Observable events, such as the social effects of a climate extreme, are interconnected with causal mechanisms operating at a deeper level of reality (Bhaskar, 2015). However, as social phenomena exist in open social systems (i.e., not lab conditions), specific conditions in each context may cause the mechanism to manifest differently. This approach suggests the importance of analyzing everyday experiences for causal roots that may extend across specific contexts but may take slightly different forms in each place.

The intercontinental research project “Vulnerability and Adaptation to Climate Extremes in the Americas” (VACEA) was funded by the International Development Research Centre (IDRC) and conducted by a team of natural and social scientists with the goal of integrating climatological and social determinants of climate vulnerability. This paper draws upon one key component of the project: community vulnerability assessments (CVAs) conducted in three countries of the project. Similar to other types of participatory research —e.g., “practical analysis” (Smit & Wandel, 2006), “participatory vulnerability assessment” (Fazey et al., 2010), “place-based analysis” (Turner et al., 2003), and “bottom up” analysis (van Aalst et al., 2008; Smit & Wandel, 2006) —we used an inductive approach to identify key issues through the participation of stakeholder groups and individual participants. Each CVA focused on rural communities, with attention to agricultural producers.

Each country team used a semi-structured interview method, beginning with open-ended questioning that allowed participants to identify and discuss their greatest stressors rather than using categories imposed by the researchers (van Aalst et al., 2008); these questions were followed by more specific questions about climate extremes in each region. In each study area, researchers partnered with local organizations on fieldwork, interpretation, and knowledge mobilization. Interview data were transcribed verbatim and coded using a flexible directed technique (Hsieh & Shannon, 2005; Gilgun, 2011) that identified very general aspects of vulnerability (e.g., social, economic, environmental) while also capturing other ideas outside these categories. Key findings were later evaluated and discussed in follow-up workshops with participants and community members in each country. During these workshops, participants were also asked to comment on climate scenarios generated as part of the research project.

To compare emerging themes and facilitate cross-case analysis, the researchers met annually for in-person international team meetings. For this paper, we identified and compared the key vulnerability themes (e.g., climatic, economic, social) in each of the three study areas. Although the precise form of vulnerability varied across these countries, we noted the prominence of economic and financial issues as a dominant theme causing producers’ vulnerability in all three contexts. As we describe below, in each case financial precarity of producers can be linked to the broader political economic structure.

The Mendoza River basin was the study area selected in Argentina. The region is in central-west Argentina. The river is mainly snowfed and agriculture is only possible through a complex system of irrigation infrastructure. The research team conducted two focus groups of ten participants each, along with 41 interviews with farmers throughout the districts of the basin and 25 interviews with representatives of national, provincial, and local governance institutions.

In Canada, ethnographic community-based research was conducted in four rural communities in two different watersheds of the South Saskatchewan River Basin: Taber and Pincher Creek in Alberta and Rush Lake and Shaunavon in Saskatchewan. Study areas included towns and their surrounding rural municipal (RM) district(s). All four communities are part of the semi-arid Palliser’s Triangle region —an area known for its history of severe drought (Marchildon et al., 2009; Warren & Diaz, 2012). In-depth interviews were conducted with 160 farmers, ranchers, rural residents, and local governance representatives (e.g., RM councillors, representatives of local watershed groups).

The Colombian research was conducted in the Chinchiná River Basin in the Andes, which exhibits a steep topography and a marked altitudinal gradient from 5,300 to 800 meters above sea level. The basin covers five municipalities: Manizales, Villamaría, Neira, Chinchiná, and Palestina. Coffee cultivation, the most important economic activity, is developed in the steep hill slopes, thus it is highly prone to erosion and highly vulnerable to climate extremes. In-depth interviews (n = 44) were conducted with potato growers living in the moorlands, coffee farmers living in the middle basin, and cattle ranchers, citrus fruit growers, and miners who extract sand and stones from the riverbed near the mouth of the Chinchiná River. Twenty additional interviews were conducted with managing directors of water and energy companies, public environmental authorities, leaders of risk management systems, reforestation company heads, and researchers on coffee and climate change.

