The convergence of HIV/AIDS and customary tenure on women's access to land in rural Malawi

Abstract

This paper examines the convergence of HIV/AIDS and the social processes through which women access customary land in rural Malawi. Data were collected from focus group discussions with women in patrilineal and matrilineal communities. Women's land tenure is primarily determined through kinship group membership, customary inheritance practices and location of residence. In patrilineal communities, land is inherited through the male lineage and women access land through relationships with male members who are the rightful heirs. Conversely in matrilineal matrilocal communities, women as daughters directly inherit the land. This research found that in patrilineal communities, HIV/AIDS, gendered inequalities embedded in customary inheritance practices and resource shortages combine to affect women's access to land. HIV/AIDS may cause the termination of a woman's relationship with the access individual due to stigma or the individual's death. Termination of such relationships increases tenure insecurity for women accessing land in a community where they do not have inheritance rights. In contrast to the patrilineal patrilocal experience, research on matrilineal matrilocal communities demonstrates that where women are the inheritors of the land and have robust land tenure rights, they are not at risk of losing their access to land due to HIV/AIDS.

Abstraite

Cet article examine la convergence du VIH/sida et des processus sociaux par lesquels les femmes accèdent au terres coutumières au Malawi. Les données ont été collectées à partir de groupe de discussion avec des femmes de communautés patrilinéaires et matrilinéaires. La possession des femmes est principalement déterminée par l'appartenance au groupe de parenté, les pratiques d'héritage coutumier et le lieu de résidence. Dans les communautés patrilinéaires, les terres sont léguées à la lignée masculine et les femmes accèdent à la terre par l'entremise de leurs relations avec les hommes, les héritiers légitimes. Inversement, dans les communautés matrilocales matrilinéaires les femmes héritent directement la terre. Cette recherche a révélé que dans les communautés patrilinéaires, le VIH/sida, les inégalités sexo-spécifiques intégrées dans les pratiques d'héritage coutumier et le manque de ressources se combinent pour affecter l'accès des femmes à la terre. Le VIH/sida peut provoquer la fin de la relation d'une femme avec l'héritier légitime à cause de la stigmatisation ou de la mort de celui-ci. La fin de ces relations augmente l'insécurité foncière pour les femmes dans les communautés où elles n'ont pas droit d'hériter un accès à la terre. Contrairement à l'expérience patrilocale patrilinéaire, la recherche dans les communautés matrilocales matrilinéaires démontre que lorsque les femmes sont héritières de la terre et ont de robustes droits, elles ne sont pas à risque de perdre cet accès à la terre à cause du VIH/sida.

Keywords:

• HIV/AIDS
• land rights
• women
• customary
• matrilineal
• patrilineal

Mots-clés:

• VIH/sida
• femmes
• Droits de la terre
• coutumier
• matrilinéaires
• patrilinéaires

Introduction

It is impossible to isolate HIV/AIDS in Malawi as a purely physiological condition. With HIV prevalence rates of 10.8% (10.2%–11.4%) among adults (15–49), this disease has affected both the social and economic frameworks of the nation (UNAIDS 2012). At the village level, the most visible affects include the increasing number of widows due to HIV/AIDS. These observable affects are contrasted by the impacts on social processes, which are often hidden and complex.

HIV/AIDS affects women differently than men. Women have gender-specific social responsibilities, which are often increased due to HIV/AIDS. In Malawi, it is usually the women who care for the sick. HIV/AIDS increases women's burden as caregivers and reduces their time available to grow food (Mbaya 2002). Gender inequalities also increase women's susceptibility to HIV infection. Women face gendered difficulties in negotiating safe sex. In addition, food insecurity may lead to risky behavior and increased susceptibility to HIV Infection. Poverty and hunger cause some women in Malawi and elsewhere to engage in risky behavior, such as transactional sex, that consequently increases their risk of HIV infection (Bryceson & Fonseca 2006; Kishindo 1995; Oglethorpe & Gelman 2008).

For women living in Malawi's rural areas, access to land is of upmost importance. Access, defined as the ability to utilize a specific parcel of earth, is necessary for securing housing, cultivating subsistence crops and growing commercial crops. In these rural areas, women access customary land through social processes specific to the cultural community in which they live. Access to land is closely linked to inheritance. Malawi exhibits both patrilineal communities where men inherit land and matrilineal matrilocal communities where women inherit the land.

The purpose of this paper is to identify the social processes through which women access customary land in rural Malawi and to examine through dedicated fieldwork how their access is affected by HIV/AIDS.

Setting the research agenda: HIV/AIDS, gender and land rights

Research on HIV/AIDS, gender and land rights is a recent phenomenon that emerged after 2000 (Drimie 2003; Floyd, Crampin, Glynn, Mwenebabu, Mnkhondia, Ngwira, et al. 2008; Food and Agriculture Organization 2006; Hilhorst 2000; Izumi 2006; Oglethorpe & Gelman 2008; Walker 2002, 2005). This type of research is significant as it acknowledges that the effects of the HIV/AIDS epidemic extend beyond physical symptoms. By layering HIV/AIDS onto the research platform of gender and access to land, researchers are able to examine how the epidemic has affected societal social structures through which women access land. Scholarship on the social and economic affects of HIV/AIDS in southern Africa evolved in the early 1990s. Research in this field of study has been conducted in many Southern African countries including Kenya, Zimbabwe, South Africa and Malawi. In Malawi, government, NGOs and academia have conducted research on this topic. For example, a study by a consortium of government and NGOs examined the link between HIV/AIDS and food security in Malawi (Ngwira, Bota & Loevinsohn 2001). The study identified land tenure and HIV/AIDS as an important research agenda, ‘knowledge about marriage systems and customary land tenure systems and how these influence widow's rights to land would be important for policy decisions and formulation of area-specific strategies to mitigate suffering among AIDS affected families' (Ngwira et al. 2001:19).

Other research from northern Malawi has supported the theory that HIV is a widow-creating disease by showing that women who are married to men with HIV have a higher incidence of widowhood (Floyd et al. 2008). This retrospective cohort study conducted by Floyd et al. examined the social and economic impacts of HIV on the spouses of HIV infected individuals in Karonga district (2008). Relevant to women's access to land, the study reports information regarding household dissolution and land inheritance specific to widows. Approximately a quarter (26%) of the wives of HIV-positive men experienced household dissolution as a result of their husband's death compared to 5% of women of HIV-negative men (Floyd et al. 2008). However, the authors report that there is not any evidence that suggests that dissolution is more likely after the death of an HIV-positive spouse (Floyd et al. 2008). The study reported equal percentages of widows who inherited the land after their husband's death between widows of HIV-positive and HIV-negative men. These results should be interpreted carefully due to small sample sizes, and lack of information about the location of the land that these women inherited.

