The Interface Between HIV/AIDS Status, Household Nutrition, Agricultural Production & Household Welfare in Uganda
BY DAWN C. PARKER WITH MACTION KOMWA
Although HIV/AIDS has no boundaries, the most affected region is sub-Saharan Africa, where 25 of the 40 million people globally living with the virus live. The epidemic has eroded the ability of rural African households to produce food and other agricultural products, generate income, and care for and feed family members. Due to high adult mortality, a substantial loss of adult labor and specialized knowledge is occurring. Labor losses come not only from reduced productivity during prolonged illness of the Person With Aids (PWA), but also from time spent caretaking and mourning by survivors. However, little information is available on magnitudes of labor losses, changes in cropping patterns, and changes in agricultural production resulting from HIV/AIDS infection. Consensus is emerging around one point however— heterogeneous local conditions, including relative scarcity of production inputs, crops grown, growing conditions, gender roles and household financial resources will interact to determine actual impacts. There is no one simple answer to the question of how HIV/AIDS will affect the varied aspects of agricultural production in sub-Saharan Africa.
A team lead by Mason’s Drs. Parker and Jacobsen, in collaboration with doctoral candidate Maction Komwa (supported by a Bourlag LEAP fellowship), the International Food Policy Research Institute (Washington, DC), Thomas Berger and Pepijn Schreinemachers (University of Hohenheim, Germany), is conducting research on relationships between HIV/AIDS, agricultural production, food security, and livelihoods in agricultural households in Uganda. Project activities include qualitative and quantitative household surveys and multi-agent systems simulation modeling. We explore the following questions:
- Can labor shortages at the household level be mitigated through interactions with other households and communities?
- How do land rights and land ownership dynamics associated with HIV/AIDS affect the welfare of individual households and the distribution of wealth between households?
- What effects does the presence in the household of a PWA have on the farming, non-farm and self-employment activities and education of the caretakers of the infected individual?
- What are the effects of an individual living with HIV/AIDS on the health and nutritional status of the person with AIDS and other members of HIV/AIDS affected rural households? Specifically, what feedbacks exist between nutritional status, health status under HIV, agricultural production and household welfare?
In 2006, Komwa conducted qualitative, semi-structured stakeholder interviews at both district and local levels in Kampala and two villages of Mayuge district. The fieldwork provided a rich body of data to understand how members of the household cope with the social and economic burden posed by a household member infected with HIV through the progression of AIDS to death. Results from the fieldwork (summarized here) mirrored results from other similar studies, but provide additional detail in some areas, and even contradictions in others, leading to new testable hypotheses.
The Household: Overall, our general results are quite similar to other studies, with a reported increase in female, elder, and child headed households. This often resulted in new, higher responsibility roles for more vulnerable family members (young household heads and female children), who often left school, entered the labor market, and even migrated to ensure survival of themselves and their households. It is also clear that a crisis of orphan care is developing. A change in family structure to smaller, more independent families was described in our study area, an effect not seen elsewhere in sub-Saharan Africa.
As widely reported elsewhere, our results detail that caretaking responsibilities fall most heavily on women and girls. These responsibilities can absorb the majority of the caretaker’s time as illness progresses, representing a substantial loss of labor for the household. Our work-hour estimates demonstrate that women consistently work a higher number of hours than men regardless of the activity. Labor is also lost due to the PWA’s lowered ability to work. We found, however, that often these labor reductions were a voluntary effort to preserve health status, based on the advice of medical professionals. Voluntary work reductions may not be possible for poorer households, potentially leading to a downward spiral of declining health and increased poverty.
Consistent with many other studies, we found that women and girls are highly vulnerable to loss of land, household capital, and in some cases their home within a given community. Widowed households without a male child to inherit the land are most vulnerable. While the traditional land tenure system serves to maintain control over scarce land resources within the extended family unit (clan), this traditional system substantially increases the vulnerability of female-headed households. New community organizations have developed to protect the inheritance rights of women, and the success of these organizations will likely influence the well being of affected households and potentially the distribution of land holdings within the community.
As have other authors, we find that expenses and loss of labor from HIV have a severe negative effect on household welfare, with effects most pronounced in poorer households that lacked resources even before HIV infection. Effects are seen both during illness, as households cannot afford drugs, other medical care, and appropriate foods, and also after a death, as they may receive less help with funeral expenses.