The findings from this intercontinental study illustrate the potential of qualitative research to identify causal structures of vulnerability, even across very different contexts. Taken together, these findings clearly reveal how a particular political-economic paradigm constitutes a major source of vulnerability in all three places. Climate extremes interact with existing inequalities to create different levels of vulnerability between producers (e.g., small versus large; highly indebted versus low indebted). These vulnerabilities are further exacerbated by neoliberal policy changes and market rationalities and by worsening climate extremes.

Findings from the three countries converge on neoliberal political-economic trends as a major cause of farmers’ vulnerability to climate extremes. Neoliberalism has been defined as “a theory of political economic practices that proposes that human well-being can best be advanced by liberating individual entrepreneurial freedoms and skills within an institutional framework characterized by strong private property rights, free markets, and free trade” (Harvey 2005, p. 2). Neoliberal political-economic paradigms are often marked by deregulation, financialization, and corporatization. In Argentina, deregulation has facilitated economic concentration and capital accumulation into an oligopoly system, causing financial precarity for grape farmers who are pushed out. A similar trend is occurring in Canada, where farmers are increasing the size of their farms to stay profitable amidst rising costs due to strong corporatization and financialization in the agri-food sector. Such growth, however, is often premised on debt and more expensive inputs, thus increasing farmers’ vulnerability when crops are lost due to climate extremes. For some Canadian farmers, off-farm employment offers stability and increased adaptive capacity, while in Argentina many have moved into tourism. In both cases, such adaptive efforts do little to address the deeper unsustainability of the agri-food system itself.

In Colombia, union entities do provide supports for coffee growers. Unfortunately, however, even the FNCC promotes the dominant discourse of sun-grown coffee monocultures as key to productivity. In the midst of this monoculture model, inequity of land distribution and international market fluctuations render peasant farmers particularly vulnerable to climate extremes that threaten their survival. Reflecting trends in the other two countries, Colombian youth choose off-farm employment due to the precarity of the dominant system.

Climate scenarios indicate increased risk of future extremes in each basin. In all three countries, farmers are exposed not only to climate extremes, but also to financial insecurity and uncertainty as a result of neoliberal economic trends, which then increases their sensitivity to climate events. Loss of producer supports and increased corporate market power exacerbate this exposure and contribute to reduced adaptive capacity.

In response to these structural factors, Colombian and Canadian farmers are pushed toward individualized and technocratic solutions like genetically modified seeds and chemicals to maintain their productivity. In Colombia and Argentina, climate adaptation measures are oriented to technological efficiency and market solutions (e.g., climate-smart agriculture), which depoliticize vulnerability and involve an excessive reliance on technological solutions to environmental damage caused by the economic model. Some authors have described this phenomenon as the “ecotechnocracy” of climate change, a bureaucratic apparatus in which socio-environmental problems produced in the framework of capitalist development have a technical-based solution, and positivist science becomes the only way to produce valid knowledge about climate change and adaptation (Sevilla-Guzmán, 2006). At the same time, ecotechnocracy promotes market solutions that accommodate globalization and associated patterns of rural development. Technology then is not only a proposal to tackle climate change, but the bastion of a development model.

In response to political-economic pressures, individual farmers may also pursue financial adaptation strategies. In Canada, the shift to specialized (sometimes patented) seed and chemical combinations and more sophisticated machinery has pushed some farmers further into debt. Large vertically integrated agricultural corporations seek to increase their profit both upstream (e.g., patented seeds) and downstream (processing, marketing and exporting) in the agri-food sector. As farms grow ever larger to adapt, larger farms and more expensive crops may not necessarily decrease vulnerability to climate risks.