An Oxfam study on HIV/AIDS and land issues in Malawi demonstrated that lack of land, HIV/AIDS and poverty are interlinked (Mbaya 2002). Shortage of land is related to poverty, which in turn is related to increased susceptibility to HIV/AIDS (Mbaya 2002). The study argues that there is a ‘gender element' in the effect of HIV/AIDS on a ‘household's economy' as, ‘since most communities in Southern Africa are patrilineal and patriarchal, a household's access to land is frequently dependent on the presence of an able male adult' (Mbaya 2002:3). Thus Mbaya argues that when the male household head passes away due to HIV/AIDS, access to land becomes uncertain. The most significant contribution of Mbaya's (2002) study are the findings that HIV/AIDS affects the quantity/quality of land held; HIV/AIDS-related illnesses effect productive use of familial land holdings and lengthy illness and death due to HIV/AIDS affect crop preference (2002).

This paper builds on the work of Ngwira et al. (2001) and Mbaya (2002) by isolating the means through which women in both matrilineal and patrilineal research communities access land and collecting information from women on how HIV/AIDS might impact their access.

Defining land rights and land tenure

Women's land rights as a concept pertains to the right to access and utilize land. For the authors of this paper, ‘access to land' refers to the ability to utilize a specific parcel of earth. This definition relates to use and is not exclusionary in that several persons may be utilizing the same parcel. It is recognized that customary systems exist in a state of flux with user rights being constantly allocated and negotiated. Women's rights to use customary land can also be described as permission or entitlement to use a specific parcel. Building on the work of Sen (1981, 1990) and Vaughan (1987), Bezner Kerr (2005) has defined ‘entitlements' as ‘legally and socially defined rights to resources' (55).

Land rights take different shapes and forms. These rights extend beyond ownership as women often have user rights, which may not relate to possessory title. Different institutions usually enforce user rights and formal land title. In customary areas user rights are often overseen by family members and traditional leaders, while formal title, if it exists, is under the jurisdiction of local and national government agencies. At the informal level, women's user rights are often informed by local custom, as well as gender relations between male and female family members.

Relevant to land use is ‘land tenure’, the manner in which land rights are held. Land tenure has been defined by Bruce and Migot-Adholla (1993:252) as a ‘bundle of rights’. This bundle is divided into three elements which may be present in various combinations: duration of rights, protection of rights and robustness of rights (Bruce & Migot-Adholla 1993). In examining women's access to land, this study looks at the ability to use land as well as the tenure of this arrangement.

Land tenure in Malawi

In Malawi, access to land is important for subsistence farming and commercial agriculture. This landlocked country in southern Africa is densely populated with 114 people per square kilometer (Benson, Kaphuka, Kanyanda, & Chinula 2002). Land continues to play an important role in the country's economy, as 35.5% of the national GDP in 2006 came from agriculture (World Bank 2007). Malawi's largest commercial export is also agricultural. In 2006 the country exported 245 million US dollars worth of tobacco (World Bank 2007). Beyond the commercial markets, agriculture contributes significantly to the informal or subsistence economy as many households grow produce for their own use. Maize is the dietary staple. The high population to land ratio and the importance of land for agriculture combine to cause heightened land pressure. This situation is further aggravated by an estimated 28% of land that lies unused as documented using aerial photography (Saidi 1999). As a resource land has become increasingly scarce due to rising population pressure and finite availability. Historically, the amount of land available has been limited by the presence of large tea and tobacco estates. Kanyongolo (2004) gives an insight into the colonial land policies under which large tracts of land were transferred from customary systems to white colonial farmers (121). To this day, land distribution in Malawi remains highly unequal. This has been recognized as a policy issue by the Malawian government and has resulted in the formulation of the National Land Policy (Government of Malawi 2002).

In Malawi's rural areas most of the land is held under customary tenure systems. Under the Land Act (1965), customary land is categorized as public land. Thus the government is responsible for its administration. This responsibility has been delegated to chiefs at the local level who are acknowledged by the government as custodians of the land. Over time, the lines between custodianship and ownership have blurred and it is often perceived that the chiefs are the owners of the land. However, under Malawian law the President appoints each Chief and also has the ability to demote him or her. Defined under the Land Act (1965), private land differentiates clearly from public land as it is held under freehold title, leasehold title, a Certificate of Claim or is registered as private land. In contrast, customary land is held communally without formalized individual titles.

In Malawi, both patrilineal and matrilineal kinship groups exist in the country, with Lambya, Tonga, Ngonde, Tumbuka, Nyanya and Ngoni patrilineal ethnic groups primarily in the north and Chewa, Yao, Ngoni, Nyanja and Lomwe matrilineal groups predominantly in the central and southern regions (Berge, Kambewa, Munthali & Wiig 2014; Takane 2008). There is also the Sena who are patrilineal ethnic group in Chikwawa and Nsanje which are geographically located in the shire region of southern Malawi (Berge et al. 2014). Data from the National Statistical Office of Malawi shows that there are two additional ethnic groups, the Senga and the Nyakyusa, who both have relatively small populations (as cited in Berge et al. 2014:63).

While Ibik (1970) classifies the Ngoni people as matrilineal, there are exceptions (as cited in Berge et al. 2014:63). For example, Takane (2008) conducted research in a patrilineal Ngoni community in northern Malawi. Similarly, the Nyanja are classified as matrilineal, but the 2007 National Census of Agriculture and Livestock reported on by Berge et al. (2014) found Nyanja villages in northern Malawi that practice patrilineal inheritance.

Both patrilineal and matrilineal groups have important implications for women's ability to access land. Within Malawi's customary areas, inheritance and dispossession often overlap in terms of women's land user rights. Women's rights to access land are shaped by kinship inheritance practices and residency patterns. As kinship varies throughout the country, so do women's rights to access customary land. For example, women living in matrilineal matrilocal communities have the right to inherit land, while women living in patrilineal patrilocal communities do not. In terms of residency, matrilocal (uxorilocal) refers to the residence pattern where a married couple live with the woman's community, which is common among the Yao, Nyanja, Lomwe and Ngoni matrilineal communities of southern Malawi (Berge et al. 2014). In patrilocal (virilocal) residence, the married couple goes to live with the husband's community. According to the 2007 national census of agriculture and livestock, virilocality is common among the Chewa ethnic group of central Malawi (Berge et al. 2014). In some areas uxorilocality and virilocality coexist which is more prevalent in the central region (Kishindo 2010). In both patrilineal and matrilineal communities, residency creates variations in inheritance traditions (Berge et al. 2014). Residency can subsequently change a female child's status as an inheritor of the land. Takane (2005) provides an example of a Chewa community that was traditionally matrilineal but had a flexible inheritance system where women and men obtained their land matrilineally and patrilineally. Takane describes these flexible inheritance practices as an adaptation to land shortages (2005). Despite these variations, it is widely recognized that matrilineal inheritance, succession and uxorilocal marriages are predominant in central and southern Malawi (Kishindo 2010).

