Chapter 5: Conclusions and recommendations

Selling vegetables and dried fish at a market in Northern Zambia

This final chapter draws together the main conclusions from the case studies, discusses follow-up initiatives under the IP, and makes general recommendations on how to respond to the epidemic, such as mainstreaming HIV/AIDS, developing multi-sectoral responses and introducing social protection.

The case studies illustrate the gap between the epidemic’s impacts on rural people’s lives and the policy responses that take place at the national level, which include agriculture, gender, HIV/AIDS and the legal protection of property. Clearly, more needs to be done to close this gap and foster an enabling policy environment. The challenge is clear. How can countries support increasing numbers of vulnerable households? What can be done to reverse the trend towards increasing destitution? What can be done to maintain the agricultural and rural sectors intact? In order to achieve a holistic and consolidated approach to solving these issues, a comprehensive and collaborative response that effectively mitigates the impacts of the HIV/AIDS epidemic on rural livelihoods needs to be implemented at the macro, meso and micro levels (Figure 11).

Figure 11: Interdisciplinary responses to the epidemic

A complex picture: conclusions of case studies

Agriculture provides a livelihood for a vast majority of the rural population through small-scale agricultural activities that are short-term, seasonal and poorly rewarding. HIV/AIDS-related illness and death bring additional costs associated with decreased household labour and increasing health care expenditure. These lessen the already meagre return to households, lead to a slow depletion of asset-based wealth and fundamentally affect long-term food security.

The three case studies demonstrate that the impacts of the HIV/AIDS epidemic go beyond rural agricultural production. Under conditions of increasing AIDS-induced poverty, access to resources becomes increasingly important; inequalities in access to inherited property, for example, further marginalize AIDS-affected households. This marginalization means that many households are not “picked-up” by national processes of reform, such as the Plan for the Modernization of Agriculture.

A complex picture is beginning to emerge and demands a comprehensive assessment of the nature of the HIV/AIDS epidemic. The stage and pattern of the epidemic within a country need to be examined. An epidemic that has stabilized, such as in Uganda, will have different impacts and demand different responses from one that is spreading rapidly (e.g. Namibia or Zambia). Other factors influencing poverty also need to be taken into account, as the impacts of the HIV/AIDS epidemic emerge over a long period of time and interact with other processes that can increase vulnerability.

FAO recognizes that HIV/AIDS is interwoven with many other determinants of food insecurity and poverty (Figure 12) (FAO, 2003). The current food-related crisis in Southern Africa, for example, is the culmination of a number of factors including policy mistakes, mismanagement and poor governance, inappropriate market reforms, a lack of extension and other support services for farmers, the removal of consumer protection, and political instability, in addition to unfavourable weather conditions and HIV/AIDS (Morris, 2002). Gender roles and relations influence the course and impact of the HIV/AIDS epidemic. Gender-related factors shape the extent to which men, women, boys and girls are vulnerable to HIV infection, the ways in which AIDS affects them, and the kinds of responses that are feasible in different communities and societies (UNGASS, 2001).

Figure 12: The context of HIV/AIDS’ impact on livelihood strategies and outcomes

A household’s wealth and livelihood strategies will largely determine how well it reacts to external shocks, such as drought, floods and fluctuations in market prices. HIV/AIDS contributes to the vulnerability of rural households by weakening their asset base. In addition to wealth, social mechanisms, such as gender inequality, affect access to resources and services, all of which affect livelihood outcomes. The complexity of the situation makes it difficult to measure the impact that is attributable to HIV/AIDS. A United Nations Regional Inter-Agency Coordination Support Office (RIACSO) consultation was held recently to address this issue and to exchange ideas on measuring vulnerability in the light of the HIV/AIDS epidemic. A number of organizations are currently working on developing more sensitive indicators to be used at the national and household levels, which will assist in designing a more effective response.¹²

IP: interdisciplinary and agricultural sector responses

National stakeholder workshops reviewed the results of the baseline surveys and identified a range of response strategies. Feedback meetings at the district and community levels, involving local government, civil society organizations, faith-based organizations, traditional leaders and extension workers, gave further direction to priority areas. These participatory processes ensure the representation of local needs and ownership of the products.

