Why should households care?

A rejoinder

Anke Niehof

01 Sep 2004


This rejoinder addresses the comments on the article “A micro-ecological approach to home care for AIDS patients”. The key issue of my reply is the rationale for choosing the household as a level of analysis in the MEAH framework, since critical questions were asked about this in the comments. Furthermore, the rejoinder discusses the methodological implications of applying the model and attempts to provide further clarification on the significance of external linkages for households and the concepts of vulnerability and coping.

The concept of household

The validity of the central position of the household as a unit of analysis in the MEAH framework is questioned in the reviews. The critique of Tanja Müller is especially focused on this point. She wonders whether it would not have been better to replace the concept of ‘household’ with that of ‘cluster’. However, I am not convinced that this will solve the problems of boundary and control over resources, even in a situation in which HIV/AIDS is pandemic and households dissolve and disappear. I will try to argue my position below.

Müller and Pool are not the first and will not be the last to question the applicability of the concept of household to rural communities in sub-Saharan Africa. Burch et al. (1993: 20) note the following: “We assume that census concepts of household are generally comparable from one time and place to another. This assumption is valid for the vast majority of population censuses, although problems arise for particular cases […]; the assumption may be most problematic for Tropical Africa.” Goody (1990: 130) comments in the same vein: “One of the simplest variables from which to obtain information on African families is mean household size, which forms a constant of most census records. But the results are frequently difficult to use, since they are usually based on the concept of hearth (cooking unit), which may be embedded in wider groupings, namely the farming group, and on the concept of household or dwelling group.” In a well-known paper Bruce and Lloyd (1992) point to the importance for households of kinship networks and other support networks in which they are embedded. In her research in rural Kwa-Zulu Natal (South Africa) Mtshali (2002) comes to the conclusion that in that area, ‘household’ can be equated with ‘homestead’ because the latter is the effective unit for resource management and decision-making (under the patriarchal leadership of the eldest man). Russell (1993) sketches a census enumerator’s nightmare in an area with a high level of labour migration in rural Swaziland. The patrilineal farmstead (umuti) encompasses groups of women with children who have their own kitchen. Sometimes the men join these groups, at other times they area away, working as migrant labourers. Male labour migration but also polygamy causes these ‘kitchen units’ to be flexible in size and composition and even to overlap at times. In my own research in Madura, Indonesia (Niehof 1985), I refrained from using the term household but, instead, used the terms ‘kitchen units’ (using separate kitchens) and ‘cooking units’ (sharing a kitchen but budgeting and cooking separately) for the group sharing resources and food on a daily basis.

The examples testify to the universality of primary groups within which daily life is organized, daily needs are attended, and shelter and care are provided, whether such groups are called household, homestead, kitchen unit, or anything else. Their boundaries may be permeable and shifting, their composition changing over time, the capabilities they comprise may be more or less adequate, but still such groups or units represent empirical manifestations of household. Kabeer puts this well in her summary of her rationale for retaining the concept of household: “The empirical significance of household relationships in the daily management of resource entitlements, and as the routine context of people’s lives, suggests that it has a certain facticity, despite its shifting guises” (1994: 114). I also agree with Davidson (1991) that we do not need to establish a rigid and all-purpose definition of household but that we can agree on a working definition, as long as we take into consideration the fluid and dynamic nature of households. For the household base underlying the MEAH framework I find the working definition of household by Rudie the most suitable one. Rudie (1995: 228) describes the household as “a co-residential unit, usually family-based in some way, which takes care of resource management and primary needs of its members.” At the same time I fully endorse the warning note of Davidson (1991: 13) that “household as an entity represents but a moment in the dynamic process of its continual formation and reformation.”

In a situation of high HIV/AIDS prevalence, as in Southern Africa, this reformation may be very drastic. In such circumstances households may change fundamentally with regard to the gendered division of reproductive and productive labour, allocation of care-giving and care-receiving roles, access to entitlements, and even power structures and decision-making. The Zulu household might lose its patriarchal character, for example. But if we do not keep the concept of household, however much deconstructed, we will not be able to see those changes. The argument above to justify sticking to the concept of household in the MEAH framework does not imply turning a blind eye to the problems of its delineation (Pool), intra-household inequities, and beyondhousehold linkages (Müller, Pelto, Young). However, these are problems at the level of application of the concept and validity of its operationalisation, rather than at the conceptual level.