The extent to which women in matrilineal communities have authority over land has been an issue of debate. The debate surrounds the extent of male influence on women's land-use decisions in matrilineal communities. In contrast, the literature regarding patrilineal communities is far more harmonious (Mbaya 2002). In patrilineal patrilocal societies, women have user rights to land which belong to their husband's kinship group. Women who do not have the right to inherit land under customary law access it only through their husbands and are frequently dispossessed of such land upon divorce or the death of their husband (White & Christian Literature Association in Malawi 2002).

Kinship may be broadly defined as, ‘relationships between persons based on descent or marriage' (Stone 2006:5). These relationships are presumably based on some level of rights and obligations. Schneider (1961) isolates kinship and descent as two separate but interrelated concepts. ‘Kinship' consists of the rules and principles of a group, while ‘descent' relates to membership in accordance with the rules (Schneider 1961:2). In regions that practice ‘unilineal' descent, such as Malawi, Schneider explains that descent units can be either patrilineal or matrilineal. In this paper, the terms patrilineal and matrilineal are used to describe the descent units/groups as identified by Schneider (1961). In patrilineal lineages descent is traced through males, while in matrilineal lineages descent is traced through females.

Kinships as social institutions are shaped by the residency patterns of their members as well as the configuration of post-marital habitation. Residence is important as it impacts the power kinship members have over available natural resources. For example, a woman who relocates to the husband's kinship community theoretically has less right to access local natural resources than if the husband had moved into her kinship area. Kinship descent and residency patterns combine together to create a complex array of possible rights and obligations for individual members. These interwoven patterns are further intensified by a variety of marriage typologies. In the communities that contributed to this research paper, there exists monogamy in the matrilineal communities and polygamy in the patrilineal communities. In polygamous unions, there are more competitors for familial resources than in monogamous marriages.

Land tenure for women in patrilineal communities

Patrilineal systems are often patrilocal, in that upon marriage the wife moves to the husband's community. In patrilineal patrilocal communities, men are most often the leaders and decision-makers. Under the patrilineal system of descent, lineage is transferred through the male line. Right to land (either ownership or use) also follows the system of descent. Customary law within patrilineal patrilocal kinship groups in Malawi dictates that:

women can only access land through their husbands and sons. Upon divorce with the husband the woman loses rights to cultivate her field and has to return to her own village. Upon the death of the husband the woman as long as she is unmarried can use the land her husband owned as she grows old, as the sons grow old, she shares the land with her sons and may be squeezed out of land. (Shawa 2002 in Mutangadura 2004:11)

This paper uses Mutangadura's description of women in patrilineal patrilocal communities accessing land ‘through' an individual who has inheritance rights (2004). Joireman (2008) describes women in this situation of having ‘secondary rights' as they cannot access land ‘in their own right' (1238). The majority of customary tenure systems in southern Africa are patrilineal and this type of land tenure for women is commonly described in the literature (Joireman 2008; Oglethorpe & Gelman 2008; Peters 2010). While there may be some deviations in who inherits the land in patrilineal, patrilocal communities, it is most often those in the male lineage. Takane (2008) cites several examples from communities in northern Malawi where strong relationships led to non-patrilineal land transfers; however, these are exceptions to the rule.

It is important to note that a woman's land tenure in patrilineal patrilocal communities is often closely associated with the authenticity of her marriage as determined by the payment of a lobola or bride price (Takane 2008). Takane conducted fieldwork in patrilineal communities and found that widows for whom the husband had paid a lobola were permitted to stay and use their late husband's land with their children as a custodian (2008).

Land tenure for women in matrilineal communities

In Malawi's matrilineal communities land is inherited almost exclusively through the female lineage. Peters (1997) who conducted research in a matrilineal village in southern Malawi explained that young women work the field of their mother or grandmother and would receive a field when they married or had a child. When their matrikin passed away, they would inherit those fields. Contrastingly, fields were rarely inherited by males. Peters documented that land left to sons in this village caused conflict (Peters 1997). Based on his fieldwork in Malawi's matrilineal communities, Takane (2008) found that land was primarily inherited by females with land inherited by males as the exception. Takane argues that inheritance by males in a matrilineal matrilocal community may be an adaptive strategy to deal with land scarcity (2008).

Residency, and decision-making power affect matrilineal women's land use in Malawi. The husband may come and live in the wife's community (matrilocal), or less frequently the wife may go and live in the husband's village (patrilocal). Shawa (2002) describes that in matrilineal communities where a women is living in her husband's village, should the husband die the wife may be chased from the village (Shawa 2002 in Mutangadura 2004:11). On the other hand, where a husband comes to live with his wife in her matrilineal community, the inverse is true. Should the wife die, the husband will be chased from the village and will lose his land rights (Mutangadura 2004:11).

When the couple lives in a place that belongs to neither the husband's nor the wife's kinship group, this may reduce the firmness of women's inheritance rights as it is easier to dispossess a woman living in a neutral location than under matrilineal matrilocal residence where her kin people are close by (White & Christian Literature Association in Malawi 2002).

The scope of decision-making authority that matrilineal women have is a topic of debate. Mutangadura (2004) describes that in Malawi some women in matrilineal systems whose husband moves to their community ‘do not have full rights over the land because men especially the uncles control the land owned by women' (Mutangadura 2004:11). The uncle, known as the mwini mbumba, can be described as a caretaker of a group of sisters (Peters 2010). Academic Pauline Peters has conducted fieldwork in matrilineal uxorilocal communities in the Shire highlands of southern Malawi for over two decades (Peters 1997). In her study Peters (1997) found that matrilineal succession and inheritance, especially of land (in a very densely populated region), along with the uxorilocal residence, provide women with a considerable scope for authority and independent action in comparison with men, both as brothers and husbands. This ‘scope of authority’, a product of matrilineal descent and matrilocal residence as described by Peters (1997), contrasts with the assertion that gender relations in Malawi's customary areas have generally taken on the form of patriarchy (White & Christian Literature Association in Malawi 2002). In further work, Peter cites the matrilineal matrilocal experience as ‘proof of the viability of strong land rights for women' (2010:180).