IP stakeholders identified a wide range of possible interventions to respond to the HIV/AIDS epidemic. These include the promotion of labour-saving technologies, improving nutrition, encouraging labour pooling arrangements, reinforcing community-based mechanisms to preserve local knowledge, and diversifying livelihoods. A few of these were selected as pilot activities to be implemented under the IP - different activities were selected for each of the three countries. In Uganda follow-up activities focus on supporting the mainstreaming of HIV/AIDS into the agricultural extension services. In Namibia and, to a lesser extent, Zambia, resources are still directed towards bringing the epidemic under control. As a consequence, the IP activities in these countries focus on improving household food security and nutrition and preventing property grabbing. These are discussed in more detail below.

Extension workers’ role in addressing HIV/AIDS

As HIV/AIDS becomes more prevalent in communities, the traditional household structure changes. The predominance of male-headed households gives way to one in which many households are headed by single parents, elderly grandparents, older children and the physically weak. Thus, the clientele for agricultural extension changes from predominantly male farmers to female, young and elderly farmers. At the same time, extension organizations and their staffs are being affected by increased AIDS-related attrition (Qamar, 2001). The challenge for agricultural extension strategies is to target new clients with decreasing human resources. One strategy could focus on developing the awareness of existing extension workers and building their capacity, so that they are better equipped to deal with the complexities that HIV/AIDS brings. This would require a revision of the existing extension training curriculum to reflect better HIV/AIDS’ impacts and possible responses related to agricultural production and food security.

In Uganda, the IP is developing an HIV/AIDS resource guide for agricultural extension and community workers and conducting capacity building workshops at the district level.

The guide can be used during training and as reference material while working with communities. It provides:

a clear picture of the dynamics of HIV/AIDS in different agricultural production systems;
practical skills for assessing and analysing the nature of HIV/AIDS’ impact on vulnerable groups;
practical guidelines for working with communities to develop appropriate response strategies for HIV/AIDS-affected households, for example, related to nutrition, labour-saving technologies, community mobilization;
useful resources and ideas that can be adapted.

Extension workers in Northern Namibia

HIV/AIDS and agricultural messages

Messages about HIV/AIDS focus on control and prevention of the disease. To date, little information has been developed on the links between HIV/AIDS and agriculture. An IP study, conducted in Uganda in 2001 found that the most effective vehicles for disseminating messages are the radio and the community worker, partly owing to the accessibility of these media and the preference for hearing messages by “word of mouth” (IP, MAAIF, 2001). Extension workers also considered delivering messages as part of their professional role.

In Zambia and Uganda, the IP is designing and delivering pilot communication campaigns. Appropriate multimedia messages will be designed to suggest practical ways in which communities can address some of the impacts of HIV/AIDS on agricultural production. Messages are designed by local communities and facilitated by a multi-sectoral team. Specific groups are targeted and messages are communicated through a variety of media, from radio to community theatre, and by extension workers using visual materials.

Improving household food security and nutrition

Household food security is an essential component of any HIV/AIDS strategy. A well-balanced diet should be an integral part of the care of people living with HIV/AIDS. Actions to promote better nutrition through on-farm processing and the utilization of indigenous crops should be encouraged. Home gardening should also be promoted as it can assist in providing a wider variety of foodstuffs. In addition, school gardening can play a role in ensuring that agricultural skills and knowledge are passed on to youth. Schools can also teach the importance of healthy eating.

In Zambia and Namibia, local ministries of health and agriculture are collaborating on activities to build extension and home-based care workers’ capacity to support the food security needs of households affected by HIV/AIDS. The manual Living well with HIV/AIDS (FAO, WHO, 2002) and local recipe books are being adapted to the local context and are providing the basis for inter-active training and guidelines. Appropriate messages are also being developed and disseminated by a wide range of media.

Protecting the property of the most vulnerable

It has become increasingly important to design or revise existing laws and to enforce laws that ensure women are accorded full and equal rights to own land and other property, including the right to inheritance. In some countries, legislation has been passed providing women with equal inheritance rights to land when their husbands die. While this is an important legal precedent, the enforcement of the law over local customary practices is equally critical. In this regard, the capacity of local officials needs to be supported so that they are able to negotiate this delicate process.

In Namibia, the IP is collaborating with the Ministry of Women’s Affairs and Child Welfare and the Legal Assistance Centre to implement a pilot project that addresses asset stripping in Ohangwena Province. The centre has several years of experience in the legal litigation regarding property, and the ministry is committed to enforcing the Married Persons Equality Act. This project is sensitizing and training local leaders and ministry staff, and para-legal training in the use of appropriate materials on the rights of women, orphans and vulnerable children is being provided for community-based support workers. Training material on property stripping has been adapted and translated into local languages. A national advocacy campaign on mitigating property stripping will be launched prior to a national conference in 2004.

