Draft THE HOUSEHOLD VULNERABILITY INDEX FRAMEWORK (HVI) By Dr L.M Sibanda, T Kureya and U Chipfupa Vulnerability assessment for better programming REGIONAL SECRETARIAT 141 Cresswell Road, Weavind Park 0184 Private Bag X813, Silverton 0127 Pretoria, South Africa Tel: +27 12 845 9100 Fax: +27 12 845 9110 Email: [email protected]www.fanrpan.org
60
Embed
HOUSEHOLD VULNERABILITY INDEX · PDF file01.08.2014 · Draft Financial capital- HIV and AIDS evoke changes in household financial assets, investment choices, household income and
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Draft
THE HOUSEHOLD VULNERABILITY INDEX FRAMEWORK
(HVI)
By
Dr L.M Sibanda, T Kureya and U Chipfupa
Vulnerability assessment for better programming
REGIONAL SECRETARIAT 141 Cresswell Road, Weavind Park 0184 Private Bag X813, Silverton 0127 Pretoria, South Africa
Given an example of three households that is household A, B and C.
Household A is
Headed by an 18 year old child who is uneducated and no longer attends school
He takes care of 5 brothers and sisters who are going to school
They own 3 cattle
The mother is alive but has relocated back to her kin group because of sickness
They are currently getting support from NGOs in terms of food, seed packs, etc
And on average they harvest 80 kg of maize per season.
Household B is
Headed by a single parent – a woman who has never been married
she has three kids all of whom are going to school
She owns 5 cattle
She is involved in informal work such as gardening
The household harvest an average of 120 kg of maize per season
And she is the sole income earner.
Household C
Has both parents available
The father is bed-ridden
they own no cattle
They have 4 kids who do not go to school
They are involved in casual work
They practice dry tillage And harvest an average of 40 kgs of maize per season. The question is which one of the three families is most vulnerable and which one is least vulnerable and why? Given the information in the example, the HVI would rank
Household C as the highest in terms of vulnerability despite the fact both parents are alive. This is because the household
o does not own any productive assets o they have limited labour available for off farm work o they have low agricultural productivity o they have low literacy levels and hence bleak future
Household B will be ranked the least vulnerable despite the fact that it is headed by a single mother The household
o can afford to send children to school o owns cattle a very important livestock in agric production o has a diversified income source o has labour vital for farm and off farm work
The human, financial and physical capital indices in the HVI model will also show that any meaningful intervention to assist Household C will focus on three issues that is:
o labour saving technologies o livestock projects to replenish household assets
Income generating projects to improve household income
Draft
Chapter 2: Review of Literature
When the HVI was initially developed it was designed to quantify vulnerability introduced by HIV and AIDS on household livelihood. It was at a later stage that it was also discovered that the same model could apply to other vulnerability shocks as they all affect the household livelihood in a similar manner. However literature that is discussed in this section relates to HIV and AIDS and its impact on household livelihoods. This literature was used to inform HVI development and design.
2.1 HIV and AIDS, Household Vulnerability and Food Security
HIV and AIDS affect rural households, most of whom depend on agriculture as a source of
livelihood (Mano and Chipfupa, 2005). Mutangadura et al (1999) and Shapouri and Rosen (2001)
state that HIV and AIDS is a major threat to agriculture and food security because it reduces
agricultural productivity and diminishes the availability of food through direct loss of family
labour, reduction in time allocated to farming, sale of farm assets, cultivation of marginal land
and marginalization of surviving widows from land ownership by customary land tenure systems.
HIV and AIDS cause spending to rise particularly on medical care and funeral expenses (Bates et
al, 2004). FANRPAN’s study on Impact of HIV and AIDS on Agriculture and Food Security also
confirmed the above findings, and generally showed that food production and income declines in
HIV and AIDS affected households. The pandemic exposes rural households to poverty mainly
through its effects on agricultural production and food security.
The extent of this exposure that ics how households are vulnerable to the impacts of HIV and
AIDS depends on their socio-economic and political status. Households are bound to have
varying degrees of resilience and ability to cope and this has implications on the policy
recommendations intended to mitigate the impact of the condition. As refuted by Bates et al
(2004), vulnerability is too broad a concept to enable effective targeting of the most vulnerable
especially when resources are scarce. In their guidelines for vulnerability mapping the World
Food Programme (WFP) (1999) stressed the need for creating a vulnerability database that is
useful to identify both chronic and transitory vulnerabilities that is, groups that are permanently
vulnerable and those that are temporarily vulnerable must be differentiated for appropriate policy
action. This cements the need to develop an appropriate method of quantifying the levels of
vulnerability of each household.