Research context

In examining both matrilineal and patrilineal in 10 communities from the southern tip of Malawi to the northern town of Karonga, this research collected information representative of diverse geographic and cultural areas. However, there are some socioeconomic commonalities. For example, in all communities farming was identified as the main economic activity. While differentiation between subsistence livelihood farming and cash crop production was not ascertained, the importance of farming as the main livelihood is relevant in understanding the research context. In all research communities a high ratio of women to men was reported. Poverty, hunger and land shortages were common community challenges. Many of the matrilineal and patrilineal research communities also reported challenges accessing secondary and post-secondary education due to funding shortages and distance. In addition, in some communities transportation difficulties and challenges accessing clean water were reported.

Methods

Prior to the fieldwork, qualitative research tools were developed for the overarching research project including a community profile questionnaire, a focus group discussion guide and a key informant interview guide. Interviews and focus group discussions were chosen as the primary method of data collection as they allow the collection of data specific to the research foci while maintaining flexibility to explore new pathways of inquiry. This paper primarily reports the results of the community profiles and focus group discussions and does not contain the full results from the key informant interviews.

Research ethics approval was obtained from the University of New Brunswick Research Ethics Board. The field research took place in January and February of 2008 in cooperation with the Department of Land Management at Mzuzu University. In Malawi the research team consisted of a master's student from the University of New Brunswick (Canada), a lecturer at Mzuzu University and a research assistant who was an undergraduate student at Mzuzu University. The research assistant also acted as an interpreter.

Malawi has both patrilineal and matrilineal kinship groups. Patrilineal groups are primarily in the north and matrilineal in the south. The research team visited one patrilineal community in the south (Nsanje district) and five in the north (Rumphi and Karonga districts). Fieldwork also included visits to two matrilineal communities (Mulanje district), as well as two matrilineal communities in the Community-Based Rural Land Development Project (CBRLDP) (Mangochi).

Field sites were selected in consultation with the Department of Land Management on the basis of the factors of` kinship groups, known land shortage, location in a rural area and accessibility. As research was conducted in the rainy season, only those communities with passable roads could be included. Multiple strategies were used to gain access to the research communities. In most cases the research team contacted the local district office who then suggested Community-Based Organizations (CBOs) who in turn introduced the researcher to local communities. In some cases focus groups were organized by CBOs. In the resettlement communities, access was gained through the Community-Based Rural Land Development Program officers at the local district office.

In each community visited, the research team collected community profile information to provide contextual information to describe the research locale. Convenience sampling was used to identify a participant to complete the community profile questionnaire. Profile data were collected from a variety of individuals including chiefs, advisors to the village headmen, civic leaders and focus group participants. In examining the socio-economic conditions of each research community, the profiles were constructed using a standard questionnaire with nineteen criteria about social institutions, livelihoods, accessibility to basic infrastructure, demographics and community challenges.

Focus group discussions were held with women in each of the communities visited. Before each discussion, oral consent was obtained. Participants were asked to describe how women in their community accessed land and how this access is affected by HIV/AIDS. The focus group interview guide contained questions on the following themes: land rights, inheritance, land use, interventions to protect women's land rights and HIV/AIDS.

Focus group size fluctuated according to participant availability and ranged from 4 to 60 participants. In some cases it was difficult to elicit participation and in others it was culturally imprudent to limit participation. In focus groups with participation numbers under eight, women in the community often had other immediate responsibilities including tending the fields and attending funerals. The largest focus group with 60 participants was an anomaly. In this large focus group, it was challenging to elicit information from each participant as certain individuals dominated the discussion. In an effort to offset this bias, the moderator attempted to seek involvement with other focus group participants. During the fieldwork, a range of 10–20 participants proved to be optimal as there was sufficient range of experience and enough time to illicit information from each participant.

The research data were analyzed using content analysis. Content analysis is ‘a careful, detailed, systematic examination and interpretation of a particular body of material in an effort to identity patterns, themes, biases and meanings' (Neuendorf 2002 in Berg 2007:304). In preparation for the analysis, the interviews were transcribed verbatim by the researcher. The text was then manually coded into predetermined themes relevant to the research inquiry.

Results

How women access land

In the research communities studied, the methods through which women access land for housing and subsistence agriculture were not homogenous. Land is often accessed through one's relationships with the lineage that owns the land. Relationships include inheritance, marriage, as well as family and community membership. Land tenure varies with the nature of the relationship through which land is accessed. Inherited land has the strongest associated ‘rights’. Beyond accessing land through one's own social relations, land is sometimes accessed through an informal rental market.

Access in matrilineal research communities

Inheritance

In the matrilineal matrilocal research communities in Southern Malawi, land is inherited primarily through the female line. The exception is where a family has had only sons. In this case a male child can inherit. Women have a strong self-described sense of ownership. When asked about access to land for female-headed households, a focus group participant (Mulanje A) explained, ‘Because we own the land from our parents so it doesn't matter much. We get the land from our parents so when the husband dies, we are still staying on our own land’. This comment reflects the land tenure rights women have as inheritance beneficiaries. While male children are not likely to inherit land, it was indicated that if there was a surplus of plots available, then they would also benefit. This is interpreted as temporary user rights for males without the tenure security afforded by inheritance.

Marriage

Matrilineal focus group participants explained that men typically move to live on their wife's land. Residency was used to justify female inheritance as ‘the male ones will go and use their wive's land wherever they go’ (Mulanje A). One interviewee (Mulanje A) described that males do not inherit as they ‘are still taken as children. And they benefit there and they go’. It was explained that inherited land is passed down through the family and that ‘no one can come from outside of that family and claim ownership of that land’ (Mulanje A). Due to the prominent residency pattern of men moving to live in their wife's village, they are considered outsiders and as such are not privy to ownership of their in-laws land. Thus, women living in matrilineal communities rarely access customary land through marriage. Husbands typically access land through marriage with their wife.

Alternatives

In common with the patrilineal communities visited, in the matrilineal research communities people can be allocated land by a traditional leader. While none of the female participants reported taking this approach, they mentioned that if their parents wanted to move to another area, they would go and ask the village headman for land.

