Resilience of Rural Households: Insights from a Multidisciplinary Literature

Abstract

Due to the increase in both the frequency and severity of shocks and stressors that rural households are faced with, a vast array of studies has emerged in recent times on issues surrounding their resilience. Scholarship in this field has benefitted from insights from multiple disciplines including applied economics, disaster risk management, and human ecology, to name a few. These diverse contributions, although beneficial and necessary, have led to what appears to be discrepancies in the research arena. Using a bibliometric approach and thematic clustering, we shed light on the research landscape of rural household resilience. Furthermore, this article synthesizes some of the main contributions from various subfields and offers a systematic way for studies to approach the subject matter. From issues such as the objective and subjective dimensions of resilience characterization and measurement, to the degree of integration of the broader social ecological systems within which, households exist, the systematic classification of related concepts carried out in this article helps to organize knowledge, as well as aims to provide a guiding framework for future research and review efforts.

1. Introduction

Rural households are faced with various shocks that adversely affect their welfare and are, therefore, constantly negotiating ways to mitigate risks and respond to the changing conditions around them [1]. Such perturbations could range from environmental upheavals (e.g., floods or droughts) to economic turmoil (e.g., price fluctuations) or social disruptions (e.g., political conflicts or changes in government policy). It is important to note that these disturbances are not usually one-dimensional (e.g., solely environmental). They have various qualitative and time dimensions, and they elicit either ex ante or ex post responses from households [2].

In general, resilience is considered to be the capacity of individuals, groups, or systems to limit vulnerability to various shocks and stressors in either an anticipatory or reactive manner, without jeopardizing long-term viability [3,4,5]. However, directing the focus of resilience studies to humans and their livelihoods puts people at the center of the discussion and takes into consideration the capabilities and agency people possess [4].

It has since been noted that the household is often the appropriate unit of analysis in rural areas for crucial decisions such as consumption and labor supply [6]. Specifically with regard to social-ecological analysis, the household’s operations represent a more significant unit than the individual [7]. In addition, the household provides the relevant social, cultural, and historical context within which individuals are situated, as well as offers a suitable lens for examining a variety of motivations that influence the observed actions of individuals [6,8]. Therefore, from a rural household perspective, resilience can be viewed as comprising the various characteristics, actions, and strategies that positively moderate the relationship between shocks and development outcomes [9].

Due to the increase in both the frequency and severity of the shocks and stressors that households face, a vast array of studies has emerged on the issue of rural household resilience in recent times. Researchers from a wide range of fields including applied economics, disaster risk reduction/management, human ecology, and human geography, among others, have embraced the task of investigating this important subject. However, this increased interest and scrutiny from practitioners and scholars has resulted in a lack of consensus in its definition and measurement, among other aspects, as well as in our ability to synthesize findings and recommendations [3,10].

Sometimes, the extent of the divergence noticed in the literature makes one wonder if it is the same concept being studied or whether it is just a case of different dimensions of resilience being examined [11], or even more, if resilience is so remarkably contextual that it leaves little to no room for the cross-comparability of studies. This paper is an attempt to address some of these concerns and to offer a systematic approach to further carry out research on the subject matter.

Aims of This Study

With variations in resilience conceptualizations, as well as in the resulting approaches to investigating how rural households can build resilience [3,12], it is important to synthesize existing knowledge in this subfield to better direct future studies. In the words of Elinor Ostrom,

“Without a common taxonomy of core variables, research conducted by scholars from multiple disciplines tends to focus on variables of major interest to their own disciplines without recording, measuring, controlling for, or even thinking of other variables that might account for the patterns of interactions and outcomes observed”

[13] (p. 15186).

The study of rural household resilience benefits from insights from multiple disciplines. However, there has been limited scrutiny on the issue from a more holistic point of view. This study aims to fill that gap and highlights the complementary nature of research from different fields. The present study does not attempt to take on the ambitious goal of laying out all the relevant pieces to consider in an inquiry of rural household resilience. Additionally, the goal of this paper is not to carry out a quantitative systematic review of the literature. It has been pointed out that such a task would be of limited utility due to the significantly non-representative nature of wellbeing indicators, shocks, and study regions about the topic found in the literature [12].