Grandmother with her grandchildren in Zambia

Beyond the IP at the meso and macro levels

General lessons learned from the IP are related to two processes that have received much attention recently: mainstreaming HIV/AIDS and developing multi-sectoral responses. The rest of this section will look briefly at these two responses and at another mechanism - social protection - which is designed specifically to support vulnerable groups.

Multi-sectoral collaboration

Multi-sectoral collaboration has become a prerequisite for the effective addressing of HIV/AIDS. Many of the global initiatives to address HIV/AIDS require countries to have multi-sectoral national AIDS committees. Implementing multi-sectoral activities is a challenging process, and some of the key elements for success are outlined in the box below. Country HIV/AIDS programmes used to be based in the Ministry of Health, whereas the new multi-sectoral committees often have a different institutional home. This creates the risk of duplicating activities if good coordination is not maintained. Moreover, multi-sectoral initiatives may exist at the national level, but not at the level of implementation.

Multi-sectoral collaboration

Multi-sectoral collaboration is actively promoted as a strategy for addressing HIV/AIDS comprehensively. For multi-sectoral initiatives to work in practice, the IP identified the following elements as essential (IP, Mueller and Wiegers, 2002):

a clear vision and clarification of the primary roles and responsibilities of different sectors;
clear benefits from participation;
sufficient resources and adequate time;
support from high-level management;
agreement on the institutional home for multi-sectoral initiatives.

Mainstreaming HIV/AIDS

Mainstreaming is designed to create an enabling policy and resource environment. Mainstreaming should ensure that HIV/AIDS priorities become an integral part of the development agenda and that the response to HIV/AIDS becomes institutionalized. HIV/AIDS should feature prominently in the resource allocations of the Ministry of Finance and/or Planning and should be linked to activities and targets (UNDP, 2002). In the mainstreaming process, policies are reviewed to assess how sensitive they are to the impacts of the epidemic, to review existing strategies and to design additional activities if necessary. If policies and programmes already address the needs of vulnerable groups, then mainstreaming merely reorients them.

Many countries have “mainstreamed” gender and HIV/AIDS-related concerns into policies, strategies and programmes, but there are still gaps in implementation. This is partly because mainstreaming demands capacity building and active engagement over a period of time. An additional failure of mainstreaming is that the meso level, which implements strategies and policies, is often overlooked. Sector-wide approaches (SWAps) are structures for implementing mainstreamed policies and reforms, but few countries have agricultural SWAps. In Zambia, health, education, road transport and (to a lesser degree) agriculture are currently leading the SWAps (DFID, 2003), and many lessons could be learned from this process. Furthermore, HIV/AIDS itself is affecting the human resource capacity of governments and the ability to implement reforms and to decentralize activities effectively.

Mainstreaming steps

These steps can be followed for the mainstreaming of gender, HIV/AIDS or the concerns of vulnerable groups.

Step 1: Ensure that the impacts of the HIV/AIDS epidemic on rural livelihoods and poverty are thoroughly understood.

Step 2: Identify the internal impacts of the HIV/AIDS epidemic on the organization (e.g. ministry and/or departments).

Step 3: Identify how HIV/AIDS will affect the key goals of the organization.

Step 4: Identify how the organization can play a key role in reducing the spread of the epidemic. This step could include building capacity to review existing policies and programmes in order to identify what elements need to be revised.

Step 5: Identify how the organization can assist in mitigating the impact of the epidemic on rural communities. This step could include building capacity so that additional sector-specific activities that directly support those most affected by the HIV/AIDS epidemic can be designed and/or scaled-up.

Step 6: Identify monitoring and evaluation tools. For example, this could include activities ensuring that all current and future policies and programmes pass a gender and HIV/AIDS sensitivity test.

Adapted from: Health Economics and HIV/AIDS Research Division. 1999. AIDS toolkits. Durban, South Africa, University of Natal. HIV-AIDS-STI Knowledge Programme, HEARD. 2003. HIV/AIDS mainstreaming: a definition, some experiences and strategies. United Kingdom, Liverpool School of Tropical Medicine, DFID.

