Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Introduction
The household is a key unit of data collection and analysis in multidisciplinary, and often interdisciplinary, environmental social science research. Several disciplines and fields under this broad umbrella often rely on collecting and analyzing data from households to understand topics such as climate change impacts and adaptation (e.g., Agrawal et al., 2009; Hertel et al., 2010), agricultural productivity and food security (e.g., Jones et al., 2014), food safety (e.g., Pretari et al., 2019; Bauchet et al., 2021), land use (e.g., Bluwstein et al., 2018; Keane et al., 2020), anti-poverty interventions (e.g., Banerjee et al., 2015), and environmental conservation and sustainable development (Bennett et al., 2017; Kareiva & Marvier, 2012; Nguru et al., 2020; Silva & Topf, 2020).
Scholars have long discussed the complexity and changes of household composition as it relates to social-environmental processes. Families and household are frequently discussed in terms of urbanization (Burch, 1967; Buzar et al., 2005; Gore, 1990), differences in composition across space and time (Bongaarts, 2001; Ruggles & Heggeness, 2008), and shifts in partnerships, fertility, and mobility (Pasley & Petren, 2015; Smock & Schwartz, 2020). Such complexity and changes of household composition make it challenging to use household as a unit of study, and the related limitations have been long established (Deaton, 1997; Casey and Lury, 1981; Wilk & Netting, 1984). Nevertheless, definitions commonly used in the academic literature continue to lack the flexibility needed to capture diverse and nuanced household structures (Casimir & Tobi, 2011; Niehof, 2011).
Advertisement
Outside of academia, the United Nations (UN) World Population Census defined a household as “persons living together who make common provision for food and other essentials for living” (United Nations, 2008, p. 128). This definition is widely used by large international organizations such as the Food and Agriculture Organization. Grosh and Glewwe (2000), in a guide to designing household surveys, presented a slightly different definition: “a household is a group of people who live together, pool their money, and eat at least one meal together each day” (p. 135). Nationally distributed censuses often have their own definitions as well. Furthermore, alternatives to the traditional household concept such as the kinship support tree (Madhavan et al., 2017), modified-extended-household (Gartaula, 2013), and dual and divided households (Owusu, 2001), have been used to broaden the conceptualization of households; however, none has been widely adopted in the field of environmental social science. Indeed, a review of 59 academic publications using household data found that none provided a theoretical or operational definition of a household (Casimir & Tobi, 2011).
The variation and lack of clarity in household descriptions has important consequences in research and research-informed practices, potentially leading to undercounting, bias, and misrepresentations (Randall et al., 2011). Notably, using different household definitions has been shown to affect how household assets, consumption, and production are measured (Beaman & Dillon, 2012), or how migrants are included or excluded from household membership (Ünalan, 2005). People with unstable housing, housed in long-term care, or who are nomadic are also often left out of traditional household units (Carr-Hall, 2015). Yet, very few studies, including in the field environmental social science, provide clear and explicit information about how they conceptualized households.
In response, we use cases from Pakistan, Bolivia, Peru, and Senegal to illustrate the complexity of household structures, and to argue that the current lack of clarity in household definitions in the field of environmental social science leads to implicit assumptions about household structures and living arrangements. We then highlight the need to establish a clear, yet nuanced process to describe households and their structures. Our findings inform the development of a protocol for explicitly conceptualizing and describing household structures. We view this protocol as a first step toward a shared process (rather than a fixed, singular shared definition) that researchers can adopt to identify, describe, and potentially compare household structures, based on households’ composition, resource sharing, and distribution and movement across time and space. Our proposed protocol helps researchers to specify assumptions embedded in their definition of households and make explicit choices about how they describe a household. It simultaneously captures the variations present across study contexts while facilitating standardized communication of household characteristics that allows for comparison across studies. As such, this protocol can help improve the credibility and transparency of data collection.