This situation is similar in Argentina where, together with limitations in the availability and access to water, producers are affected by the pricing and marketing system imposed by the neoliberal model of agribusiness. The dominant model of rural development and agribusiness excludes those farmers who fail to integrate into the marketing chains. Ensuring water access will not be sufficient to modify vulnerabilities of less powerful producers, if it does not go hand-in-hand with policies that support small producers. In Colombia, farmers traded traditional polycultures for monocultures of genetically modified coffee varieties to increase productivity and resistance to pests and diseases. These cultures require more investment in fertilizers, but coffee prices at the New York Stock Exchange are not high enough to cover production costs —a similar issue to the “cost-price squeeze” reported in Canada. In both countries, international markets generate large profit to intermediaries, corporations, and foreign roasters.

The cases analyzed in this paper clearly reveal that vulnerabilities to climate change are produced under the framework of broad political-economic structures.

Global changes in agri-food systems exacerbate farmers’ vulnerabilities to climate change, especially in certain sectors or for some groups already experiencing inequality. This inability is manifested in unequal power over domestic market prices and trade dealings in general and makes the most vulnerable farmers fight for minimum subsistence levels. While in some cases, adaptation strategies challenge the dominant system (e.g., producers of shade-grown coffee in Colombia), in other cases, they may reinforce the very structures that produce vulnerability in the first place (e.g., Canadian farmers turning to costly biotechnological inputs to increase production; Argentinian producers abandoning agricultural activities under pressures from other sectors).

In this analysis, a Critical Realist approach helps us see commonality across contexts. In the midst of contextual specificity, CR encourages the search for “demi-regularities” (i.e., identifiable trends or patterns, but not causal laws) that can be found in complex “open” social systems (Danermark et al., 2002). Demi-regularities point toward pervasive structures that are relatively perseverant over time, although they may manifest differently depending on context and local conditions. Neoliberal political economic structures have led to similar patterns in each country, although institutional and individual responses may differ, and social differentiation means that some groups are affected differently than others. The structures we identify across contexts may ultimately be internally necessary to farmers’ current vulnerability (Danermark et al., 2002). In other words, in a completely different political-economic system, vulnerability might manifest differently or not at all. Moving beyond the proximal causes of vulnerability to the structural ones (Ribot, 2014) also helps us question the shortcomings of superficial solutions that focus only on the surface level of vulnerability (e.g., livelihood diversification, climate-smart solutions) and are not sufficient for addressing deeply rooted vulnerability.

Our findings suggest future actions that reduce the extent of corporatization or facilitate fairer international trade relations. Nonetheless, one caveat to our approach is that although such solutions help struggle against “deep vulnerability”, they are, by virtue of their depth, much more difficult to achieve. Thus, more short-term and context-specific solutions remain important but should be combined with structural critique. While the current analysis has relied primarily on qualitative data to identify relevant structures, future research in this area may seek to integrate quantitative data into structural critiques.

Drawing on the findings of an intercontinental study in Argentina, Canada, and Colombia, this paper has illustrated the usefulness of qualitative analysis for identifying deeply rooted causes of vulnerability amongst agricultural producers facing climate change. Interviews and group discussions with over 300 community members reveal that neoliberalism —as a deeply engrained political-economic structure that emphasizes trade liberalization, deregulation, corporatization, and a reduction in state supports for farmers— has been a major cause of vulnerability for agricultural producers in each country, exacerbating existing inequalities and reducing adaptive capacity to climate change. Identifying structural causes of vulnerability helps us move beyond individualized or technocratic solutions (for example, income diversification or expensive technological inputs that promise higher yields). In all three countries, over-reliance on such solutions has actually proved to be maladaptive for many farmers, in some cases pushing them further into the very structures that cause their vulnerability. Individualized responses are mostly reactive, and although they may work in some contexts or for some groups, deeper solutions are needed to address vulnerability at its very roots.

The authors declare no competing interests. Funding for this research was provided by the International Development Research Centre (IDRC) and the Tri-Councils, grant number 106372-008, 009.