Rather, this paper conducts a more general knowledge synthesis and offers a systematic classification of existing knowledge on the subject from several subfields. In this regard, a major objective of this paper is to produce a typology that captures different elements of these research strands and offers a framework that can be applied across various settings. Such a task is necessary since resilience has been pointed out to be an arena for generating integrative science and interdisciplinary collaboration in promoting a more sustainable development for humanity [14]. Furthermore, such a typology can be useful in guiding subsequent (systematic) review efforts in the field of rural household resilience, as well as help build a pipeline of studies that produce evidence that can be replicated across a range of settings.

2. Methodology

This study begins by providing a broad overview of knowledge on rural household resilience from different disciplines. Several reviews along different dimensions of resilience relating to rural households have already been carried out in the literature. For example, Serfilippi and Ramnath [3] provide an overview of competing resilience definitions used by researchers and development organizations. Additionally, they compare resilience with vulnerability and sustainability; evaluate some of the common frameworks employed in the analysis of resilience; and present a list of indicators that could help in its measurement. Douxchamps et al. [10] conduct a similar review and focus more on the existing resilience assessment tools that have been created for the context of climate change and agricultural development. On the other hand, Ansah et al. [11] provide a review of resilience concepts, methodologies, and empirical evidence strictly from a food security perspective.

More recently, and more closely related with some of the goals of this present work, Barrett et al. [12] carry out a scoping review and examine how resilience has been conceptualized in the international development literature. They probe the accompanying analyses that follow such theoretical underpinnings, as well as the purposes, findings, limitations, and implications of empirical studies on the subject matter. They consider both the individual and household level studies and focus on low- and middle-income countries.

The review carried out in this study is distinct in some of its objectives and approach. It begins by conducting a bibliometric analysis to intentionally uncover the breadth and nature of disciplinary contributions to the study of rural household resilience. Bibliometric analyses are often employed to analyze the pattern and context of knowledge production in a subject area [15]. They are also helpful in identifying subdomains of research fields. The keywords used in the bibliometric search were informed by the literature and additional suggestions were received from other researchers in the field. Search strings included phrases like {resilien* (The wildcard symbol (*) represents zero or more non-space characters. This ensures that all words beginning with the suggested phrase are captured) capacit*}, {livelihood* resilien*}, {resilient household*}, {development resilien*} {livelihood* adaptat*}, {socio-ecological/socio-environmental resilien*}, {socio-ecological/socio-environmental adapt*}, etc. The systematic search was conducted in Scopus database between January and February 2021 and the resulting analysis was carried out using Microsoft Excel and VOSViewer 1.6.11. (Additional details of the bibliometric analysis and output are available from the authors upon request).

Following the bibliometric review, a qualitative thematic clustering was carried out. The goal was to provide a general overview of the field of rural household resilience. For this, results from the bibliometric analysis were complemented with a “snowball” reference selection. That is, other relevant articles not captured by the systematic search (e.g., due to different keywords provided by the authors) were also incorporated into the document reviewed for this study. For instance, articles that lend insights into the theoretical/conceptual underpinnings of a research strand were included in the discussions, even if they did not constitute a part of the database for the bibliometric analysis.

The themes used for the clustering were informed by elements of resilience studies found in the literature. Specifically, studies were broadly grouped based on the major outcomes of interest in those studies and by how resilience is both conceptualized and operationalized. The paper then culminates by offering a typology and discussing various issues that future studies on rural household resilience should carefully consider. The proposed typology is informed by practical considerations involved in conducting a study on the resilience of rural households, such as deliberations surrounding how resilience is framed and made operational for a study.

3. Overview of the Literature

The research landscape on resilience of rural households is vibrant and has been experiencing exponential growth over the past decade. The dates of publications spanned about three decades, with the earliest work published in 1992. However, among all the articles used in this study, only three were published in the 90’s. Over 95 percent of the articles were published from 2010 onwards, signaling the increased appeal of household resilience to researchers.

About 3000 authors contributed to the documents used in this analysis. The individually most cited authors were Berkes F., Twyman C. and Stringer L. C., with 615, 419 and 306 citations, respectively. The most cited work in the database was by Berkes and Jolly (2002), titled: “Adapting to climate change: Social-ecological resilience in a Canadian western arctic community”. There were over 70 journals with at least three publications, and with over 2000 citations and 38 articles, Ecology and Society was top-ranked (

Table 1. Most impactful journals.