Methods
We used a case study methodology to describe and illustrate the complexity of households across four study sites. A case study is frequently used to assess how a phenomenon occurs within a specific context (Yin, 2018); thus, this approach is particularly well-suited to investigate household characteristics as they are embedded in specific political, cultural, and environmental contexts. We employed Seawright and Gerring (2008)’s diverse case selection method to illustrate how a variation in household characteristics can influence household data collection and understanding of resource ownership and use. We acknowledge that our four cases are located in the Global South with a focus on low-income rural households, and are not representative of all households in all contexts. They do, however, allow us to illustrate the importance of explicitly conceptualizing and describing household compositions and structures in collecting accurate household data, particularly related to resource ownership and use.
Advertisement
Co-authors of this Research Brief led the cases under analysis. While methodological and conceptual approaches varied, ranging from those grounded in feminist theory, to mixed-methods, to a positivist experimental design, each case investigated common behaviors of resource management and use among households in rural communities. Household compositions and structures were not the focus of these studies but emerged as a common theme during data collection and analyses. As such, these cases were selected to be included in this Research Brief. Once case selection concluded, we applied deductive reasoning and cross-case analysis (Yin, 2018) to analyze household compositions and structures and examine how they diverged from commonly used household definitions. Building upon assessments of existing household definitions (Beaman & Dillon, 2012; Casimir & Tobi, 2011; Randall et al., 2011), our cross-case analysis incorporates an environmental social science perspective and informs the development of a protocol that can be used to support clarification and operationalization of household definitions in the field of environmental social science.
Case studies of varying household compositions and structures in environmental social science
Joint-family structures, headship, and resource ownership in Pakistan
We draw from research conducted in the Khyber Pakhtunkhwa (KPK) province of northwestern Pakistan to illustrate how commonly used household definitions do not address the varying structures and related resource use in multi-generational households (Nixon, 2022a, b). Multi-generational households can be stem (i.e., one or both partners, one of their children, and their child’s partner and children) or joint families (i.e., one or both partners, two or more of their children, and their children’s partners and children) (Ruggles, 2010). Differentiating between joint and stem households is uncommon, including in the UN’s definition of multi-generational households (United Nations, 2019a), despite the prevalence of both structures in many parts of the world (Ruggles & Heggeness, 2008). In fact, neither the national census nor the international data sources we examined have collected data on joint families in Pakistan. We argue that this limits our ability to accurately report shared ownership and decision-making processes present in joint families.
We surveyed 448 households in KPK in 2019. We specifically asked about household structures and defined joint families as multiple related nuclear families living together in a house or compound (multiple houses joined physically or in the same enclosed area) (Suleman et al., 2012). Fifty-eight percent of households surveyed reported a joint-family structure, which usually consisted of adult brothers and their families living together with their parents (Nixon et al., 2023). Our data indicated that this joint-family structure impacts households’ access to, ownership of, and management of various forms of capital; topics of concern for many in environmental social science (Agrawal & Perrin, 2009; Shukla et al., 2019; Smit & Wandel, 2006). For example, two nuclear families living in a joint-family household may own separate plots of land, but inputs (e.g., labor, finances) and outputs (e.g., income, yield) might be shared. In our study, two brothers in a joint-family household owned separate but adjacent plots of land and pooled their resources to install a tube well for irrigation. This means that if data are collected from each nuclear family, their report of infrastructure such as this tube well, as well as income, livestock, and other sources of capital may not accurately reflect their lived experiences if shared in the joint families. For example, both two nuclear families in a joint-family household may report owning land, even though only one has formal ownership. Likewise, one may not report ownership of a land, but have access to the shared land. Thus, lack of clarity in household structure can lead to double-counting resources or misrepresenting access to capital.
The joint-family structure also challenges assumptions about household decision-making and the related question of headship. For example, in our study, one member of a joint-family household worked in aquaculture and another ran a restaurant. They combined their incomes to support their joint families, but each made decisions for their own livelihood strategy. It became unclear who, then, should be identified as the household head or primary decision-maker because household-level analysis often assumes that there is one decision-maker (Snyder & Kilgore, 2018). In a joint-family structure, it is possible to have multiple members functioning as household heads, each heading their own nuclear family, but sharing responsibility or deferring to an older generation for the joint-family decisions. Thus, the assumption that each household has one head, or that headship is shared by two partners, may result in inaccurate data and misrepresentations of decision-making and resources (Snyder & Kilgore, 2018).