The dynamic nature of households and its methodological implications

Maria Koelen is much more positive about taking the household as the unit of analysis. She even sees this as a strength of the model, because it fills a gap that is left open in much health promotion research. However, she raises a relevant methodological issue. If, as in a situation of HIV/AIDS, households tend to break up and dissolve because they could not cope and therefore cannot be traced with traditional household surveys, then how can one apply a model in which the household is the unit of analysis? The answer is that we can only trace the history of such households and former members who have become part of other households or have founded new households (orphanheaded households, for example), when we insert a temporal dimension in the study design. This can be done by including retrospective questions in the survey, conducting case studies and life histories to supplement the survey data, or use longitudinal (panel) data. The latter will be done in a Wageningen PhD project on care for AIDS patients, which is to be carried out in the province of Kwa-Zulu Natal, South Africa. For a large number of households in Kwa-Zulu Natal, such panel data were collected by a research organization for the last five years at three-months intervals. This research project, which has just begun, is intended to shed light on the processes whereby AIDS-afflicted households dissolve and their members become part of other or new households. It will also examine what implications for care these shifts entail. In short, this project will attempt to capture the mechanisms of the flux.

When households are viewed as a set of social relationships geared towards a common purpose and jointly managing resources to this end, it follows that the intra-household bonds will loosen when the members no longer experience or perceive this commonality of purpose. I agree with Müller that in such cases individual members may be better off just fending for themselves, as in the case of Liza. Then, the external ties of the household (including the social relationships in the cluster) become more meaningful to the individuals. As a result, the household cannot be kept together anymore and falls apart. I agree, therefore, with Gretel Pelto about the significance of external linkages, which she says I leave rather implicit. However, unlike Pelto, I believe that a distinction should be made between micro-meso linkages (visible in the case of Liza and relating to community structures and beyond-household support networks) and micromacro linkages (the relation of the household to the political economy in which it embedded, for example).

Vulnerability, coping, care and inequality

The problem with an individual focus for explaining coping behaviour, as Müller proposes, is that it ignores the functioning of the asymmetric support relations within households. Because of its family base, the household is a context of generalized reciprocity in which some give more and others take more; the balance shifts through time. Very vulnerable persons cannot take care of themselves and cannot give care to others. Households are expected to provide care to their most vulnerable members, such as AIDS patients or infants, and this care provision is part of the household’s resource allocation and management. It is true that Sam no longer take part in joint resource management and does not contribute to the resources of the household, but that is also true of infants, children or frail elderly in households. Managing these asymmetries for the benefit of all is part of the morally and culturally underpinned mission of the household. This only works if the common benefit and the right to care of the most vulnerable are acknowledged and the demands for care and support do not exceed the capacity of the household and its caregivers to provide it. In the case of Sam and Liza, and in the case of many households with members suffering from AIDS, those two conditions no longer hold true.

Maria Koelen raises questions about the concept of coping. I agree with her conclusion: If households cope they lose (because they have to cash in their assets and use the last resources they have) and if they cannot cope they lose as well (they are lost). I see coping as a short-term non-premeditated response to a crisis. The point is that households may be able to cope for a short period, after which external support or a change of circumstances may come to their help and may enable them to survive. Elsewhere I phrased this in relationship to livelihood systems as follows. Extremely vulnerable livelihood systems break down in a situation of stress because of lack of assets and an inability to develop effective coping strategies, and are dependent upon external support for survival (Niehof & Price 2001). The same applies to households that are vulnerable as a consequence of living with HIV/AIDS. They can only cope and survive with external assistance. Such assistance may be provided by NGOs or by new organizations especially formed to help households cope with AIDS in areas where prevalence is high (Lwihula 1998).

Gretel Pelto rightly points to the importance of care for child nutrition, and refers to the famous UNICEF framework. I am glad that she raises the issue of nutrition because originally I had planned to use nutrition as a second example that the MEAH framework could be applied to. But since the format of one article was too short for this to be feasible, I restricted its application to AIDS. In a recent PhD thesis at Wageningen University (Balatibat 2004) a model was used in which the UNICEF framework was combined with a livelihood approach, to look at household food security and child nutrition in the Philippines. The model proved useful for eliciting the underlying causes, at household level and beyond, of child malnutrition. To explain why even in food secure households children may be malnourished it is essential to look at patterns of child care within households.