Journal	Total Link Strength	No. of Publications	Citations
Global Environmental Change	57	25	1476
World Development	56	18	289
Ecology and Society	40	38	2101
International Journal of Disaster Risk	32	25	299
Land Use Policy	28	21	254
Regional Environmental Change	26	21	378
Geoforum	24	6	282
Environmental Science and Policy	22	9	309
Sustainability	21	31	141
Food Security	18	8	89

). However, Global Environmental Change and World Development had the greatest link strengths (i.e., in terms of cross-citations from other journals) with 57 and 56, respectively (Figure 1).

Resilience, climate change, adaptation, vulnerability, adaptive capacity, livelihoods and food security were found to be the most commonly used author keywords in the database used for the bibliometric analysis. Figure 2 below displays the co-occurrence network of author keywords. The figure reveals some of the common themes that have received greater attention in the field ranging from specific livelihood practices/strategies (e.g., agroforestry, migration) to countries/regions (e.g., Bangladesh, Africa) and shocks/stressors (e.g., climate change, floods).

These results from the bibliometric analysis help to highlight the multidisciplinary nature of research on rural household resilience. This article then proceeds by providing a brief overview of the literature to shed light on some of the major insights and theories that are informing and being produced by research studies. For this paper, academic work on the subject matter is broadly grouped into two research strands: the Human Systems Research Strand and the Social-Ecological Systems Research Strand, and, where applicable, corresponding sub-themes under a strand are referred to as Clusters.

It should, however, be noted that the groupings delineated in this overview are not necessarily mutually exclusive, as it is common for studies to incorporate elements of the various research classes—albeit to different degrees. The goal of such a classification is simply to illuminate different aspects researchers take into consideration when investigating rural household resilience, and to provide a framework that highlights elements of the different approaches. These groupings then provide a springboard for which the typology emerges. The article proceeds by briefly describing these broad approaches:

3.1. Human Systems Research Strand

3.1.1. Cluster 1

In general, this research cluster considers resilience as a set of conditions or attributes that enable households to achieve resilience and attempts to quantitatively measure the concept for research purposes, as well as for the monitoring and evaluation of development projects [16]. Resilience is often linked to outcomes such as food and nutrition security, asset accumulation, health status, and other observable measures of household wellbeing.

A major line of scholarship within this research cluster adopts the concept of development resilience, defined as “the capacity over time of a person, household or other aggregate unit to avoid poverty in the face of various stressors and in the wake of myriad shocks” [17] (p. 14626). Resilient units are those that maintain this high capacity over time. In many instances, an asset-based approach is often favored over income- or expenditure-based poverty measures in this cluster [18]. Assets are broadly considered to be “state/stock variables used to generate income” [19], and they include “productive and financial wealth, as well as social, geographic and market access positions that confer economic advantage” [18] (p. 179).

The theoretical underpinning for this research theme is that of stochastic welfare dynamics or poverty traps/dynamics, with insights from the literatures on risk and vulnerability [20]. It is argued that in light of initial ability and asset levels, poverty can become self-reinforcing due to structural and behavioral mechanisms such as market failures, social exclusion, underinvestment in higher-return assets/inefficient investments, etc., and these cause low living standards to persist [21].

Wellbeing outcomes—which are subject to random shocks but determined by choices that are constrained by nature, human institutions, and resource availability—fall within certain ranges (or stable states) that are not expected to change over time without interventions [17]. Since the concept of development is normative by nature, it is argued that resilience in this context implies preventing a movement to lower undesirable states for those that are currently nonpoor, as well as enabling a transition to better wellbeing regimes for the current poor [17]. In this subdomain, resilience has also been interpreted as the (non)persistence of shocks, i.e., how long a household experiences the effects of a shock [22].

Some other bodies of work are also prominent in this subfield, such as that of the Resilience Analysis and Policies team at the Food and Agricultural Organization (FAO) through the introduction of the Resilience Index Measurement and Analysis (RIMA) tool (and subsequently, RIMA-II). Consistent with its organizational mandate, the FAO RIMA approach directs its focus to the food systems, and specifically to the food security status of the household [23]. This interest on household food security has been the focus of several studies [24,25,26]. However, other scholars have adopted the approach and used it in other non-food security contexts [27].