Overall, this case demonstrates complex resource access and environmental decision-making processes present in joint-family households that do not fit within the traditional understanding of nuclear families and multi-generational stem families. These nuances become especially salient for researchers examining environmental social science, highlighting the need for researchers to clearly distinguish joint-family household structure in household data collection when this structure is present.
Seasonal migration and land tenure in Bolivia
This case draws from a household survey in Santa Cruz, Bolivia (McWherter et al., 2022), to illustrate how commonly used household definitions may not capture seasonal labor migration, educational migration, and land tenure structures. In developing the survey, we defined households as “people who normally live, eat, and sleep in the house, including those not there that day (e.g., are traveling or have left for the city),” informed by Grosh and Glewwe (2000). Utilizing this definition, we collected 1030 surveys from self-identified household heads. During follow-up qualitative interviews, we realized that understanding households’ natural resource management decisions in the region requires a more nuanced understanding of household members’ movements and land sharing arrangements.
First, our case shows the need to better account for complex movements of household members. Specifically, seasonal migration in Bolivia is a relatively common practice for wage labor and educational purposes among agricultural households. For example, traditional cattle ranchers rely on multiple ranches dispersed across communities, and some farmers would spend months away from their home as farm labor during peak agricultural times (Bastia, 2009; Kirshner, 2010). Additionally, agricultural households often send their children to high school in urban areas due to lack of educational access in rural areas. In this situation, children may live alone with extended family, or a parent (usually the mother) will move temporarily to care for the child (Punch, 2002). Thus, household composition may vary depending on when a household is surveyed due to these movements. In addition, this seasonal migration can result in an inaccurate understanding of households’ financial resource sharing by obfuscating income and labor inputs. For example, many households were financially supporting family members living outside of their home, suggesting they should have been counted as household members to more accurately capture resource sharing within the household.
The second lesson we learned from this case is that Bolivia has a diverse set of land tenure structures, including public, private, communal, and shared (‘indiviso’) tenures (Assies, 2006; Lastarria-Cornhiel, 2009). This variety of land tenures goes beyond a purely private regime in which one household neatly matches one (or several) well-defined plot(s) of land. Land ownership and use is important for understanding household wealth, agricultural production, and conservation practices, yet our research revealed that conventional household definitions often overlook land plots outside of those owned by household heads and members, especially when the plots are shared. For example, in our context, ‘indiviso’ land tenure (e.g., adult siblings or neighbors share a plot of land) is common. However, applying our household definition or other commonly used definitions such as “persons living together who make common provision for food and other essentials for living” (United Nations, 2008, p. 128) may result in researchers failing to connect a household to an ‘indiviso’ land plot, or potentially double count that land when surveying both households. Such survey results would likely mischaracterize land-use decisions and the associated land-use negotiations between individuals within a household or between households.
Overall, this case demonstrates that seasonal migration and land tenure can influence how researchers understand household structures and compositions, which also affect their understanding of households’ wealth, income, access to financial and other resources, labor inputs, and land-use decisions. This has implications for how researchers examine natural resource management and conservation practices at the household level. Thus, this case demonstrates the need for a holistic and comparable process to characterize and describe household composition that considers seasonal migration and resource (including land) sharing within and among households.
Mobility and social-ecological change in Peru
Drawing on a study of four communities (Majes, Madrigal, Cabanaconde, Caylloma) in the Caylloma Province of Peru, we illustrate how within-household mobility may be overlooked when using traditional household definitions. We conceptualize mobility as all forms of movements of persons, including tourism and internal displacement due to climate change (International Organization for Migration, 2019, pp. 93–94). In this study, we conducted 130 semi-structured interviews with farmers, miners, pastoralists, and farmworkers and analyzed how interviews’ intersectional identities related to their capacity to adapt to social-ecological changes impacting the Caylloma Province, including glacial melt, mining development, and the implementation of a large-scale water transfer project (LWTP) in Majes (Erwin et al., 2021; Bury et al., 2013; Moraes et al., 2023; Popovici et al., 2021).