Reaching the poorest - social protection

The question remains as to whether policies are designed in order to respond to the needs of the rural poor. They may provide guidelines and objectives, but they are often not operational frameworks. If, as is highlighted above, multi-sectoral and mainstreamed operational frameworks are also not running smoothly, the outcome for poor people will be bleak.

A number of countries have recognized the limitation of sector policies in addressing poverty, and have designed sector-specific activities to target the most vulnerable. Social protection is a broad concept, which can be loosely defined as strategies that are not primarily intended to promote economic growth but to alleviate poverty or reduce vulnerability (DFID, 2000), such as pensions, employment schemes, “safety-net” types of interventions (e.g. input vouchers, food price subsidies, supplementary feeding).

In addition to public forms of social protection, the private sector may be an efficient mechanism for putting the poor on the agenda. Not only is the private sector affected by the HIV/AIDS epidemic, it also often: has greater skills and resources than governments; can mobilize capital; has good access to workers, customers and supply networks; has a vested interest in a healthy workforce; and, perhaps most important, has substantial influence over government policy.

Examples of social protection in the IP-countries

Zambia has a public welfare scheme that provides health clinic vouchers, school bursaries and limited food distribution in a number of districts. Zambia also has a social investment fund (ZAMSIF), which is a decentralized coordination mechanism that can channel resources to the community level. A community investment fund finances small projects that are identified, planned, implemented, managed, operated and maintained by communities, and a district investment fund strengthens local government capacity and accountability.^[11]

Namibia has an Orphans’ Emergency Fund, which was established in 1998 under the Ministry of Health and Social Services and the Ministry of Women’s Affairs and Child Welfare. This is implemented with the assistance of a national non-governmental organization (NGO), Catholic AIDS Action. The fund is meant to serve as a “stop-gap” measure for orphans who are in the process of applying for government assistance, and is designed to meet the expenses of school and examination fees, food or outstanding debts. Namibia’s pension scheme also helps to lessen poverty among the elderly and households that include elderly Namibians.

In Uganda, resources from the Poverty Action Fund are channelled to the district and village levels to support members of households affected by the epidemic through income-generating activities, nutritional support, training and improved access to schooling for orphans (UNDP, 2002).

Future steps

The Gender and Population Division (SDW) serves as FAO’s focal point for gender, HIV/AIDS and population ageing. SDW promotes the mainstreaming of these concerns into the activities of the Organization, and acts as an important catalyst in the follow-up to these innovative IP initiatives, in collaboration with partners inside and outside FAO.

Follow-up activities will be related to:

Advocating and mainstreaming HIV/AIDS concerns, targeting policy-makers and the public. The IP has produced well-documented cases on how to tackle and alleviate some of the negative impacts of HIV/AIDS on people’s livelihoods, agricultural production and food security. These approaches can be used when FAO is providing its Member Nations with policy advice and technical assistance to develop agricultural sector strategies that counter the consequences of the epidemic and prevent its further spread.

Replicating the pilot activities in other countries of Southern and East Africa, and obtaining the financial and human resources to continue developing the response strategies identified. One such example is the Improving Rural Livelihoods of HIV/AIDS-Affected Households in Northern Province, Zambia project, which was funded by the Government of Ireland and implemented through the FAO office in Zambia, with technical backstopping from SDW. The Zambian Government and FAO headquarters have also developed a Technical Cooperation Programme proposal to build institutional capacity in mitigating HIV/AIDS’ impact on the agricultural sector. Similar processes are taking place in Namibia and Uganda.

Continuing action-oriented research. The IP, SDW and other programmes in FAO have learned that the research required to address issues of HIV/AIDS’ impacts is complex.¹⁵ As a response, SDW organized a workshop in Uganda in collaboration with the Association for Strengthening Agricultural Research in Eastern and Central Africa (ASARECA). The workshop emphasized gender-disaggregated data (GDD) and social research methods in the context of HIV/AIDS and gender. The IP methodology has been updated and refined for the household livelihood research related to the project in northern Zambia. For example, the refined sampling strategy includes household types that are at different stages of the AIDS epidemic, such as female- and male-headed households taking care of people living with AIDS (PLWA), female- and male-headed households taking care of orphans, and non-affected households as a control group.

Disseminating information and experiences to create awareness about the linkages among HIV/AIDS, gender inequality, food security and agriculture. This will help to sensitize and influence policy-makers and development workers in FAO and partner countries on the importance of interdisciplinary collaboration for addressing the complex problems of HIV/AIDS and rural development. It will also help to raise awareness of IP experiences among a wider audience. Information will be communicated through various channels: on the FAO and IP Web site (www.fao.org/sd/ip/), in reports and articles and at relevant national and international events, meeting and conferences.