The RIMA-II and other similar frameworks [16] attempt to capture the possible pathways through which households may be able to moderate the impact of a shock on their wellbeing. Pillars of resilience are constructed using time-variant and time-invariant characteristics and activities. In general, the pillars of resilience are taken to be absorptive, adaptive, and transformative capacities [28], and indicators that reflect these capacities are then chosen. However, RIMA-II uses a practical classification based on its food security outcome of interest, thus resulting to unique resilience pillars such as access to basic services, assets, social safety nets, sensitivity, and adaptive capacity [23].

3.1.2. Cluster 2

Within the human systems research strand, a theme around the subjective dimension of resilience has emerged over time. Scholars in this field argue that there is often an assumption of a consensus on “desired states” [4]—which are usually defined by the prevailing power dynamics—or that a desired state even exists, and such thinking is sometimes viewed as “akin to ‘imposed rationality’ that is alien to the practice of ordinary people” [29] (p. 109).

Accordingly, calls have been made to complement standard objective indices with subjective measures in resilience studies, as it is believed that people can self-evaluate their capacities, capabilities, and limits [30]. Likewise, the existing variation in resilience conceptualizations and the value-laden nature of some of its constituent objective indicators have been argued to give further grounds to the potential of subjective approaches [30]. Furthermore, it is stated that subjective measures may prove to be better at capturing less-tangible features of resilience such as entitlement and power [31], or even courage, optimism, and faith [32].

For instance, a subjective approach to resilience highlights people’s differences in perceiving risks, which could shape their resulting anticipatory actions [4]. Risk attitudes—which comprises of the “perceptions about the probability and severity of risk associated with change as well as the costs and benefits associated with adapting” [33] (p. 54)—are considered to be subjective and socially constructed, and impose a limit to adaptation actions of decision-makers [34]. Similarly, perceptions of the riskiness of possible responses following the event of a shock are also believed to shape resulting behavior [35].

Additionally, people’s perception of their ability to handle hazards has been shown to influence their adaptation responses to adverse events [36]. For example, it was found that the decisions of some households in Bangladesh to remain or leave their community in the face of environmental disasters was partly due to their self-confidence and belief in their ability to bounce back [37]. It is posited that some other factors such as strong occupational identity or place attachment can as well limit people’s ability to re-imagine themselves in other roles [33]. This could prevent them from exploring alternative livelihood strategies, thereby undermining their overall resilience.

In essence, this research cluster has emanated to help capture aspects of resilience that objective indicators may miss or inadequately account for. Studies in this cluster often highlight the role of cultural and societal norms and values in shaping the response of individuals and groups, and ultimately, their resilience.

3.2. Social-Ecological Systems Research Strand

Studies in this research strand are characterized by their presentation of the intricately interconnected nature of human and natural systems. Social-ecological systems (SES) are complex and comprise multiple subsystems and internal variables that interact to produce outcomes that feed right back into these (and other) subsystems and their components [38]. In addition, these systems are adaptive, i.e., they tend to evolve in their “genetics, behaviors, or spatial distributions” based on the outcomes of the interactions that occur within them [14].

It has been pointed out that adaptive systems are process-dependent and organic, often with multiple stable states (or regimes); while regime shifts—changes to alternative states with different structures and functions—are the consequences of the erosion of resilience being often caused by human actions [14,39]. Studies adopting the SES approach attempt to illuminate the manner in which human activities alter and are affected by the structure and function of ecosystems [33]. Some of these feedbacks and non-linear dynamics can result in the deepening of undesirable states, thereby leading to a social-ecological trap i.e., a situation whereby poverty and environmental degradation are perpetuated by reinforcing social-ecological dynamics [40,41].

Furthermore, an SES approach often helps in pinpointing the identity of the system in question and measuring thresholds of its elements. The elements of such an identity include: Components—the human and non-human actors; relationships—the interaction among the components; innovation—the elements of the system that lead to change; and continuity—the elements that embody memory and help maintain the system [42,43]. These identity elements do not necessarily represent all the variables influencing the system but taking them out of the system would lead it to a qualitatively different state [43].

It has been noted that the delineation between social and ecological systems is often arbitrary and artificial, since humans exist in nature [44]. Therefore, a lens that is solely focused on the adaptive abilities of the social system may not be sufficient to guarantee desired sustainability outcomes. It has also been observed that the actions of decisionmakers taken in response to change and uncertainty could well undermine the ability of the ecosystem to sustain such adaptation strategies [14]. Thus, studies in the SES research strand emphasize the linkages across subsystems, making them better positioned to reveal the long-term implications of shocks and resulting responses on household resilience.