The role of the uncle in matrilineal communities

Within the data collected, there is a debate about the level of control an uncle asserts over land in matrilineal communities. Key informants working with NGOs argued that while on the surface it appears that women are in control over matrilineal property, it is actually the maternal uncle who controls the land. This perspective was contrasted by information collected in the matrilineal research communities as well as several key informants. In the research communities, female focus group participants argued that their uncles play a crucial role in making sure that land stays in the family of the wife, but have no say over how the land is used. They described the role of the uncle as twofold: firstly, to make sure that land stays within the clan and, secondly, to assist with relationship problems they might have with their husband. For example, if a woman wants to divorce her husband, she would first go to her uncle and explain the situation. The woman's uncle would then consult with the husband's uncle. Divorce among the Yao ethnic group was described as common. This is linked to the husband's tenure insecurity among the wife's lineage:

the notion of children belonging to the female lineage of the matrilineal is very important. And therefore the men are essentially seen as individuals that come to reproduce. Right? You help the woman to reproduce and of course you help the woman do farm work, right? If this person fails in this role, then he can go.

The uncle is described as a ‘mwini mbumba’, a man with the responsibility of protecting his sisters’ interests. A key informant, himself a mwini mbumba, explains that in this role he has to make sure that the land does not fall into the hands of strangers, but ‘doesn't control what goes on in the individual land holdings’. He explains that the woman's brother also plays an important role, while the uncle is:

Important to settle differences between sisters for example, or related people in the matrilineage … But at the level of the household the brother, the sister's brother is important … The uncle will be important where you are talking about the lineage as a whole. Right? So there's that type of division of labor, but in many cases that division is not so clear cut because the brother can also be the mwini mbumba. Right? The mwini mbumba simply means guardian of a group of sisters.

The uncle in the matrilineal context could be termed as an ‘administrator' of the lineage. He is described by one of the interviewees as an ‘institution' responsible for resolving conflicts over land ownership within the kinship group as well as household arguments. Information on the history of a clan's land boundaries is held with the uncle as he ‘is supposed to be a custodian of the household, he should know how this group came in, where the boundary was, what process was there when they were coming in’. The uncle uses oral history to argue against those encroaching on his clan's land.

In this research it was reported among matrilineal research communities that the ‘mwini mbumba’, a descent group male, keeps land within the kin and makes sure that the land is properly divided, while it is the husband who has a role in land-use decision-making. Furthermore, Schneider's (1961) concept that matrilineal systems have strong descent bonds and weak marital ones corresponds with the fieldwork data collected. In one relocated research community the researcher visited, women explained that they could easily get rid of their husband if he is not good. This is a stark contrast to the patrilineal communities visited where women explained that it is difficult to divorce their husband.

Access in resettlement communities

Both of the resettlement communities visited in Mangochi belong to matrilineal kinship groupings. As such, land is traditionally inherited by women. However, in this case as they are living on resettled land as part of the CBRLDP, both women and men are accessing land through the resettlement project. From a legal perspective, land is held by a ‘land trust' of which beneficiaries are members.

When asked if they would maintain the practice of female inheritance in this area, one group said that they would but the other asserted that ‘here the land belongs to the household' (Mangochi B). In this quote the household is taken to represent both male and female members. The idea that land belongs to the household diverges significantly from belonging to the lineage or the kinship group. As the groups are only newly relocated, it is too early to collect information on inheritance or land separation after divorce. Will there be a shift away from matrilineal inheritance structures? This is an important topic for future research.

Does the resettlement project give women better access to land? As it is targeted at the ‘land poor’, it should provide better access for those who have previously had to rent land. This was the experience for women in one of the focus groups. A key informant familiar with the project goes further in arguing that, ‘They have better access, because if I register my own right as a beneficiary of the project, the land is in my name, it is not in the name of the family’. Here the argument is that women will have more power to make decisions about land registered in their name. However, research data collected show that the project has more male than female registrants.

Access in patrilineal communities

Research participants from patrilineal communities indicated that women can access land through marriage, family, requesting land from the chief or renting. Land is inherited primarily by male children. In addition this section examines the importance of the lobola for women's land tenure.

Inheritance

In the patrilineal communities visited in Southern and Northern Malawi, land is primarily inherited by male children. The exception is where a husband and wife have only female children. In that case the land can be inherited by the girls, but must be passed on to males in the next generation. Inheritance may rely on the legitimacy of the marriage. For example, if a lobola was not paid, the male child from that union may not be able to inherit some of his father's land. Males who have inherited the land maintain strong tenure rights. They may ‘lend' land to a female relative to use but retain the ability to retrieve it should they need the specific parcel of land.

Marriage

As males most often inherit the land, women typically marry outside of their village and relocate to live on the husband's land. They gain access to this land ‘through' their husband. This relationship is dependent on the continued relationship with their husband as well as the nature of relations with the in-laws. Should the marriage end in divorce, the wife may be expected to return to her parent's village. A wife's land tenure after her husband's death is more complex. Ability to continue to access her husband's land is dependent on the wife's relations with her husband's family, as well as her decision to maintain widowhood. Respondents often answered that their capability to stay on their husband's land after his death depended on the ‘personalities' of their in-laws. Some send the widow home to her family, while others let her to stay on the husband's property. Furthermore, permission to remain on the husband's land is conditional on the wife remaining a widow and not remarrying. There are strong sentiments about keeping land within a family/kin system. Women who stay on their deceased husband's land cannot remarry as there is concern that the land would end up under the control of the new husband who is an outsider to the kinship group.

Many of the widows who participated in this research project reported that they had access to a portion of their husband's land. This means that they had been partially dispossessed of the land they shared with their husband. This indicates that there is a range of dispossession from being dispossessed of part of the land to being pushed off the land altogether. Partial dispossession of land user rights is equated with a change or renegotiation of ‘permission' to use the land.

Family

Women who are sent back to their village as the result of divorce or the death of their husband often access land through their family. They will access it through their parents or through their brothers if land has already been divided.

While women are given land to access, the size and tenure arrangements may not be favorable. Widows reported that they could not access enough land for cultivation, and were only given enough for housing. Another concern was that they might not be able to secure long-term user rights. Women using ‘family land' are not able to lay future claim to the land or to pass it on to their own children. One woman (Rumphi B) who was openly HIV positive explained how she had planted trees on a plot of her father's land where she had built a house for herself. She was attempting to lay claim by demarcating the area with trees. Her father told her to remove them, saying they would destroy the natural environment. She felt that he was rejecting her attempt to lay claim to this land.