Through the data collection and analysis stages, we found that many households owned multiple homes, often across two or more districts, in other Peruvian provinces, or abroad. Due to the Majes LWTP, many households were allotted land in Majes District, either because they won it through a lottery system, or they were given land as compensation for their highland farms being inundated for constructing the Majes LWTP (Stensrud, 2016; Vera Delgado & Vincent, 2013). Some households also obtained land in informal settlements in Majes (Erwin et al., 2022, 2023). While these households had a house in Majes, many also kept their house in their town prior to settling in Majes. Following qualitative interviews, we realized that mobility would pose significant challenges to our planned household survey across sites. Which household would be sampled could depend on where the household was in their cycle of mobility. Depending on when a household and its members would move from one place to another, we could miss a household, mis-capture the composition of a household, survey the same household multiple times, or count the same individual as a member of different households that were in fact the same household.
This case also reveals how traditional household definitions may limit researchers’ ability to accurately document how communities are affected by and adapting to social-ecological changes. Most household surveys like ours are designed to measure how households are adapting in situ by asking what changes a household has made to their livelihood practices without specifying where these changes occur. The assumption of such questions is that households are static, not dynamic (Greiner, 2012; Wilk & Miller, 1997). However, these questions may miss the broader picture of how social-ecological changes affect people’s lives and livelihoods across their homes and various forms of adaptation they adopt at different home sites. For example, in Majes, households depended on the Majes LWTP for water, and increased temperatures and evapotranspiration had resulted in escalated water needs for crops and livestock. These households adapted by building reservoirs to store water. The same households, when in their homes in the district of Caylloma, relied on natural bogs to provide water for their alpacas and llamas; however, climate change had decreased available water in these bogs. As such, in Caylloma they responded by selling their alpacas and llamas, or in some cases, moving water from rivers onto their property. In brief, the same households adapted to the same changes (increased temperatures and evapotranspiration) in different ways across their home sites. Therefore, using a traditional household definition such as “a group of two or more persons living together who make common provision for food and other essentials for living” (United Nations, 2008, p. 128) without considering mobility may limit researchers’ ability to examine people’s experience of social-ecological changes, the associated impacts on people’s lives and livelihoods, and adaptation to social-ecological change at the household level.
Overall, this case demonstrates how mobility can undermine many commonly used household definitions (Greiner, 2012; Wilk & Miller, 1997). While our case examines the Peruvian context, mobility is a common adaptation strategy throughout the Global South (e.g., Fernández-Casanueva, 2021; Radel et al., 2018). There is a need for researchers to consider mobility in their sampling design and use more inclusive ways to describe households that incorporate mobility as a household feature.
Polygamy in Senegal
This case illustrates how commonly used household definitions do not explicitly address polygamous household structures, which has implications for understanding households’ management of environmental and natural resource challenges (e.g., drought, flooding, soil health) with survey research. Worldwide, 2% of the population lives in polygamous households, but the rate is substantially higher in some regions. In sub-Saharan Africa, 11% of the population lives in polygamous households (Kramer et al., 2019). In a 2016 survey of 1981 smallholder farming households in southern Senegal, we found that about one-third of households practiced polygamy (Bauchet et al., 2021). Of these, 80% included two wives, 16% three wives, and the remainder four to six wives (average, 2.2 wives).