The SES perspective helps to capture how changes in environmental variables relate to changes in the human system dynamics. The focal point here is often about sustaining the productivity and functioning of natural resource systems, as well as household wellbeing and social institutions. Such inquiries could range from examining the complex interactions between the economic decision-making of farming households and soil fertility in the face of environmental shocks [45], to assessing the potential for the livelihood transformation of households and the prevention of social breakdown in the presence of climate change [46].

The SES research strand has also played a vital role in expanding our understanding of transformation in the resilience discourse. The intensity of shocks and stressors could be strong enough to warrant an alteration in the qualitative state of the system through the introduction and/or loss of state variables. Research in this field illuminates the changes in the structure and function of a system that can occur, and the factors that can prevent or facilitate such changes—e.g., when transformation is desired and deliberately pursued [47]. Such insights can be beneficial within the rural household context in that the conditions that create for longer lasting resilience can be identified. For instance, a household may need to explore an entirely new livelihood direction (as opposed to say, trying new farming practices) and understanding the requisite behavioral shifts, technological innovations and institutional reforms to achieve this will be paramount [28,47].

In a nutshell, we have employed the broad groupings of the Human Systems and the SES Research Strands to collate different aspects of scholarship on rural household resilience. The overview of the literature that has been discussed so far provides the setting for the rest of this paper. Following the bibliometric analysis and thematic clustering, we develop a systematic mapping of knowledge on the subject matter. In the section that follows, we propose a typology that is informed by various elements found in the research strands presented above.

4. A Typology of Studies on Rural Household Resilience

Since the landscape of research on the resilience of rural households has agreeably been referred to as being in an “early, somewhat noisy and disjointed stage” [12] (p. 19), it therefore becomes instructive to construct a typology for this subfield of resilience. Typologies help to organize knowledge in a subject area by systematically classifying related concepts and help to guide further research [48]. This evolving knowledge domain of rural household resilience, with significant contributions from a wide range of disciplines, is ripe for an attempt at integrating some of the diverse approaches presently found in the literature.

Below, a typology is proposed (Figure 3), which will hopefully facilitate stronger multidisciplinary efforts geared towards a more robust understanding of the resilience of rural households. This mapping is not a mere extension of the classification discussed in the previous section. Rather, it is the outcome of dissecting the research strands and clusters into integral parts and grouping them under related strata—such as is done for a taxonomical classification. The typology is also informed by some of the practical considerations that go into conducting a study on rural household resilience. The intent of such a system of categorization is to assist researchers in recognizing the options available at different stages of the research endeavor, as well as to enable them to clearly disclose the choices they are making—choices which are hopefully based on the best fit or purpose of the study.

As advised in the literature, the place to start when attempting to investigate the resilience of rural households is to address the question: “Resilience of what, to what, and for whom?” [49]. As the focus of this treatise is rural households, the researcher may need to be explicit about the shocks and/or stressors that threaten the outcome of interest—and these hazards may be covariate (widely experienced) and/or idiosyncratic (individualized) in nature [9,50]. Similarly, there is need to recognize the economic, political, sociocultural, and ecological contexts within which the households are situated, and to consider how these enable or limit their resilience. It is also imperative to be clear and consistent about how resilience is being conceptualized and subsequently operationalized. The sub-sections that follow revolve around these issues.

4.1. Resilience Conceptualization

A few dominant approaches to conceptualizing resilience are currently found in the literature. They are broadly classified by Barrett et al. [12] as: resilience as capacity; resilience as a normative condition; and resilience as a return to equilibrium.

The resilience as capacity (or combination of capacities) approach considers resilience to be a multidimensional set of features that enable a household to limit the adverse wellbeing effects of shocks and stressors. As pointed out earlier, the constituent pillars of resilience in this approach are usually taken to be absorptive, adaptive, and transformative capacities—or any other combination of similar capacities (e.g., in the FAO RIMA framework). This approach endeavors to incorporate a wide range of indicators from social and ecological systems that limit households’ vulnerabilities or enhance rapid recovery. Here, resilience is seen as an ability to withstand, recover from, or alter the structure and function of a system in the presence of shocks in a way that promotes longer-term resilience.

