It is commonly agreed that HIV and AIDS have contributed to the depth of problems faced by rural
households in southern Africa in the context of the 2002 food emergency. What is much less understood is
the extent of that contribution and how it varies according to the demographic structure, mortality and
morbidity profile of households. The purpose of this study is to help fill this information gap and to further our
understanding of the impact of HIV/AIDS on acute food insecurity in southern Africa. Data generated from
emergency food security assessments conducted in Malawi and Zambia in August and December 2002 and
from Zimbabwe in August 2002 were used to study the relationship between HIV/AIDS proxy variables and
food security parameters.
Although the HIV/AIDS pandemic is of global concern, it is in Africa where the effects of the disease are most
acutely felt. Of all global HIV infections, roughly 70% are located in Africa, where an estimated 28.5 million
people live with HIV/AIDS (UNAIDS, 2002). The disease is now responsible for more annual deaths in Africa
than any other cause. The southern Africa sub-region, in particular bears a disproportionate burden of
HIV/AIDS cases. It is here that the world’s highest rates of HIV infection are to be found: in a number of
southern African counties, the adult prevalence rates are over 30% (Botswana, Lesotho, Swaziland, and
Zimbabwe (UNAIDS, 2002).
Southern Africa is currently experiencing the worst food security emergency in a decade. Each of the three
countries in this report has been receiving large amounts of international food aid and other humanitarian
assistance since mid-2002. Since December 2002, at least 25% of their entire population have required
food assistance. In Zimbabwe the food security crisis is particularly severe, with over half of the country’s
population requiring assistance. Current indications are that food aid will again be required in parts of the
region, particularly in Zimbabwe, during the coming consumption year of April 2003 to March 2004 (SADCFANR
It is now well recognised that household food insecurity in rural and urban southern Africa cannot be properly
understood if HIV/AIDS is not factored into the analysis. Carolyn Baylies (2002) notes that HIV/AIDS can, on
one hand, be treated in its own right as a shock to household food security, but on the other, it has such
distinct effects that it is a shock like none other. Livelihoods-based analysis of linkages between food
security and HIV/AIDS show that the impact is systemic, affecting all aspects of rural livelihoods (Haddad
and Gillespie, 2001); and that effective analysis of the causes and outcomes of HIV/AIDS requires a
contextual understanding of livelihoods unique to a given area and/or social groups (FEG, 2000).
The results presented in this report clearly indicate that households affected by adult morbidity, mortality and
with a high demographic load are significantly more vulnerable to food security shocks than are other
households. Insofar as these indicators suggest the presence of HIV and/or AIDS, this analysis strongly
implies that HIV/AIDS has significantly increased the vulnerability of households to acute food insecurity in
2002-03. The analysis has shown that these households suffer from marked reductions in agricultural
production and income generation, leading to earlier engagement in distress coping strategies, and,
ultimately, a decline in food security. The cumulative impacts of HIV/AIDS on food availability, food access,
and coping capacity are compounded, resulting in amplified negative impacts on overall household food
security. The analysis further demonstrates that different morbidity, mortality and demographic profiles have
different effects on food security processes and outcomes. Key differences are seen according to whether or
not the household has an active adult present or a chronically ill person, whether the head of household is
chronically ill, whether there is a high dependency ratio, or whether the household has taken in orphaned
children. Each of these characteristics has further nuances that are affected by age and gender.
In addition, there are differences between wealth groups in the extent to which proxy indicators affect food
security processes and outcomes. There is some evidence that the presence of proxy indicators has a
significantly greater impact on poorer households than they do on better-off households, although this is not
always the case (vis. the effects of chronic illness or death in adults on incomes). In this way, already
vulnerable households become even more vulnerable with the affects of HIV/AIDS.
This study suggests that the impacts of HIV/AIDS on food security in the context of the 2002 food emergency
are strong and negative. It also suggests that these impacts are complex and require urgent and innovative
responses in the 200-04 marketing year and beyond. The critical question for programming, policy,
advocacy and research is: what can be done to prevent, slow or even reverse a downward spiralling
livelihood trajectory for HIV/AIDS affected households?
Implications for Programming:
A “three-pronged attack” is suggested: consumption side support;
productivity enhancing support; and support to household and community safety nets.
Consumption Side Support: Chronic illness of the head of household and elderly headed households, in
particular those headed by women, are indicators that might be used in conjunction with food aid targeting.
If these indicators are to be used it is very important that they be cross-checked with wealth group analysis.
Taken in isolation from wealth status they may not be robust indicators of vulnerability.
Due to the decreased mobility of households affected by HIV/AIDS, special efforts will need to be made to
reach them. Simply distributing food at a central distribution point may not be enough. Agencies will need to
consider how they can work with communities to ensure that HIV/AIDS affected households receive their
quota. This may involve provision of transport and/or increasing the number of distribution points.
School feeding programmes have the combined benefit of ensuring that more children consume a healthy
meal at least once per day, as well as reducing the dropout rate. Keeping children in school has the obvious
long-term benefit of promoting education and empowerment, as well as reducing idle/unsupervised time of
children, especially young girls, who can become vulnerable to exploitation and thus increased exposure to
Awareness of the decline in quantity and quality of labour should be an integral part of the programme
design in areas with high HIV/AIDS prevalence. Food for work programmes, for example, should be
designed such that the type of labour opportunity is consistent with the capacity of the elderly and/or adults
that are not at their peak health. Even more appropriate would be the design of food for assets programmes
that are not labour dependent, but are oriented to skill development or awareness campaigns that are
accessible and appropriate for the elderly and children.