Sera Young questions the role of the household in the Madura example in my article. Yes, of course, it is the individual who goes to the healer because it is the individual who is ill, just as individuals and not households are admitted to hospitals. But the household has to bear the costs and, if needed, make the arrangements. This is also a form of care, belonging to Tronto’s second phase of care: ‘taking care of’. In decisionmaking about the treatment of ill household members, cost is an important factor. Another factor in the choice for treatment is the prevailing folk etiology of illness and disease. It is not only this individual with these symptoms who goes to this kind of healer, but other individuals in the same community and in the same situation will likely do the same. This is where Pelto’s belief-related factors come in. The issue of the household allocating the means for treatment of ill household members also touches on the power issue that Gretel Pelto would like me to have paid more explicit attention to. In societies with a high degree of gender inequality such resources are reserved for men rather than women, and for boys rather than girls; women are unable to do much about it. The literature on the poor state of women’s reproductive health and the continuing high rates of maternal morbidity and mortality in many countries, testify to the unequal distribution of power and care between the sexes both within and beyond households.

For the conceptualization of care in the MEAH model the work of Tronto (1993) was used. Maria Koelen’s question about the status of the fifth requirement of good care in the MEAH model can be answered very briefly. Yes, absolutely, integrity is a condition for good care, not just a requirement. If I did not make that clear in the article I hope I have done so now. In countries where care involves several specialized actors apart from those at the household level, this is exactly where care may not be good care, in spite of all the efforts put into it.

Pathway analysis?

The suggestion of Young to apply pathway analysis to care-giving and care-receiving relations is an interesting one, but its application would have to cover the broad definition of care that I used in the MEAH framework. It should also be able to make visible the patterns of ‘caring about’ and ‘taking care of’, as well as the degree of integrity of the care process. Perhaps this is possible. It is certainly worth trying. For such an analysis it is helpful to see the household as a condensed web of particularistic, manystranded and more or less ascribed relationships within which care is provided. When moving beyond the household along the nexus described by Young, the web widens and unfolds and the relationships become less particularistic and more single-stranded. The different phases of care may be located at different positions on this nexus, and the same may apply to different types of care, like daily care, emotional care, instrumental care, medical care, and so on. Furthermore, the location on the nexus may vary according to the magnitude and nature of the care demand. It might well be that by using pathway analysis these patterns can be captured and visualized. Even so, I still think that the household should be the starting point for the analysis, not the needy individual. I agree with Koelen that the core of the MEAH model is about the interaction of needs and their provision and patterns of care emerging from this.

Emic-etic and adequacy of care

In his criticism of my use of the emic-etic distinction Robert Pool touches upon a crucial issue, namely the problem of judging the adequacy of care. He is right that a value-free objective assessment is impossible. However, the whole point of applying the MEAH framework is to be able to assess adequacy according to the different phases of care and their specific requirements, the way and the extent to which these phases are integrated, and the adequacy of the resources that are available and mobilized to address care needs. For all these components valid indicators have to be found. In this way, the concept of care becomes better amenable to investigation, and judging the adequacy of care will be based on comprehensive conceptualization and transparent operationalisation. Pool is right that in the table the emic-etic distinction seems to be more or less narrowed down to intentions and measurable health effects and that the latter can be emic as well. Indeed, the crux of the distinction lies in the vantage point, the etic one being that of the “community of scientific observers” (Harris 1968: 575) – for judging measurable health effects the bio-medical scientist – and the emic one, being that of the “actors [care givers and care receivers] themselves” (Harris 1968: 571). Of course, the MEAH framework also represents the etic vantage point of the social sciences, which is why it is an interdisciplinary one. The interdisciplinarity and different perspectives make the framework far from simple, but care is about complex and resilient empirical realities, the understanding of which offers no easy shortcuts. While Pool questions the validity of the table referred to above, Koelen underlines its usefulness. She points to the fact that the third category in the table, namely that of ‘HHPH practices not intended as health care but yielding measurable health effects’, is one that is often overlooked in domain-specific models with a focus on health production.

Other points of criticism

Pool suggests that I “over-rely” on the book by Barnett and Whiteside (2002). If he were right it would not be such a bad thing, given the comprehensive nature of the book and the broad scope of recent literature reviewed in it. There is a strong body of literature on AIDS and rural livelihood, which connects easily to the household resources perspective in my framework (e.g. Barnett 1995; Haddad & Gillespie 2001; Loevinsohn & Gillespie 2003; Müller 2004) and there is the anthropological synthesis by Schoepf (2001), but theoretical work that highlights the care angle is much more difficult to find, which is why I am grateful for the references to the work of Farmer.

Then there is the pertinent question of Pool about what is ‘new’ about the MEAH model. I would to respond to that by saying that every invention builds on pre-existing building blocks but makes new combinations of these. Even in the physical sciences, inventions did not descent like thunderbolts from a clear sky. Arthur Koestler has written a beautiful book about this, called “The Sleepwalkers”, which I highly recommend. I leave the evaluation of the novelty of the framework, and of the relevance of posing the question, to the community of scientific peers. I hope to have contributed a heuristic and applicable framework for the analysis of care that takes the level of the household as its point of departure.