On the other hand, the other two categories in the Barrett et al. grouping take resilience to be a condition, or a qualitative state, that is either normatively determined or defined in reference to a pre-existing state. The approach that treats resilience as a normative condition anchors the concept to wellbeing standards that have been normatively defined. That is, thresholds for a standard of living indicator (e.g., minimum levels of food consumption or asset stock) are used as a reference point in assessing whether households are resilient or not.

Conversely, the resilience as a return to equilibrium approach focuses on a household’s recovery from a shock to its ex-ante state. That is, it considers the success of households in bouncing back to their pre-shock status. In principle, these two classifications could be viewed as variants of a resilience as a condition (or resilience as an outcome) approach. The main difference between the two sub-approaches is a consideration of whether the initial state of the household was desirable in the first place.

From the ongoing discussion, it appears as though fundamentally, the approaches to seeing resilience as either a capacity or a condition are directly linked, in that the condition of resilience is simply a reflection of an underlying capacity (or capacities). It can be argued that the magnitude of loss in the event of a shock and the speed and extent of recovery are in essence connected to the capacities of the household. As Christopher Béné [51] puts it: “resilience capacities are input to the resilience process, while resilience per se is the (intermediary) outcome, contributing to the longer-term final outcome (which itself is measured in terms of wellbeing)” (p. 811).

In this vein, the degree to which the distinctions in resilience conceptualization matter may be limited to how they inform the resulting operationalization of resilience in studies. Therefore, a researcher’s interest in either assessing the multiple capacities that enable rural households to be resilient, or in assessing the likelihood of those households (re-)achieving some established standard of living will ultimately determine the analytical approach to be adopted.

4.2. Resilience Operationalization

In addition, the researcher will need to make decisions as to what extent the views or perceptions of the population being assessed ought to be reflected in the studies. That is, judgements are needed about whether, and to what degree, resilience ought to be defined and/or evaluated by experts and external agents (objective approach), or by the subjects themselves (subjective approach). In addressing this issue, Jones [52] proposed that the objective and subjective measurement approaches lie on a continuum and stated that “when it comes to resilience measurement, it is crucial to recognize that subjectivity and objectivity are neither binary nor mutually exclusive” (p. 4).

This view suggests that the two methods may be seen as being complementary to one another and evaluators may best be served if consideration is given to the two approaches. That is, it will be wise for researchers embarking on an investigative journey on the resilience of rural households to be armed with a toolkit of both objective and subjective measurements. However, final decisions should ultimately be informed by the suitability of the preferred approach to help them arrive at the goal of their inquiry. Nevertheless, such a conclusion may only be arrived at after adequately considering the pros and cons of each approach.

Some of the factors to consider in making this decision include the bounded rationality of a single set of stakeholders; the practicability of minimizing the various cognitive and response biases such as priming and extreme response bias (as in for subjective approaches); the desire for comparisons across households or studies (e.g., the impact of interpersonal or cross-cultural differences in standardized questions); the time and resource constraint of the study (e.g., the amount and types of data required for objective approaches); the ability to identify indicators that accurately reflect the vulnerability contexts and capabilities of households; etc.

When employing subjective approaches, it is important to remember that people’s interpretations of what resilience is to them or how they assess resilience may be dependent on several factors including cultural and personal elements. However, there are tools and techniques that can be utilized to reduce the worries associated with how people subjectively assess themselves and their experiences [53]. For instance, anchoring vignettes have been shown to be helpful in measuring complicated concepts and in correcting for differing interpretations of identical questions [54]. In illustrating how this tool can be applied to a resilience context, Jones [52] discusses how evaluators may begin by describing a hypothetical person’s experience with a shock and then asking the respondents to rate the made-up scenario before rating their own situation using the same response scale. This strategy provides a basis for benchmarking individual responses, as well as helps in controlling for cross-cultural differences.

Furthermore, the choice of indicators and analytical method (if applicable to the study) should also be carefully determined since measurement decisions have sometimes been shown to influence the conclusions on wellbeing outcomes. For instance, working with the same longitudinal survey data, Vaitla et al. [53] discovered that conclusions about food security status were sensitive to the cut-off points used by the researchers. That is, the threshold that the researchers used to specify whether a household was food secure or not influenced the conclusions from the analysis. In addition, the study showed that drivers of resilience differed depending on the choice of food security measure (i.e., the suggested causes of food insecurity when researchers used the Food Consumption Score measure were different from when the reduced Coping Strategy Index was utilized). The authors concluded that the two measures employed in the study capture “overlapping but not identical aspects of food security” [53] (p. 147).