As the UN definition does not explicitly address the relationship among “persons living together” (United Nations, 2008, p. 128), the common assumption of monogamy can lead to inaccurate representation of households. Household survey data are often collected from a self-identified household head (Kleinjans, 2013). Doing so without acknowledging the potential polygamous household structure poses two challenges to producing accurate data that can help understand households’ relation with the environment. First, survey questions often rely on the head to accurately remember detailed information about each household member such as age, education level, occurrence of sickness, time use, or natural resources use. Even errors in the simple count of household members can be magnified in polygamous households. In the case of Senegal, higher counts of household members were more likely to be recorded for polygamous households. The 95th percentile, 99th percentile, and maximum numbers of household members were all higher in polygamous households than in monogamous households: 33, 57, and 102 versus 24, 32, and 70, respectively. Larger households are a natural consequence of polygamy, but the likely inaccuracy in these numbers is more pronounced for polygamous households. Similar or possibly larger errors likely apply to more complex questions such as those about natural resource use, environmental management, and/or related decision-making processes.
The second challenge arises when researchers attempt to collect data from more than one household member. In the Senegal study, we surveyed the household head and his/her spouse, as commonly done in other household surveys. To contain survey time and cost, we only asked to talk to the head’s main wife. The concept of a main or primary wife is culturally well-established in Senegal (Lépine & Strobl, 2013); however, information may have been missed when the main wife did not know certain aspects of the life of other wives and their children such as their educational practices, health and health care utilization, or natural resources use. Information may also be biased when characteristics and decisions of the main wife were not representative of those of other wives, which is unobservable to the surveyor (Gibson et al., 2007). In the context of rural Senegal, the main wife may enjoy superior access to water sources, agricultural or forested plots, or inputs (fertilizer, pesticides, information/training). She may also have priority in using tools or hiring agricultural help. As a result, information captured from the main wife may have led to biased generalizations about female members of the household, or about the household in its entirety.
Overall, this case demonstrates that overlooking polygamous household structures in describing households compromises researchers’ ability to make inferences about household resource use and environmental management behaviors, at least in some parts of the world. Researchers who work in societies where polygamy is practiced need to explicitly incorporate polygamous household structures in order to more accurately collect data from multiple household members who may have different levels and types of knowledge about the household as a whole.
Next step: A protocol to identify and describe households
The cases from Pakistan, Bolivia, Peru, and Senegal highlighted how the current lack of specificity in household definitions impact both our understanding of households’ interactions with their natural environment, their resource access, use, and decision-making processes, and the quality of household data collection. These cases encapsulate many common complexities associated with collecting data from households around the world. Specifically, multi-generational (and the related joint-family) and polygamous households have long been prevalent structures in multiple parts of the world (Ruggles & Heggeness, 2008; United Nations, 2017), and multi-generational households are increasing in places where they have not been the norm (Bengtson, 2001; Easthope et al., 2017; Keene & Batson, 2010). Seasonal migration and mobility are increasingly used by resource-dependent communities globally for multiple, complex reasons, including adapting to impacts of social-ecological changes (Burnham & Ma, 2016; Erwin et al., 2022; United Nations, 2019b). We found that existing household definitions and guidance on creating household descriptions (i.e., United Nations, 2008) have not sufficiently attended to these complexities.
To address this challenge, we developed a protocol (Fig. 1) based on the case studies and literature presented above to guide researchers toward explicitly and systematically considering various structures and characteristics of households in their research context. Descriptions and references for the protocol are in Table 1. This protocol highlights a series of considerations in the form of questions about three components of a household (see the top three panels in Fig. 1): (1) household composition; (2) distribution and movement, and (3) resource sharing and ownership. These questions are not meant to be exhaustive, nor does every question in the protocol apply to every case. Indeed, the terms used in this protocol are in and of themselves difficult to define and not applicable in all settings. Similarly, the answers provided in the protocol (see the bottom three panels in Fig. 1) represent possible but not exhaustive answers to the questions posed to describe households in any given study area. Further, this protocol is not meant to replace broadly used definitions that allow for international comparisons; rather, we present it as an additional way to conceptualize a unit of data collection and communicate research. As such, we seek to introduce an explicit, nuanced, and shared process of conceptualizing and describing households that may be iteratively revised as other studies use and add to the protocol.