Given the fundamental decline in income and agricultural production experienced by HIV/AIDS affected
households, the analysis supports the need for continued consumption-oriented assistance to these
households in the form of safety net programmes, even after the immediate emergency has passed.
HIV/AIDS affected households will take longer to “recover” from a shock, and may never fully do so.
Accessing food will continue to be a foremost and formidable challenge of HIV/AIDS affected households
long after a crisis subsides.
Productivity Enhancing Support:
This needs to be purposefully tailored to household types and individuals
with a focus on improving productivity of HIV/AIDS affected households in general, with a special focus on
households headed by elderly women and those with a chronically ill household head. Existing productivity
enhancing interventions should be adapted to make them relevant for and accessible to both adults (over
and under the age of 60), and to children.
Rapid introduction of interventions with a high food access to labour ratio is recommended. These may include
on-farm food production oriented, on-farm cash generating and off-farm cash generating interventions.
Household and Community Safety Nets: In order to strengthen household level safety nets, micro-enterprise
service provision should be stepped up to help HIV/AIDS affected households to strengthen their economic
resources. However, microfinance is not a panacea for mitigating the economic impact of HIV/AIDS or for
alleviating poverty. It works best in areas which are well served by markets, and not in remote communities.
Also, the financial service does not create the economic opportunity, the client does. Different types of
service are required for households at different stages of HIV/AIDS impact. In relation to the household
types investigated in this report, a tentative conclusion is that, all other things being equal, micro-finance
appears most appropriate for households headed by HIV negative couples aged between 18 and 59 which
have taken in older orphaned children; for households headed by elderly HIV negative couples which have
taken in older orphaned children; and for households headed by HIV positive adults aged between 18 an 59,
where the infected adults are in the asymptotic phase of infection (which can last 3-5 years), with or without
When households do not want to or are unable to repay debt, and when households are in remote areas or
areas where income is highly unpredictable, savings schemes or savings led credit initiatives may be more
appropriate interventions than micro-finance. Whilst they may be more appropriate than offering credit,
managing them can be more complicated. The analysis in this report tentatively indicates that (all other
things being equal) this type of intervention may be particularly appropriate for HIV negative households
headed by elderly males (no spouse) with no other adults, but with children; and for households with a
chronically ill adult who is not head of household.
The economic stress caused by HIV/AIDS can become so severe upon a household that engaging or
continuing income generation activities no longer becomes an option. At this point, the community and
extended family’s role becomes critical. In order to create a source of funds that are sustainable over the
long term, communities will need to embark on an on-going resource mobilization campaign to identify and
mobilize internal resources first and then tap into external resources. There are essentially two aspects to
community resource mobilization campaigns: (i) strengthening social capital through tools that strengthen
community participation, awareness and empowerment and (ii) fundraising activities. From the data
analysed in this report, all other things being equal, the following household types are more likely to be
reliant on community support than others: households in which both the head and the spouse are chronically
sick; households in which the head is chronically sick; and households headed by single elderly females
In many if not most cases, optimal support to safety net activity at community level will have both individual
household and community elements. The targeting and blending of these support elements will depend on
the stage that households and communities have reached in terms of HIV/AIDS impact.
Implications for Policy:
A key element of response will be to view policy through an “HIV/AIDS lens”. The
“lens” essentially refers to an approach to viewing potential solutions to a problem (such as poverty) that
derives from evolving knowledge of the important linkages with another problem (such as HIV/AIDS)
(Gillespie et al, 2001). It also means that polices need to be designed, evaluated and implemented with a
view to their impact on HIV/AIDS prevention or mitigation – in this way, HIV/AIDS can be “mainstreamed” in
the policy process. The HIV/AIDS lens needs to be applied to all policies which have an impact on the lives
of rural people.
Implications for Advocacy:
The role of HIV/AIDS in the 2002 southern African food emergency highlights
the fact that the disease is a critical livelihoods and rights issue, seriously compromising access to food at
the household level. This fact should be amplified at the highest levels, building on the momentum
generated by the UN Special Envoy’s joint statement of January 2003 and the establishment of a
Commission on HIV/AIDS and Governance in Africa (CHGA) by the UN Secretary General in February 2003.
In addition, the question of whether or not HIV/AIDS, in its own right, warrants emergency programming even
if the 2003 and subsequent harvests are good, needs to be placed squarely on the table. This is particularly
important given the expected exponential growth in the disease over the next 10-12 years. The need for a
“paradigm shift” in the way that development and emergency programming is implemented in the region to
tackle effectively this growth is an issue that needs to be raised at all decision-making levels.
Implications for Research:
This study highlights the need to differentiate between households according to
the type of impact that HIV/AIDS has had or is having. The relationship between poverty and AIDS impact is
a strong one, this is known already. What is less understood is the relative importance and dynamics of
different AIDS related morbidity and mortality profiles within and across wealth groups in relation to
household food security. This study identifies some possible relationships in this regard, but much more
needs to be done in a more controlled and focussed research environment than was possible here. A key
area of research will be to track HIV/AIDS infected and affected households of different types through time to
see how resilient or vulnerable they are to livelihood shocks (such as the 2002 food shock) and longer-term
trends – such as gradual land degradation and economic decline. Research should take into account extrahousehold
factors, such as kinship and other forms of social capital in livelihoods trajectories.