Why should households care?

“Why should households care?” is the title I gave to this rejoinder. The framework makes clear that households should and do care because for most people they provide the context and the resources for meeting daily needs, including care needs. In a situation in which there is a lack of alternatives to institutional care, where can AIDS patients seek refuge other than in the households of their families or relatives? If they don’t find it there they will be ‘socially dead’, such as the case of Judy in an article on the access to anti-retroviral medicine in Uganda (Meinert et al. 2004) tells us. After losing her husband and being separated from her son, she is ostracized by her brothers and sistersin-law and the rest of the community, and has only her mother to turn to. Judy’s main concern is not medical treatment – she has access to ARV medicine but she is lax in taking it – but food. She worries about who will cook for her and find her the only food that she is able to eat. Once she is denied this care, she stops eating and dies. It is a sad example of what happens to people when they are no longer part of a household that duly fulfills its most important function: taking care of the primary needs of its members.


Balatibat, Emelita M. 2004 The linkages between food and nutrition security in lowland and coastal villages in the Philippines. PhD Thesis Wageningen University. 
Barnett, et al. 1995 The social and economic impact of HIV/AIDS on farming systems and livelihoods in rural Africa: Some experiences and lessons from Uganda, Tanzania, and Zambia. Journal of International Development 7: 163-76. 
Barnett, T. & A. Whiteside 2002 AIDS in the Twenty-First Century: Disease and globalization. New York: Palgrave Macmillan. 
Bruce, J. & C. Lloyd 1992 Finding the ties that bind: Beyond headship and household. Working Paper 41. New York: The Population Council. 
Burch, T.K. et al. 1993 Measures of household composition and headship based on aggregate routine census data. In: J. Bongaarts et al. (eds.), Population research methodology, Vol. 4, Chapter 15. Chicago, Illinois: UNFPA Social Development Center. 
Davidson, A.P. 1991 Rethinking household strategies. Research in Rural Sociology and Development 5: 11-29. 
Goody, J. 1990 Futures of the family in rural Africa. In: G. McNicoll & M. Chain (eds.), Rural development and population. New York: Oxford University Press, pp. 119-47. 
Haddad, L. & S. Gillespie 2001 Effective food and nutrition policy responses to HIV/AIDS: What we know and what we need to know. Journal of International Development 13: 487-511.
Harris, M. 1968 The rise of anthropological theory. New York: Thomas Y. Crowell Company. 
Lwihula, G. 1998 Coping with the AIDS pandemic: The experience of peasant communities of the Kagera region, Tanzania. In: G. Mutangadura, H. Jackson & D. Mukurazita (eds.), AIDS and African small-holder agriculture. Harare: SAfAIDS. 
Kabeer, N. 1994 Reversed realities: Gender hierarchies in development thought. London: Verso. 
Meinert, L., M. Whyte, S. Reynolds Whyte & B. Kyaddondo 2004 Faces of globalization: AIDS and ARV medicine in Uganda. Folk 45: 105-23. 
Mtshali, S.M. 2002 Household livelihood security in rural KwaZulu-Natal, South Africa. PhD Thesis Wageningen University. 
Loevinsohn, M. & S. Gillespie 2003 HIV/AIDS, food security and rural livelihoods: Understanding and responding. RENEWAL Working Paper No. 2. 
Müller, T.R. 2004 HIV/AIDS and agriculture in sub-Saharan Africa. AWLAE Series No. 1. Wageningen: Wageningen Academic Publishers. 
Niehof, A. 1985 Women and fertility in Madura, Indonesia. PhD Thesis, Leiden University. 
Niehof, A. & L.L. Price 2001 Rural livelihood system: A conceptual framework. UPWARD Working Paper Series No. 5. Wageningen: WU-UPWARD. 
Rudie, I. 1995 The significance of eating: Cooperation, support and reputation in Kelantan Malay households. In: W.J. Karim (ed.), ‘Male’ and ‘female’ in developing Southeast Asia. Oxford, Washington DC: Berg Publishers, pp. 227-47. 
Russell, M. 1993 Are households universal? On misunderstanding domestic groups in Swaziland. Development and Change 24(4): 755-87. 
Schoepf, B.G. 2001 International AIDS research in anthropology: Taking a critical perspective on the crisis. Annual Review of Anthropology 30: 335-61. 
Tronto, J.C. 1993 Moral boundaries: A political argument for an etic of care. New York, London: Routledge.