Fig. 1
Proposed protocol for creating a flexible process to identify and describe households
Table 1
Descriptions, examples, and references for the proposed protocol for characterizing and describing households
Category
Sub-category
Description/definition
Example
Reference
HOUSEHOLD COMPOSITION
Nuclear
Household composed of one or two related generations
Two partners and their children
de Laiglesia, J. R., & Morrisson, C. (2008). Household Structures and Savings: Evidence from Household Surveys. Zurich: OECD Development Centre.
Multi-generational
Household composed of three or more related generations
Two partners, one of their parents, and their children
Ruggles, S., & Heggeness, M. (2008). Intergenerational coresidence in developing countries. Population and Development Review, 34(2), 253–281.
Joint
One or both partners, two or more of their children, and their children’s partners and children
Two partners, two of their sons, their son’s partners, and their children
Ruggles, S. (2010). Stem families and joint families in comparative historical perspective. Population and Development Review, 36(3), 563–577.
Stem
One or both partners, one of their children, and their child’s partner and children
Two partners, their son, his wife, and their children
Monogamous
Partnership of two adults
One wife and one husband
de Laiglesia, J. R., & Morrisson, C. (2008). Household Structures and Savings: Evidence from Household Surveys. Zurich: OECD Development Centre.
Polygamous
Partnership of more than two adults
Two wives and one husband
Unrelated
Households includes individuals not related by blood, partnership, or law
Household composed of only individuals related by blood, partnership, or law
Married adults and their children
DISTRIBUTION AND MOVEMENT
Migration
The movement of a person or persons “away from his or her place of usual residence, whether within a country or across an international border, temporarily or permanently, and for a variety of reasons” (IOM, 2020)
Movement away from place of residence to pursue economic productivity through labor
Members of the household have moved away from the place of residence, either permanently or seasonally, to pursue work in a place outside of the geographic boundary
Burnham, M., & Ma, Z. (2016). Linking smallholder farmer climate change adaptation decisions to development. Climate and Development, 8(4), 289–311.
Educational migration
Movement away from place of residence to pursue economic productivity through labor
Members of the household have moved away from the place of residence, either permanently or seasonally, to pursue work in a place outside of the geographic boundary
Punch, S. (2002). Youth transitions and interdependent adult–child relations in rural Bolivia. Journal of Rural Studies, 18(2), 123–133.
Multi-local
“Two or more spatially dispersed residential units, united in joint decision-making under the imaginary roof of one single household” (Greiner, 2012: 207)
Parents live in the country, but the children live in the city, but they consider themselves to all be part of the same household and/or they share resources
Greiner, C. (2012). Can households be multi-local? Conceptual and methodological considerations based on a Namibian case study. DIE ERDE–Journal of the Geographical Society of Berlin, 143(3), 195–212.
Mobile
“The movement of an entire household across space to a new location” (Agrawal, 2009)
In pastoralism, households often move together between different locations to care for animals
Burnham, M., & Ma, Z. (2016). Linking smallholder farmer climate change adaptation decisions to development. Climate and Development, 8(4), 289–311.
RESOURCES
Land tenure
Relationship (legal or customary) among people (as individuals or groups) with respect to land. It defines how property rights to the land are allocated, what access is granted to use and transfer the land
Food and Agriculture Organization. (2002). Land Tenure and Rural Development. FAO Land Tenure Studies 3. Rome, Italy: Food and Agriculture Organization of the United Nations. https://www.fao.org/4/y4307e/y4307e00.pdf.
Individually owned land
Land that is assigned to a private party. May be owned by spousal couple or household head/individual that is not shared with others
A married couple owns and has exclusive rights to an agricultural parcel
Food and Agriculture Organization. (2002). Land Tenure and Rural Development. FAO Land Tenure Studies 3. Rome, Italy: Food and Agriculture Organization of the United Nations. https://www.fao.org/4/y4307e/y4307e00.pdf.
Household shares ownership
Land that is collectively titled by multiple individuals living together in the household
A father and son living together collectively own and manage grazing land for their different cattle
Assies, W. (2006). Land tenure legislation in a pluri-cultural and multi-ethnic society: the case of Bolivia. The Journal of Peasant Studies, 33(4), 569–611.