References

Barnett, T. & Whiteside, A. 2002. AIDS in the 21st Century: disease and globalization.

CARE. 2001. The impact of HIV/AIDS on agriculture production systems and rural livelihoods in the central region of Malawi.

CEDAW. 1997. Concluding observations: Namibia. UN Document A/52/38/Rev. 1, Part II paras 69-131.

CHR. 2003. 55th meeting. Resolution 22. Commission on Human Rights (CHR).

CSW. 1998. Report of the forty-second session (2-13 March 1998). Resolution 42/1 Human rights and land rights. Commission on the Status of Women (CSW).

DFID. 2003. Key Sheet Social Protection. Department for International Development (DFID). www.keysheets.org.

DFID. 2003. Mainstreaming HIV/AIDS across DFID Zambia Programme. DFID Resource Centre for Sexual and Reproductive Health.

FAO, Engh. 2000. HIV/AIDS in Namibia: the impact on the livestock sector.

FAO. 2002. HIV/AIDS, food security and rural livelihoods: fact sheet.

FAO. 2003. Food security and HIV/AIDS: an update. Committee on World Food Security. 29th Session.

FAO, DCI, Wiegers. 2003. Improving rural livelihoods of HIV/AIDS-affected households in Northern Province in Zambia. Preliminary findings.

FAO/OXFAM-GB. 2003. Report of Workshop on Women’s Land Rights in Southern and Eastern Africa.

FAO, SDW. 2003. Gender, HIV/AIDS and Food Security. Presentation at ECOSOC Ministerial Round Table Meeting on Gender and Rural Development. New York, 30 April 2003.

FAO/SDWP & Stokes, S. 2003. Measuring impacts of HIV/AIDS on rural livelihoods and food security.

FAO, WHO. 2002. Living well with HIV/AIDS - A manual on the nutritional care and support of people living with HIV/AIDS. Rome, FAO Food and Nutrition Division

GoU. 2000. Poverty Eradication Plan. Revised vol. 1. Government of Uganda (GoU) Ministry of Finance Planning and Economic Development

GoZ. 2002. Poverty Reduction Strategy Paper. Government of Zambia (GoZ).

HRW. 2003. Double standards: women’s property rights violated in Kenya, Vol. 15, No. 5. Human Rights Watch (HRW).

IP, MAAIF. 2001. The impact of HIV/AIDS on agricultural production and mainstreaming HIV/AIDS messages into agricultural extension in Uganda.

IP, Mueller, Wiegers. 2002. Building and managing cross-sectoral collaboration for sustainable development and food security: lessons learned from the IP Programme.

MoH. 2002. National HIV/AIDS/STI/TB Policy in Zambia (draft). Zambian Ministry of Health (MoH)

Morris, T. 2002. Special Envoy of the Secretary-General for Humanitarian Needs in Southern Africa: Report of the First Mission to Lesotho, Malawi, Mozambique, Swaziland, Zimbabwe and Zambia, 24 September 2002.

NAADS. 2001. Annual report 2001/2.

NACOP/MOHSS. 2002. Presentation by NACOP. National Coordinator. Ugandan National AIDS Coordination Programme (NACOP) and the Ministry of Health and Social Services (MOHSS).

OAU. 1981. African Charter on Human and People’s Rights. Organization of African Unity (OAU).

Okuonai, S., Karamagi, H. & Kyomuhendo, S. 2003. Evaluation of HIV/AIDS policies in Uganda. Uganda AIDS Commission.

ORMT. 2002. Report from Ohangwena Regional Management Team.

PMA, Government of Uganda. 2002. Plan for Modernization of Agriculture: eradicating poverty in Uganda. Government strategy and operational framework. Ministry of Agriculture, Animal Industry and Fisheries, Ministry of Finance Planning and Economic Development.

Qamar, K. 2001. The HIV/AIDS epidemic: An unusual challenge to agricultural extension services in sub-Saharan Africa. Journal of Agricultural Education and Extension, 8 (1).

RoN. 1999. The National Strategic Plan on HIV/AIDS. Medium-Term Plan II. 1999-2004. Republic of Namibia (RoN).

RoN. 2002. 2001 Census Preliminary Results.