Another participant (Rumphi B) who was staying on her mother's land noted that she had insecure tenure and was concerned that her uncle will try to get the land once her mother passes away. (Her mother had inherited the land as she did not have any brothers. The woman's uncle is the son of another one of her grandfather's wives). While women's user rights on customary land in patrilineal research communities varied, they were often seen as temporary. One focus group participant (Karonga C) explained,

Women have difficulties in accessing land. Like when someone is divorced when she goes back to her home she will be given a piece of land usually, but when the brothers get married they will grab that piece of land from her because it doesn't belong to her.

This reflects women's tenure rights to family land in patrilineal communities. They appear to lack robustness and fluctuate according to other familial demands on land resources.

In most of the data collected, women moved to live with their husbands. However, in two cases in Nsanje it was found that the husband moved to his wife's family land. In the first case the couple had relocated as a result of water supply problems in the husband's village. In the second case the husband was from Nkhota-kota district which is largely matrilineal and had decided to relocate to the wife's village. Regarding her husband's relocation she explained, ‘it depends, if the husband wants to go and stay on the woman's side there is no problem. But if he says no it means that the wife will have to go to his place’ (Nsanje).

Alternatives

Approaching the chief and renting land were two alternatives for women seeking to access land in the patrilineal research communities. Responses on the success of asking the chief for land were mixed among the interview participants. In two communities it was felt that this would be a worthwhile effort, while in two others it was not a viable option. In the particular focus group discussion (Karonga A) that thought this could be a possible course of action, one woman mentioned that ‘If she is a widow they go by themselves, but if she has an older boy child he has to go and ask’. The groups that thought the chiefs would not give them land explained that in the current state of land shortages there is not much land left to allocate. They explained that the chief would tell the widows and divorced women to ask their relatives for some land.

In the northern districts of Rumphi and Karonga, women indicated that they often have to rent land for housing and agricultural purposes. Married women, widows and divorcees were in this situation. Due to land shortages, they needed to rent land for subsistence and cash crop agriculture. Rental was often sought in addition to customary land they were already using. Rent is not always collected in money and in some cases women can exchange part of their produce for the use of the land. Others accumulate money through small informal enterprises such as selling donuts in order to make the money required for paying rent.

Lobola and women's land tenure

The ‘lobola' is a tradition that is practiced among the patrilineal communities in Malawi but is absent from the matrilineal setting. Data collected for this research project, shows that the term ‘lobola' has many different meanings. A ‘lobola' or ‘bride price' is a payment made to the bride's family by the groom. This practice is seen as authenticating a marriage and those who have not paid lobola are not recognized as having a valid marriage. In a marriage where the lobola has been paid, the children belong to the husband's lineage. Conversely,

When you have paid nothing, every children from where you have paid nothing … that woman is not going to belong to you … There is no respect, there is no nothing, it is just as if you are a friend in the bush. (NGO Karonga)

Key informants in northern Malawi argued that payment of the lobola is really about respect, but has been misconstrued by some as buying a wife. The lobola binds a marriage and makes it difficult for a wife to divorce her husband without reason. An NGO worker (Karonga) explains:

It is very expensive and it brings ties between marriages. There is not that kind of marriage that will break down … If the woman goes back to her home they will say, you go back we don't want you here because what we (were) paid so much that we cannot give back.

Raising the funds to repay a lobola is extremely difficult for a woman who is reliant on subsistence agriculture. Women said that it is painful to return the lobola and that those who do not have understanding parents will often stay in abusive marriages because there are no funds left to return. Women may be forced to repay the lobola if they divorce their husband or if they remarry after their husband's death. The amount to be repaid is negotiable and is dependent on the attitude of the husband and his family and the relationship between the wife and her in-laws. A woman in Karonga described her situation. She was chased from her husband's home because she could not bear children after 10 years of marriage. As a result she had to return half of the lobola. In another example, a woman who had an abusive marriage was allowed to leave without repaying the lobola. It is evident that there is some flexibility regarding repayment.

Just as the paying of the lobola validates a marriage, repayment signals the end of the union. A key informant (NGO Karonga) described that women who are widowed:

Can continue living at the same place because the lobola was paid. Now if she decides to remarry then she has to leave everything apart from maybe her personal belongings … . And on top of that when she remarries and they pay the lobola she will bring back everything that I paid to them.

Once a woman remarries she must leave all of the property in her husband's village and relocate. If she chooses to remarry, she may have to repay the lobola to her previous in-laws.

How does the lobola relate to women's dispossession of land user rights? It is not the lobola, but the absence of lobola which affects women's rights. Women, who have a legitimate marriage, where the lobola has been paid, have a greater chance of maintaining user rights to their husband's land. The importance of the lobola in maintaining land rights extends to the children, as those born out of legitimate marriage may not be seen as legitimate heirs to the land. An example was given by a participant in Rumphi:

Sometimes women can have children from different fathers which are not well known at the family there. So the children are raised by their mother. They can even grow without knowing their fathers. So when it comes to land it becomes difficult for them to access land. Because the uncles will be the ones telling them why we don't know your father, we haven't seen him, we don't know him, you don't belong here.

To summarize, women and children's likelihood of being dispossessed of their land-use rights greatly increases if a lobola was not paid. In patrilineal communities women's land-use rights are closely linked to the payment of a lobola.

Household land-use decision-making

While the purpose of this paper is to examine women's access to land, it is important to note that household land-use decision-making is an important factor in ensuring food security. Each focus group was asked how decisions are made about household land use. Often this question was answered with the phrase, ‘We sit down and discuss as a family’. While taken at face value, it appears that the husband and wife have equal decision-making power during these ‘family discussions’. However, after further inquiry it is evident that unequal power relations are often present. Research participants from both matrilineal and patrilineal communities indicated that the husband often has more decision-making power than the wife in deciding how land is used at the household level. Decisions are not always harmonious as land use may be a source of conflict. Household land use in the communities visited is a contentious issue due to the differing interests between the husband and the wife. It is in the interest of the woman to plant sufficient subsistence crops as she is responsible for food security, while it is in the husband's interest to grow cash crops as he is seen as the family's financial provider. This research found that land-use decision-making is made primarily by men in both matrilineal and patrilineal communities, but that in patrilineal research communities women used their responsibility for food security as a bargaining chip to increase their decision-making power. The strategy of using previous food insecurity to allot more land for subsistence agriculture is used. Women realize that they must have access land to fulfill their role of food security providers.