Multiple households share ownership
Land that is owned by multiple households. A right of commons exists within the community whereby each household has the right to use the property of the village
Village has community land that members graze cattle on
Food and Agriculture Organization. (2002). Land Tenure and Rural Development. FAO Land Tenure Studies 3. Rome, Italy: Food and Agriculture Organization of the United Nations. https://www.fao.org/4/y4307e/y4307e00.pdf.
It is important to note that this protocol is not a survey instrument; that is, we do not suggest researchers ask the questions directly to respondents. Instead, this protocol can inform the development of context-appropriate survey questions to gather data on the three components of a household. It can also direct researchers to secondary data that relate to these considerations, or aid researchers in deciding if preliminary, likely qualitative, data, are needed to first generate a better understanding of the study context prior to developing a survey.
We propose that this protocol be used at three stages of research: (1) during research design, it can be used to operationalize the definition of a household to inform sampling strategies and the development of data collection instruments; (2) during data collection, it can be used by researchers to inform survey questions about household composition, and (3) when communicating research results, it can be used to systematically describe households in the research. To demonstrate how this protocol can be used, we applied it to the four cases presented in this paper (Fig. 2). We are not able to explicitly describe our application in this paper due to limited space, but we use Fig. 2 as an example to show how this protocol could be applied to help us better understand our study populations, particularly the three components of a household in each of our study context. We recognize that our cases are limited to rural areas in the Global South. While similar complexities are documented throughout the world (Bengtson, 2001; Easthope et al., 2017; Keene & Batson, 2010), it will be important to apply our protocol in urban areas and in the Global North. This type of further research is needed to test the use of the protocol in various contexts to assess its utility.
Fig. 2
Application of the proposed protocol in characterizing and describing households in the four cases—joint-family structures in Pakistan, seasonal migration and land tenure in Bolivia, household mobility in Peru, and polygamy in Senegal. These descriptions are based on the case studies described in this paper and therefore, may not necessarily represent the broader geographic areas
Finally, while a full discussion of the political nature of defining a household is beyond the scope of this paper, we recognize that definitions of social units such as households are not simply descriptive, are not politically or culturally neutral, and are contested among some scholars (Harris et al., 2020; Meers, 2021). Definitions of households used by those who collect household data can sometimes be prescriptive, implying certain ways of organizing societies from the political and cultural perspectives of those in power (Hamzah & Adnan, 2016). Thus, our proposed protocol should be applied with care and in collaboration with community members from the research area.
In contrast to traditional household definitions, our proposed protocol provides a process that prompts researchers and practitioners to explicitly consider and incorporate important heterogenous and nuanced household structures and characteristics into their research. Not all elements of the protocol are relevant to all research contexts, but documenting such absences using this protocol is also informative. It is our hope that the use of such a protocol will contribute to the transparency, replicability, and credibility of environmental social science scholarship that uses household data and thus improve our understanding of the questions raised within these research realms.
Acknowledgements
We thank the following organizations that supported the four case studies described in this paper: National Science Foundation (award # 1660481); United States Department of State and United States Agency for International Development (USAID) under the Pakistan—U.S. Science & Technology Cooperation Program (award # AID-391-A-17-00001); Arequipa Nexus Institute for Food, Energy, Water and the Environment, a partnership between the Universidad Nacional de San Agustín (UNSA) in Arequipa, Peru and Purdue University in Indiana, USA; and USAID’ Feed the Future Innovation Lab for Food Processing and Post-Harvest Handling [award # AID-OAA-L-14-00003]. The two points discussed regarding the political nature of household definitions and the intentional ‘illegibility’ of households in some communities were developed based on comments from an anonymous reviewer. We thank this reviewer for these two important insights.
Declarations
Conflict of interest
The authors declare no competing interests.
Open Access This article is licensed under a Creative Commons Attribution-NonCommercial-NoDerivatives 4.0 International License, which permits any non-commercial use, sharing, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if you modified the licensed material. You do not have permission under this licence to share adapted material derived from this article or parts of it. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by-nc-nd/4.0/.
Publisher's Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