In a different study, Upton et al. [54] compared the performance of three of the dominant resilience analytical methods using the same data. They found that each method marked different households as resilient or least resilient, and differences were also observed in the predictive abilities of households’ wellbeing states over time. The authors concluded by stating, “it remains unclear what these [methods] really measure nor what descriptive, inferential, or predictive benefits they yield [beyond more established wellbeing measures] …The approaches presently in play are all, at best, imperfect, and at worst deeply flawed” [54] (pp. 13–14).

Sadly, as attested to in the foregoing discussions, the observed sensitivities to the indicators and methodologies employed in resilience analyses reduces the confidence we have in the findings from studies on rural household resilience. However, heeding certain guidelines can help bring researchers closer to carrying out investigations that produce more reliable conclusions. One of such pertinent recommendations include utilizing higher frequency data for empirical studies—in contrast to the overwhelming use of cross-sectional data presently found in the literature (e.g., Barrett et al. [12] find that only 16% of the quantitative studies in their review used panel data).

In addition, researchers should employ approaches that are best suited for the question at hand. For instance, a composite index-based analytical approach may not be the most ideal empirical method if the goal of the study is to identify what variable—e.g., a particular household asset or community infrastructure—will be most impactful in building resilience [54]. Finally, it is crucial for researchers to be aware of and transparent about the assumptions and limitations of their chosen analytical approaches.

4.3. Systems Perspective

Another point of importance when carrying out a study on rural household resilience is the consideration of the relevant subsystems required to satisfactorily describe or assess resilience. In addition to the characteristics and assets of households, some studies appropriately consider the various social, economic, and institutional structures and processes that enhance resilience. However, the wellbeing of rural households and the long-term viability of their livelihoods cannot be addressed without an appreciation of the intertwined social-ecological systems within which these households exist—especially if their livelihoods are directly supported by these systems. To what extent the investigator explicitly factors this interconnectedness into the inquiry will depend on the objective of the study.

Even if the goal of the study is not specifically to describe or analyze the relationships between these interlinked systems, it will be valuable for researchers to account for them in some way. For instance, a study that seeks to identify which households are more resilient to certain shocks, or one that aims to describe the impact of a hazardous event on specific wellbeing outcomes, may not need to detail the feedback loop between human and natural systems. Even if the shocks being considered are not weather-related (e.g., sociopolitical shocks like conflict), it may still be helpful to point out how ecosystem properties influence households’ wellbeing, especially if their productive operations are somewhat dependent on natural capital.

On the other hand, if the study at hand is interested in the long-term sustainability of nature-based livelihood activities in which households in a given region participate, then it will be crucial to acknowledge the linkages between various human and natural subsystems and recognize that there will be feedbacks across multiple scales and time periods. Because of the sheer number of relevant variables in both the biosphere and social system and all the interplays among them, and because these systems are adaptive in nature—so much so that one cannot fully predict how they will behave in future—it may be practically impossible to outline all the interactions that take place within these systems. Nevertheless, it will still be valuable to identify some of the most important system attributes in an investigative effort on rural household resilience.

Quinlan et al. [55] suggest the possibility of simplifying our understanding of social-ecological systems resilience, while simultaneously retaining the richness of a systems perspective. As an example, using a meta-framing of resilience principles, they propose that consideration could be given to the system structure and/or the system dynamics. The structure of the system deals with the organization or the presence of links between components of the system, while system dynamics has to do with the complex interactions of (fast) variables within the system [55,56]. Identification of such important system elements and connectivity does not only contribute to the richness of resilience thinking from a household perspective but will ultimately help to guide their decision-making processes.

In another instance, Ruiz-Ballesteros and Ramos-Ballesteros [7] (2019) link the practices of households in the Agua Blanca community of Ecuador with the evolution and current state of their socioecological system. The authors consider how factors such as demographic behavior, household economy, household participation in the community, local knowledge, place attachment, and future outlook influence the dynamics and resilience of the SES. These examples illustrate how the need to streamline systems perspective into a research study could possibly be achieved.

