«The Role of Gender and Kinship Structure in Household Decision-Making for Agriculture and Tree Planting in Malawi Seline S. Meijera,c,*, Gudeta W. ...»
A lot of studies have tested these models and the assumptions associated with them for rural households in developing countries. There has been increasing evidence against the common preferences model and the unified household model. Although the concept of a “unitary household” is convenient, the empirical evidence to support these simplistic models is scarce (Strauss and Thomas, 1995). Alderman et al. (1995) suggest that there was sufficient evidence against the unitary model of the household. However, the unitary model is not always rejected completely (Doss, 2013). For example, Quisumbing and Maluccio (2003) test the unitary versus collective model of the household for Bangladesh, Ethiopia, Indonesia, and South Africa and reject the unitary model as a description of household behavior, but fail to reject the hypothesis that households are Pareto-efficient. On the other hand, Udry (1996) tests the concept of Pareto efficiency for farming households in Burkina Faso and finds that plots controlled by women have significantly lower yields than similar plots controlled by men, thus contradicting the Pareto efficiency of resource allocation within the household.
There is a large body of literature on the models of household decision-making, and some of the findings have been inconsistent with each other and have encouraged the further development of new and alternative models of intra-household resource allocation and decision-making. The recently developed Women’s Empowerment in Agriculture Index (WEAI) is a new tool that measures the empowerment, agency, and inclusion of women in the agricultural sector and can serve as a diagnostic tool to signal key areas for interventions to increase empowerment and gender parity (Alkire et al., 2013). Alkire et al. (2013) document the development of the WEAI and present findings of a pilot in Bangladesh, Guatemala, and Uganda. Although the authors caution that the results are not representative of the whole countries, the study finds that in Bangladesh women are empowered in 43.2 percent of households sampled (WEAI = 0.762), compared to 27.3 percent in Guatemala (WEAI = 0.702) and 41.2 percent in Uganda (WEAI = 0.800). In Malawi, the WEIA score is 0.84 and nearly 52 percent of women have achieved adequate empowerment, making Malawi perform better than Bangladesh, Guatemala, and Uganda (Malapit et al., 2014). However, the study was conducted on the boundary of the Central and Southern regions of the country, and ideally information would be collected across all three regions to understand differences within the country. The WEAI is a promising new tool; larger surveys in more contexts can help identify key decision-makers in different types of production.
Levels of involvement in decision-making can range from no involvement to sole decisionmaking, with various levels of input falling between. For example, Nosheen et al. (2008) look at men and women’s participation levels in different agricultural practices using three categories of involvement: Never, sometimes, or often. This paper employs a binary approach to assess decision-making by husbands and wives, as this approach is the simplest to understand and explain to respondents.
MEIJER ET AL -58Journal of Gender, Agriculture and Food Security Vol. 1, Issue 1, pp 54-76, 2015 Kinship Structure Two marriage structures exist in Malawi, with associated rules for inheritance and residence (the structures are referred to here as “matrilineal” or “patrilineal”). In a matrilineal society, land is transferred along matrilineal lines, which usually involves inheritance going from the wife to her daughters or nieces. In addition, it is common for the married couple to take up residence in the wife’s village, which is referred to as uxorilocal residence (Takane, 2008).
The husband will cultivate land together with his wife, but he has no decision-making power over the transfer of his wife’s land rights. In case of divorce or death of the wife, the husband loses the user rights over his wife’s land and is expected to return to his original village, leaving the children with the wife or her family (as they belong to the matrilineal kin). In patrilineal societies, on the other hand, land rights belong to men and are usually transferred from fathers to sons. The residence rules are virilocal and a wife lives in her husband’s village after marriage (Takane, 2008). Upon divorce, a woman must return to her original village while the children remain in the husband’s village. A widow may sometimes remain in the husband’s village if bride wealth was paid and if the relatives of the husband give permission for her to do so.
Studies have observed changes in the inheritance and residence rules in different regions of Malawi over the past few decades (Hansen et al., 2005; Vaughan, 1985). For example, Phiri (1983) described how influences—such as the intrusion of patrilineal immigrants, Christian missionary activities, colonialism, and the modern capitalist economy—have affected matrilineal societies in central Malawi since the mid-nineteenth century and found that uxorilocality in particular has diminished. Similarly, Takane (2008) studied customary land tenure and inheritance rules in diverse regions of Malawi and concluded that, despite the fact that most of the land transactions followed customary land tenure and inheritance rules, land transactions differed from the basic rules in a good number of cases. Reasons behind these deviations include the unique personal relationships between landholders and heirs, the increase of wives returning to patrilineal villages after divorce or widowhood, and the increasing scarcity of land (Takane, 2008).
In a detailed account of changes in gender relations in Malawian households, Vaughan (1985) describes the effects of commodity production on gender relations within rural households in southern Malawi during the colonial period, using two case studies with similar matrilineal inheritance rules and uxorilocal marriage. The first case study describes the effects of the collapse of the cotton industry in the 1930s—in which men and women played an equal role—which was then replaced by labor migration and cattle-raising. As these new activities were restricted to able-bodied men, this resulted in the marginalization of women in the economy (Vaughan, 1985). The second case study focused on the tenants of a privately-owned estate where residents had to pay a local rent (thangata) to be allowed to stay on the land. In the 1930s, the system was modified, and instead of paying rent in the form of labor, men paid in the form of tobacco grown on their land. The men now had a direct interest in land allocation and land rights slowly became invested in men; and, as a result, women lost their bargaining power (Vaughan, 1985). “In both cases, the ultimate decline in the position of women came about in part through the particular interventions of the state in molding the nature of economic relations. By placing fiscal responsibility on men, for instance, the state made central the economic activities of male household members and placed greater emphasis on the household (and hence on marriage) as the
basic economic unit of society. The price women paid for their exemption from taxation (and, in the case of the estates, their exemption from thangata) was the ultimate marginalization of their economic activities” (Vaughan, 1985).