HIV/AIDS and women's land-use rights

In each of these focus group discussions, participants and key informants were asked the main research question, ‘How does HIV/AIDS affect women's access to customary land in rural Malawi?’ Their responses follow under three categories: namely, dispossession of land-use rights in patrilineal communities, reduced productivity and stigma.

Dispossession in matrilineal communities

In matrilineal communities as women are customarily the rightful inheritors of the land, they are not dispossessed of land access rights as a result of their own HIV status or their husband's death. No examples of women being dispossessed of their land were found in the research data. However in matrilineal research communities, it was reported that women are sometimes dispossessed of household property by their husband's relatives. This would include iron sheets and other components of the house. When asked how they thought household goods should be divided, women gave different answers. Responses included that they should have an equal share with the in-laws, that the wife should have three quarters of the property and, finally, that they would prefer to be left all of the property and to distribute it as they see fit. From one participant's perspective, ‘in most cases you find that the relatives of the husband would come and take almost everything leaving the widow with maybe only kitchen utensils … . So that is one of the major problems because they are left in poverty’. In one of the communities there is a human rights committee that deals specifically with this issue and it is reported that this had decreased the negative impact of household property grabbing at that site.

Dispossession in patrilineal communities

In patrilineal communities, HIV/AIDS can cause a woman to be dispossessed of her land ‘user rights' in several circumstances. When her husband has passed away she may be sent back to her village. Secondly, women who are found by their husband to be HIV positive may also be sent back. In patrilineal communities visited, there were reports that women are being dispossessed of their land-use rights. This occurred in some cases as a result of their positive HIV status, but was more often because a woman's husband had died. Here dispossession occurs when a woman is chased or pushed off her husband's land. In the research data collected, it is difficult to differentiate between physical dispossession and voluntary withdrawal from the husband's property. This is due to invisibility of the psychological tools used to encourage women to leave. Psychological tactics used to get widows to leave their husband's land include verbal abuse and fear of witchcraft. Tactics may be direct or indirect. One woman was told directly, your husband is dead so you can go back home. Another participant explained that women have limited ability to negotiate with their in-laws as they fear they will become victims of witchcraft.

Dispossessing the widow is rationalized by the fact that the social relationship which informs her membership in the community is broken. She is told, ‘you don't belong here' (Rumphi A). Her access rights were secured through her husband who is now deceased. When asked if one's likelihood of being dispossessed rests on how their husband died, most participants answered that ‘death is death’. Many do not know the HIV status of their deceased husbands. While you might be dispossessed whether your husband died in a car accident or by HIV/AIDS, the latter has an additional stigma. An example was given by a research participant where an HIV-positive woman returns to her family's land which is held by her brothers who say, ‘this one is already dead, why should we give land to her?' (Rumphi B). Stigma compounds the initial dispossession by providing challenges for future land access.

Women, who are permitted to stay on their husband's land, may be only allowed to use a portion of that which she had used while her husband was alive. This is rationalized by the reasoning that the husband's relatives should not give his widow all of his land as it will be more than she needs. This amounts to partial dispossession of the land they can access. Table 1 displays how it was reported that women access land in patrilineal communities, reasons they are dispossessed and the type of dispossession. In some instances women face a range of dispossession from partial to full.

Table 1. Dispossession of women's land-use rights in patrilineal research communities.

Relationship to point of access	Point of access	Reason for dispossession	Dispossession typology
Wife	Husband	Polygamy	Partial
		Wife inheritance	Partial
		Divorce	Full
		HIV/AIDS stigma	Full
		Cannot bear children	Full
Widow	Husband's relatives	Husband's death	Partial/full
		Refusing wife inheritance	Full
		Widow remarries	Full
		HIV/AIDS stigma	Full
Sister	Brother	Land needed for another use	Partial/full

Research data collected from different sources at the same location were sometimes contradictory. In one patrilineal community, male village leaders and female focus group participants were interviewed separately. Both groups were asked what would happen if a husband was HIV negative and found out his wife was positive. The male leaders said that the husband would keep his wife. In contrast, the women explained that the husband would chase his wife away for fear that she would transmit the disease to him. The women answered, ‘She is saying most they would chase them. A few would leave their wife home' (Rumphi A).

While dispossession occurs at the family or kinship group level, it was indicated that a chief or district commissioner would intervene if they were notified. The chief's mandate to intervene may be limited. Research participants explained,

It's just within the clan of the husband, like others will say she should go … but the chief himself can maybe just intervene and ask why you are chasing this one? … the chief can only say, you can go and discuss as a family. (Rumphi B)

It is possible that the relatives of the deceased may not accept the authority of the chief or the District Commissioner over the deceased's property. Allocated land is seen as belonging to the clan or kinship group and is not directly under the authority of the traditional authorities.

One focus group discussed the psychological impact that getting kicked out of a village has on a woman. Research participants described, ‘They complain, so they will think a lot what to do. Because maybe when they go there they won't find a land so sometimes they will die earlier because of the thinking' (Rumphi A). Accessing land for women who have returned to their parent's village may be stressful. This psychological stress stems from dispossession, and the shortage of land in one's natal village.

Reduced productivity

HIV/AIDS reduces a woman's ability to access land in both matrilineal and patrilineal communities as a result of reduced productivity. Women who are infected or affected both face challenges maintaining high levels of agricultural productivity. Those with advanced AIDS have reduced energy to work in the fields. Likewise, those who have a family member with HIV/AIDS or another illness will have to reduce the time spent in the fields as women are traditionally ‘caregivers' in Malawian society. This could impact women's future access to land as fields left barren for several years may risk being grabbed. A key informant from the matrilineal areas argued that this type of property grabbing of barren land has been avoided as a woman usually lends land to her sisters if she is too sick to tend it. This practice protects land from lying fallow and being consequently grabbed.

Stigma

Stigma decreases women's ability to access customary land by assigning HIV-positive women a decreased value. In some cases women may be sent home by their husband as a result of their positive status. Data collected from research communities in northern Malawi found that positive HIV/AIDS status was linked to the perception that the individual was ‘already dead’. By labeling an HIV-positive woman as a ‘dead person’, it is rationalized that they are less deserving of resources. Such resources include land. An NGO informant gave the example that if an HIV-positive woman goes to the village Chief to ask for land, he may not give it saying, ‘Why do you need land if you are already dead?' Being stigmatized as a ‘dead person' also affects an individual's participation in social activities. One research participant in Rumphi district was not permitted to work on any church committees as she was thought to be already deceased. At the family level, when this woman was found to be positive, her father went out and purchased timber to build her coffin. That was 1986. Today she has outlived her father but still faces challenges due to stigma.