TNDP. 2002. Transitional National Development Plan 2002-2005. Zambia

UNAIDS. 2002. National response brief to HIV/AIDS. Namibia, Uganda and Zambia.

UNAIDS. 2003. Accelerating action against AIDS in Africa.

UNAIDS, UNICEF, WHO. 2002. Epidemiological fact sheets on HIV/AIDS and sexually transmitted infections. Update.

UNDP. 2002. HIV/AIDS and poverty reduction strategies. Policy note.

UNGASS. 2001. UN Special Session on HIV/AIDS. Gender fact sheet.

UoC. 2003. HIV/AIDS in Zambia. University of California (UoC).

White, J. & Robinson, E. 2000. HIV/AIDS and rural livelihoods in sub-Saharan Africa. Natural Resources Institute.

World Bank. 2000. Social Risk Management: a new conceptual framework for social protection and beyond. Paper No. 0006.

Annex I: Summary of IP survey methodology

ACTIVITY (July 2002_April 2003)	DESCRIPTION
Research team	National research teams were set up. They consisted of experts from various disciplines and backgrounds so that they were well equipped to cover such a complex topic as the impacts of HIV/AIDS on agricultural production, food security and rural livelihoods.
Secondary data review	Desk studies were undertaken in each country to help identify data gaps and existing mechanisms that households and communities use in response to the epidemic and to refine the research methodology. National and international reports and information on the impacts of HIV/AIDS on agriculture and food security and the interlinkages with gender/youth were reviewed. Comprehensive annotated bibliographies were compiled in each country.
Survey design	Following secondary data review, the research teams, national-level IP stakeholders and officers from FAO headquarters agreed on the focus of the survey and developed a research framework. The research adopted a sustainable livelihoods approach as a useful tool for understanding the impact of HIV/AIDS on household assets and the various responses adopted by different households. Households are regarded as possessing different sets of livelihood assets that are essential to their livelihood strategies. These are human capital, natural capital, financial capital, social capital and physical capital (FAO/SDWP and Stokes, 2002). Data domains: 1. demographic profile (household register including age, sex, education level, school dropout, orphan status, occupation, relation to household head and in/out-migration) 2. agricultural production (changes in total landholdings, amount of land cultivated, cropping patterns, land preparation methods and ownership of farm equipment over the last five years) 3. livestock production (changes in animal types, breeds and numbers over the last five years) 4. food security and nutrition (food availability, food self-sufficiency, eating patterns and access to food support) 5. health (illness, death and means to cover illness/death-related expenses) 6. labour (changes in intra-household labour allocation for productive and domestic tasks, use of hired labour and labour availability over the last five years) 7. socio-economic characteristics (changes in income/expenditure patterns and asset ownership) 8. agricultural knowledge 9. existing response strategies to HIV/AIDS effects on household assets 10. rural institutions (collapse of existing and emergence of new institutions, responses of existing institutions to the HIV/AIDS pandemic) 11. gender roles and decision-making Data collection strategy: Data was collected through both a quantitative survey (the household level) and a qualitative investigation (the household and community levels). A comprehensive quantitative questionnaire covered the first seven data domains and included a retrospective component with a five-year recall period. Data were disaggregated by gender and household type (see sampling procedure). Qualitative methods involved participatory discussions at the community level to capture the general development context and data domains 8 and 10. Focus group discussions and individual household interviews were conducted to capture in-depth information related to labour constraints, gender roles and decision-making, changes in asset ownership, inheritance and existing response strategies. Various socio-economic and gender analysis (SEAGA) and participatory rural appraisal (PRA) tools were adapted, such as historical time lines, Venn diagrams, village resource maps, wealth rankings, gender activity clocks, seasonal calendars, problem analysis charts and an income and expenditure matrices. Focus group discussions were conducted separately for women and men.
Sampling procedure	Sampling was done in two stages: Stage 1: Study sites were purposively selected. In Uganda, three districts (Luweero, Masaka, Iganga) were selected to reflect the different livelihood options (crops, livestock and fishing) in the Lake Victoria Crescent agro-ecological zone and different HIV/AIDS prevalence rates (Masaka, Iganga and Luweero represent high, medium and low prevalence, respectively). In Namibia, three districts (Engela, Eenhana, Okongo) were selected based on different levels of HIV/AIDS prevalence (low, medium, high) and represent a cross-section of the main health districts in Ohangwena region. The reason for selecting Ohangwena region was that it is the poorest region in Namibia with a large share of the population involved in subsistence agriculture and it has one of the highest prevalence rates of HIV/AIDS