In contrast to other studies, we found that consumption of highnutrient foods was highly emphasized, with purchase of these foods second in priority only to anti-retroviral drugs (ARVs). Both governmental and non-governmental organizations were reported to be providing food and medicine, but access to both seemed uneven. Our results provided qualitative details regarding in-kind contributions from relatives. We found a mixed picture, with assistance being uncertain and potentially declining over time. Again, as in the descriptions of increasing burdens of care for orphans and the resulting increase in the number of independent nuclear households, we see evidence of a support network stretched to its limits and of households needing to put a high priority on meeting their own needs rather than supporting more distant relatives.
The Farm: Our study finds a mixed picture in terms of cropping changes. As in other studies, shifts to crops with lower labor requirements and a decline in household agricultural production over the last five years are reported. However, this shift did not always occur subsistence crops. Female-headed households may be more likely to plant subsistence or food crops, but this could be due to cultural preferences or higher weight placed on household subsistence requirements, rather than lower labor availability. In contrast to other studies, our respondents reported intensification strategies that often included switches to higher-valued market outputs, such as vanilla, aloe vera, and poultry production. These strategic crop changes appeared to result from lower labor availability and, in some cases, from a subsequent reduction in cultivated area. Consistent with other studies, reported cropping changes depended on household resources, with households with sufficient labor and financial capital less likely to change cropping patterns.
Multi-Agent System Modeling: To advance our quantitative understanding of changes in labor availability, work organization, land transactions and agricultural production due to HIV/AIDS and HIV-related mortality, we are constructing computational models— quantitative multi-agent systems capable of capturing biophysical and socioeconomic complexity and spatially integrating different models of farm household functional behavior—to explore the effects of reduced labor and income on household agricultural practices. MAS models combine traditional microeconomic decision models at the household level with structured interaction environments at the community level in order to model inter-household dynamics in a simulation environment.
Our models extend earlier MAS modeling work from the University of Hohenheim, conducted for the same study area. Their model incorporates generic representations of socio-economic factors, labor management and loss, land transactions, land market activities, and cropping patterns of the smallholder farmers, integrating biophysical crop yields and soil dynamics models with mathematical-programming based economic models that simulate decision-making and poverty levels. Real-world farm households are represented analogously as agents in the model. Three objectives guide agent land-use decisions: cash income from selling farm produce and off-farm employment; in-kind income from consuming own farm produce; and future income from investments in livestock and coffee.
Modified rules for the MAS, specifically focused on changes in available resources and decision making for both HIV-affected and non-affected households, have been created based on our qualitative fieldwork. Modified rules focus on four factors: household labor availability (accounting for both PWA and caretaker labor loss); increased household expenses from medical care, high-nutrient foods and funeral expenses; and changes in land transactions (both land sales and inheritance based on kinship networks). Our modified model will allow us to conduct scenario analysis to explore the potential effect on household welfare of policy interventions such as subsidies for medicine and food for households with PWAs, strategies to protect the inheritance and land tenure for widows and orphans, and agricultural subsidies. Because our modeling framework incorporates interactions between households, effects of community level constraints on land and labor availability can be explored, in contrast to household level approaches.
Endogenizing Nutrition and Heath: Adequate nutrition is essential to maintain the effectiveness of ARVs, which may allow PWAs to maintain productive lives. The MAS model is being extended to endogenize linkages between household nutrition, health status and labor productivity for PWAs. A household survey is being conducted by Komwa in an expanded study area in summer 2007. It includes detailed questions on health status, food consumption, household labor and agricultural production. This tremendous data resource will allow us to model dynamic interactions between decreased resources (labor and wealth) for households with PWAs; decreased agricultural productivity; and decreased food consumption, household nutrition and health status— which may again reduce household labor availability. We hope that our models will help identify households most vulnerable to being caught in this downward spiral of decreasing health and household welfare and to identify policy interventions most effective at intervening in the downward spiral.
Dawn C. Parker (dparker3@gmu.edu) is assistant professor in the department of Computational Social Science (http://mason.gmu.edu/~dparker3/). This article was first published in print and citations have been removed due to space limitations, but are available from the author.