In contrast, others have argued that the matrilineal societies in Malawi have been remarkably stable in the face of external change and pressure to conform to patrilineal rules (Peters, 1997; Peters, 2010). Peters (2010) examined the matrilineal land tenure in southern Malawi and found that traditional matrilineal inheritance rules have prevailed despite a long and continuing history of prejudice against matriliny and attempts by the government to discourage matrilineal inheritance (Mkandawire, 1983; Hansen et al., 2005). Changes in gender relations and matrilineal and patrilineal residence and inheritance traditions will affect household decision-making, and hence it is important to take note of these studies.
Methods Biophysical Context and Selection of Study Areas Malawi is a small landlocked country in southern Africa, occupying an area of 11.9 million hectares, of which 22 percent is comprised of inland waters (lakes Malawi, Malombe, Chilwa, and Chiuta). The climate is tropical and rainfall is concentrated in a single wet season between November and April, with average rainfall varying from 800 mm in the lowlying areas along the Lake to 1,000 to 1,500 mm in the high-altitude plateaus. Almost all households involved in farming cultivate maize, making it the most important staple food.
Other important food crops are pulses, groundnuts, and cassava. In addition, cash crops grown for export include tobacco, tea, sugar, coffee, and macadamia. The population of Malawi was estimated to be 14.9 million in 2010, with an average annual growth rate of 3.1 percent (World Bank, 2013). The population is concentrated in the south of the country, where the population density is 184 persons per square kilometer, compared to 63 in the Northern Region (NSO, 2008).
This study focused on two study sites in Malawi: the northern district Mzimba and the southern district Chiradzulu. Mzimba district is characterized by relatively low population densities and is inhabited mainly by the Tumbuka and Ngoni ethnic groups. The population density in Chiradzulu is relatively high, and the predominant ethnic groups are the Chewa, Lomwe, and Yao. In both districts, 10 villages were randomly selected from a list of villages provided by the extension planning area (EPA). In Mzimba, 10 villages were selected in Zombwe EPA: Chinombo Jere, Kenani Shaba, Maquiko Mbizi, Yesaya Juba, Simoni Tembo, Samani Mkandawire, Mathambo Mtete, Samuel Jere, Chabwa, and Palango Mhango. In Chiradzulu, 10 villages were selected in Mbulumbuzi EPA: Luna, Nchenao, Lumeta, Nsungwi, Jonathan 1, Chiwinja, Makawa, Nyasa, Sasu, and Mbunda.
In Malawi, matrilineal succession is mainly practiced by the major ethnic groups found in the central and southern regions—such as the Chewa, Yao, and Lomwe—whereas patrilineal kinship structure is mostly associated with the Tumbuka and Ngoni ethnic groups in the north (Place and Otsuka, 2001; Takane, 2008). In matrilineal households, either the husband or the wife can be considered the household head, whereas in patrilineal households it will generally be the husband.
MEIJER ET AL -60Journal of Gender, Agriculture and Food Security Vol. 1, Issue 1, pp 54-76, 2015 Household Survey A household survey was used to elicit information on intra-household decision-making. The survey contained questions on personal and household characteristics, farming activities, and decision-making roles within the household. For 11 agricultural activities—including tree planting and tree management—the respondent was asked if the main decision-maker in their household was the husband, the wife, or if the decisions were made jointly by the husband and wife together. The same question was asked for the implementation of these 11 activities.
The questionnaire was administered between October and November 2012 to 135 married household heads who were randomly selected from the lists of village inhabitants provided by the local EPA. In Chiradzulu, the household survey was administered to 41 male-headed households and 27 female-headed households, whereas 65 male-headed households and 2 female-headed households were included in the survey in Mzimba. This study reports the findings of the households comprised of a husband and wife, and it excluded household heads who were single, separated, widowed, or in a polygamous marriage. In all cases, the head of the household was interviewed, as identified by the household after selection. If the head of the household was not available to be interviewed, another household was selected from the list using a random sampling procedure. To complement the data collected in the household survey, handheld GPS units were used to measure the area of the land belonging to each respondent to establish land size and calculate the density of trees planted on a respondent’s land.
Focus Group Discussions Focus group discussions were carried out according to the methodology described by Hennink (2007). The groups were stratified according to district (Chiradzulu and Mzimba) and gender (male and female). In both districts, separate groups were formed of male and female respondents. In each district, four focus group discussions were carried out with female participants and four with male participants, resulting in a total of 16 focus groups.
After the villages had been selected, respondents were selected randomly from the list of farming households provided by the EPA. Each focus group discussion consisted of seven to nine participants. A discussion guide was developed and translated into both Chichewa and Tumbuka and the focus group discussions were conducted in the local language of each district. The group discussions lasted approximately two hours. The focus group discussions included a group exercise on household decision-making in relation to agricultural activities.
The participants in the focus groups were asked to discuss which household members are generally responsible for the decision-making for 11 agriculturally-related household activities. For each activity, the participants discussed who the most common decisionmaker for the activity was, and consequently which household members were involved in the implementation of the decision. The focus group discussions were carried out in April 2013.
Data Analysis Respondents were divided into two kinship groups based on their ethnicity. Respondents from the Chewa, Lomwe, and Yao ethnic groups were classified as belonging to the
matrilineal kinship group, and the other ethnicities to the patrilineal kinship group.
Descriptive statistics were used to describe the characteristics of the sample population.