4.4. Other Considerations

A consideration of multiple time periods in resilience enquiries is essential in resilience analyses since resilience is a time- and event-dependent concept. Some authors have even hinted at the need for high frequency data for resilience analyses, such as statistics that are collected monthly [22]. However, collecting quality and extensive time-series data can be very expensive on a practical basis. Notwithstanding, by the very nature of resilience as a concept, at least two points in time (i.e., ex ante and ex post) should guide the investigation. It is almost impossible to think of resilience without reference to different time periods, even if it is just a recollection of a household’s asset level before and after a shock, or a projection of its wellbeing state into the future.

On a different note, when thinking of capitals, researchers should avoid the trap of limiting the indicators they choose to just the tangible assets available to households. It is important to incorporate the notion of access to various resources, as well as to see capitals as the basis of agents’ power to act, adapt and transform their situation [57]. As Bebbington [57] notes, “People’s assets are not merely means through which they make a living: they also give meaning to the person’s world… This meaning will then be one of several influences in subsequent decisions people make about their livelihood strategies” (p. 2022). Such meanings, as well as people’s values and ideologies (or cultural capital in general) have been shown to shape governance mechanisms or institutions [4,58], which in turn constrain or enhance resilience efforts [59,60].

Similarly, thought should be given to the various risk- and loss-management strategies employed by rural households (i.e., strategies employed to reduce ex-ante possibility of a loss or to mitigate ex-post consequences). Because of the missing and imperfect markets in many rural areas, households often tend to have a diversified portfolio of livelihood activities, as well as engage in consumption/asset smoothing [2,61]. The specific alternative mechanisms that households adopt are often shaped by the economic, political, and sociocultural environment in which they find themselves, which ought to be given sufficient attention by researchers as well. The level of importance of these contextual variables has made some authors advise that even if the focus is the household wellbeing, development interventions should not be targeted at the household level alone, but should be directed at institutions, infrastructure, and higher system levels as well [37].

5. Conclusions

Scholars from multiple disciplines have been actively engaged in increasing our understanding of resilience as it relates to rural households, whether it is by examining the effects of stressors and shocks on household wellbeing, or assessing the ways households respond to these threats, or evaluating the outcomes of these responses. Such diverse contributions, although beneficial and necessary, could sometimes lead to some confusion in the research arena. Additionally, academic work on the issue is usually confined to disciplinary boundaries, resulting in variations in how resilience is framed and analyzed, which has in turn often led to contrasting results in resilience analyses. Such discrepancies have warranted a more integrative approach in studying rural household resilience.

This article synthesizes some of the major elements embedded in various research strands in the literature and offers a systematic way for studies to approach the subject matter. This treatise aims to encourage scholars to employ a multidisciplinary lens in their research endeavors, a move that will hopefully result in a more robust understanding of resilience within the rural household context. The typology presented in this paper allows for different theories to inform studies on rural household resilience. This is especially useful because it has been argued that it may not be desirable to have a unified theory of resilience in the field of development since it is doubtful that such conformity will meet the practical demands of various stakeholders [12].

As a result, scholars investigating the resilience of rural households can adopt the characterization of resilience that is most useful to the goals of their enquiry, whether their focal interest is the state of resilience or the capacities that make this possible. It does not currently appear that there is a one-size-fits-all approach to resilience conceptualization and measurement. Nevertheless, the selected resilience measures for a study should be consistent with the preferred conceptualization. Furthermore, as sometimes pointed out in the literature but not always heeded, resilience within the context of rural household is not an end goal by itself. Specific wellbeing indicators ought to be the final outcomes of interest—such as the food security, health, or livelihoods of households—and these should be clearly indicated in studies.

Researchers should also endeavor to integrate elements of objective and subjective approaches in their characterization and/or evaluation of resilience in a manner that advances the objective of the study. This fusion of approaches is needed so that studies can maintain scientific objectivity, while adequately reflecting the experiences of rural households (since resilience ultimately centers around households’ qualitative states). In addition, the types of data and the analytical methods that are best suited for the research questions at hand should be the guiding factor for researchers, as opposed to convenience, funding source, existing structure of collaborative networks, or methodological allegiance.

Finally, the need for systematically synthesizing findings from studies on rural household resilience is becoming ripe. Scholars who seek to embark on this worthy task can find a framework such as the one presented in this paper to be a helpful tool. Researchers may choose to use certain elements of the typology in this paper as the basis for collating previous studies. Furthermore, this framework can also inform the generation of a stream of comparable future studies in this subfield—studies that integrate insights and approaches from multiple disciplines—thus fast tracking our cumulative scientific understanding of rural household resilience.