in Namibia. In Zambia, Choma, Monze and Sinazongwe districts in Southern Province were selected owing to their relatively high HIV/AIDS prevalence levels and their importance in terms of agricultural production. In each of the districts within the three countries, two communities were selected - bringing the total sites to six per country. Stage 2: Households were stratified and then randomly selected from the different strata in each study site. In Uganda, households were stratified according to affected and non-affected status and were further stratified by whether the household was male-headed, female-headed or orphan-headed. An operational definition of affected households includes households in which at least one family member has been lost to HIV/AIDS or HIV/AIDS-related chronic illnesses (TB and pneumonia), or households in which at least one family member was suffering from frequent or long illness due to, or related to, HIV/AIDS during the recall period (1997-2002). Non-affected households are households in which no member has died of, or is living with, HIV/AIDS-related diseases. It was agreed that a higher probability of selection (0.7) should be given to the affected households and a lower probability (0.3) to the non-affected ones in order to give more relevance to impact. As in Uganda, in Namibia households were stratified according to affected status and gender with random sampling within the different strata. In Zambia, the Standard Enumeration Areas (SEAs) of the Central Statistics Office (CSO) were used. CSO has divided Zambia’s districts into Census Supervisory Areas, each of which is subdivided into SEAs. For each SEA, preliminary information on the number of households and the population is known, and this was used for selecting enumeration clusters. Households within these clusters were stratified by gender, i.e. female- or male-headed. Survey households were then randomly selected from the list of male- and female-headed households. The final sample included 766 rural households in Zambia, 513 in Namibia and 610 in Uganda.
Data collection	Prior to conducting the actual fieldwork, the draft questionnaire was pre-tested in six to ten households, and local enumerators were trained in administrating the questionnaires and in the participatory tools used for the qualitative research. The quantitative household questionnaire was administrated in six communities over a period of one month per country, followed by qualitative data collection in the selected communities.
Data management and analysis	Tabulation plans were drafted and the data were entered and cleaned. A data management workshop was organized in Namibia in October 2002 with representatives from all three IP countries. One statistician from Zambia and one from Penn State University provided technical backstopping throughout the process. Data management was done in similar fashions in each country and basic analysis involved frequencies, cross-tabulations and descriptive statistics. Zambia: The data were entered using the Integrated Microcomputer Processing System (IMPS). The resulting ASCII data files were converted into Statistical Package for Social Sciences (SPSS) for data cleaning and subsequent analysis. The final baseline survey data set was saved and archived in both SPSS and MS Access formats for future use. Namibia: Data entry, cleaning and analysis were all done using SPSS software. Uganda: Epidemiological Information System (EPIINFO) was used for data capture and SPSS for data analysis.
Lessons learned	Focus on what aspects to study via quantitative and qualitative methods during the design phase of the study Sampling problems associated with identification of “affected” households. It is important to note that use of the CSO Standard Enumeration Areas (SEA) in Zambia resulted in a low incidence of “affected” households in the sample. This, and the respondents’ low response due to HIV/AIDS stigma, which probably accounts for the very low proportion of the sample being assigned to the “affected” categories, points to the need to devote considerable effort to constructing an adequate sampling frame. Attention is needed to construct adequate sampling frames for selecting “affected” and “non-affected” households. An integrated research methodology that sequences qualitative and quantitative research methods (QL-QN-QL) is needed, i.e. there is a need to precede quantitative surveys dealing with HIV/AIDS’ impacts with qualitative investigations to help determine what should be investigated and how to characterize the quantitative variables. Analytical/interpretive problems related to the a priori assignment of households to categories that may not be mutually exclusive (e.g., female- and widow-headed households). Analytical/interpretive problems of attributing observed effects on households to HIV/AIDS. Follow-up research is needed into inequality of property ownership among the different households types, particularly in relation to affected male-headed and male-headed households with orphans having relatively more assets.
	Please contact the IP ([email protected]) for the following documents: Questionnaires (Uganda, Namibia and Zambia); The qualitative checklist; Enumerator instruction manual (Uganda); and Annotated bibliographies.

^[14] The World Bank's Multi-Sectoral AIDS Programme (MAP) has a similar philosophy: www.worldbank.org/afr/aids/map.htm.