In rural communities HIV/AIDS stigma comes from misconceptions about transmission as well as moral issues surrounding sex. In some cases stigma appears to be decreasing, an NGO worker in northern Malawi explains,

We used to view those with HIV as those with moral decay. So these people are prostitutes. We used to point fingers at HIV as being transmitted maybe by prostitutes, female prostitutes. Maybe just females in general … But as of now, there is that understanding that HIV/AIDS is for everyone, everyone can catch HIV/AIDS … . But as for stigma it is still out there where we are working.

Key informants describe that HIV/AIDS-related stigma is on the decline and that people have become more open about their status as programs are developed to assist them. Awareness of HIV/AIDS-related stigma had reached the community level and one local leader described his own role to discourage discrimination and to make sure that individuals are not separated from the group because of their positive HIV status. Peters, Kambewa and Walker (2010) describe media statements as inducing public stigmatization of HIV/AIDS linked to ‘promiscuity' and associated fatality. The above examples from this fieldwork suggest that HIV/AIDS stigma and associated public narratives are being challenged at the village level.

Discussion

Intersection of HIV/AIDS, customary tenure and land shortages

In the context of HIV/AIDS, this research has shown that land tenure for women living in Malawi's customary areas is determined primarily by the kinship group they belong to. The impact of HIV/AIDS on women's access to land was unequal with women in patrilineal communities fairing worse than those living in matrilineal areas.

HIV/AIDS affects women's access to land in communities where females access land through relationships with male members who are the rightful land heirs. HIV/AIDS may cause the termination of these relationships due to stigma or death of the access individual(s). Termination of such relationships increases tenure insecurity for women accessing land in a community where they do not have inheritance rights.

HIV/AIDS has caused difficulties for women accessing land in patrilineal communities in two significant ways. Firstly, by causing death HIV/AIDS creates difficulties for women who access land ‘through' a relationship. Women's land tenure becomes vulnerable when their husband dies. Their ‘access point' no longer exists. Participants argued that ‘death is death’, meaning that it is not important how the man died. This research shows that while the cause of death is not often known, HIV/AIDS was reported by NGOs to be a major cause of death in rural Malawi. Here it is not HIV/AIDS that directly causes the dispossession of land-use rights; rather the disease kills the person through whom women access land. Women face increased vulnerability to being dispossessed from their land rights should their husband die of HIV/AIDS as the woman's in-laws will have multiple reasons to dispossess her (widowhood as well as HIV/AIDS stigma).

A second way that HIV/AIDS affects women's access to land is through stigma. The research shows that some women had been divorced by their husbands as a result of such stigma. Of the two methods through which HIV/AIDS affects women's access to land in patrilineal communities, the first is more serious. While participants reported that civic education is changing local perceptions about HIV/AIDS stigma, death of an ‘access point' is more severe as it relates to culturally embedded traditions of land ownership. Women accessing land ‘through' a relationship with another person is a direct result of male inheritance. As a potential intervention point, projects to reduce HIV/AIDS stigma may be easier to implement than would be changes to inheritance practices which are entrenched in cultural tradition.

For women in patrilineal communities, HIV/AIDS is a stressor on an already stretched tenure system. Land shortages are another factor that affects women's access to land in rural Malawi. Resource shortages put cultural systems under stress as there is not enough available land to allocate to all those in need. All of the research communities visited complained of land shortages. Furthermore, women in focus groups reported high numbers of children. Population is growing and yet the economy in the research communities visited remains largely agrarian. Research participants cited renting land as an option to gain access to land given current shortages. Takane (2008) has also documented land rental in patrilineal communities.

While a limitation is that this research project was not designed specifically to address land shortage, information collected suggests that this factor is impacting cultural traditions. This finding is congruent with the literature that ‘poverty and scarcity of resources tax the goodwill of family members' (Joireman 2008:1240). While Joireman (2008) is referring to the goodwill of the family of the deceased husband, a widow may also have to rely on the goodwill of her natal family. For example, in patrilineal communities where women are sent back to their family's villages after the death of their husband, there is little available land left for them. While theoretically, this system of women accessing land through their family members would work under optimum circumstances, it is being put under stress by resource shortages. Land shortage also affects women's ability to access land through traditional authorities, as chiefs may have nothing to allot.

Why the matrilineal experience matters

In Southern Africa, the experience in regard to land tenure for women living in matrilineal matrilocal communities is very different from that for women living in patrilineal patrilocal communities. This research demonstrates that HIV/AIDS differentially impacts women in matrilineal and patrilineal communities. The value of the matrilineal experience documented by this research is that it provides us with an example of an exception to the dominant patrilineal system. This exception is valuable for several reasons. First, it suggests that HIV/AIDS has a lesser impact on women's access to land in customary land tenure systems where women have robust rights. Second, it raises the point that matrilineal is something to consider when considering policy approaches to improve women's land rights. For example, broad approaches like equal inheritance laws may not be appropriate in countries where matrilineal tenure systems exist. This paper supports the suggestion of Peters and Kambewa (2007) that equal inheritance will dispossess women of land in matrilineal matrilocal areas.

Conclusion

This paper supports and furthers the existing literature that suggests that HIV/AIDS, gendered inequalities embedded in customary inheritance practices and resource shortages combine to affect women's access to land in patrilineal communities. In contrast to the patrilineal patrilocal experience, research from the matrilineal matrilocal communities demonstrates that where women are the inheritors of the land and have robust land tenure rights, they are not at risk of losing their access to land due to HIV/AIDS.

This research finding is transferable to other countries in southern Africa that experience land shortages and have both matrilineal and patrilineal tenure systems. It is important to consider the matrilineal exception in developing policies to ensure women's access to land.

In conclusion, the strength of this study is the finding that HIV/AIDS is not an independent factor that impacts women's access to land, but one that acts synergistically with customary practices and resource shortages. For women living in patrilineal communities with land shortages, HIV/AIDS is a stressor on an already stretched system. As land is a primary resource for women's livelihoods and women in patrilineal patrilocal areas have challenges gaining access, there is a strong rationale for future research that examines alternative livelihood schemes that are not dependent on land for women who have been dispossessed due to HIV/AIDS.

Acknowledgements

Thank you to Professor Ian Methven, Professor John McEvoy and Mrs Veronica McGinn at the University of New Brunswick. Thank you also to the Department of Land Management at Mzuzu University and Ms Glory Munthali our talented research assistant. This work was funded by the Canadian Federation of University Women Charlottetown Chapter and the Social Sciences Humanities Research Council of Canada.