2.2 Sustainable Livelihood Framework
The impact that HIV and AIDS have on rural livelihoods is better explored using the Sustainable
Livelihoods (SL) framework approach. According to the SL framework, households are only
viewed as being sustainable if they can adjust to threats without compromising their future ability
to survive shocks to their livelihoods and HIV and AIDS is potentially one of those shocks
Draft
(Carney, 1998). The most common hypothesis that relates the impact of HIV and AIDS on asset
accumulation and holding capacity of rural households is that the pandemic strips individuals,
households, networks and communities of assets (Gillespie and Haddad, 2001). The pandemic is
alluded to represents a potentially devastating shock to farm household survival. The illness or
death of one or more household members can affect each of the livelihood assets resulting in a
reduction in the ability of the household to adjust to future shocks (Stokes, 2003).
The SL framework (Carney 1998, DFID/FAO, 2000) has provided quite a clear basis for
understanding how HIV and AIDS can impact on various aspects of livelihoods in many
different ways. The framework depicts livelihoods as being determined in the first instance by the
range of assets available to the household. “Assets” is used as a broad term, and five categories of
assets or capital are identified which are human, physical, financial, social and natural capital.
O’Donnell (2004) has argued recently that the SL framework can provide a clear basis for
understanding how HIV and AIDS can impact on various aspects of livelihoods in many
different ways. When considering livelihoods from the perspective of HIV and AIDS, a
livelihood system analysis will take on an additional character. The analysis begins with
identifying livelihood strategies that are susceptible to HIV and AIDS, and then tracks the impact
of AIDS on livelihood assets—human, natural, financial, physical and social. Such an analysis
should reveal intervention points for reducing the risk of HIV infection and mitigating the
negative impact of HIV and AIDS, so that preventive measures can be linked to mitigation
efforts to address both the causes and symptoms of the disease (Tango International, 2003: 4-5).
Drawing on the work of Chambers and Conway (1992) a livelihood is defined as comprising ‘the
capabilities, assets and activities required for a means of living’ (Carney, 1998) As an approach to
understanding and facilitating development the SL approach contains echoes of the basic needs
approach and its evolution into concerns with food security and then poverty alleviation and
reduction (Maxwell, 1998). It also draws on the insights from integrated rural development,
farming systems research and participatory approaches in development. These various strands are
linked with appreciation first of the diversity of livelihoods of rural people, second of the roles of
different types of assets in rural peoples’ livelihoods, and third of the importance of the wider
social and political and economic environment in mediating access to assets. While increasing
evidence has accumulated that rural people engage in many different types of income generating
and livelihood activity (Taylor et al. (2000), Ellis (1998)), it is also recognized that their ability to
engage in non agricultural activities is often very dependent on their access to assets ((Reardon,
1997; Baker ,1995 and El Bashir, 1997, cited by Tacoli 1998); Dercon and Krishna (1996), de
Draft
Janvry and Sadoulet (1996)). These authors show that different types of activities require
different combinations of financial, human, social, physical and natural capital.
As Moser (1998) argues, analysis of the linkages between people’s access to assets and livelihood
diversification goes back into the literature of the late 1980’s on people’s coping strategies in
response to seasonality and famine (Corbett 1989; Davies 1989) and on the role of entitlements
and assets in these coping strategies (Sen 1981; Swift 1989). These coping strategies aim to
maintain a minimum level of consumption by:
Contributing to overall production and income
Allowing exchange and or consumption in periods when there is no income.
Hence analysis of assets in rural livelihoods therefore needs to examine the functions of different
asset types within the asset portfolios held by poor people with different livelihood strategies.
Such analysis must then progress beyond categorization of the type of capital as emphasized by
the SL framework, to identify priorities for policy and for other interventions supporting
expanded access to assets
2.3 Review of HIV and AIDS Impact on Rural Livelihoods
2.3.1 Human Capital
(a) Changes in household demographic structure and labour availability
Human capital assets represent the skills, knowledge, ability to labour and good health that
together enable people to pursue different livelihood strategies and achieve their livelihood
objectives. Smallholder agriculture is labour intensive due to low levels of mechanization. HIV
and AIDS has the potential to erode the active labour force in farming systems where women
and children already make up a higher proportion. HIV and AIDS is debilitating, and reduces
hours at work due to chronic illness. Attending funerals and other related rituals like memorial
services frequently reduces labour time input, which is important to animal well-being and crop
security.
The loss of adult on- and off-farm labour is one of the most widely discussed effects of the HIV
and AIDS pandemic (Topouzis and du Guerny 1999). The loss of experienced agricultural
workers affects both individual households and communities, resulting in labour shortages and
declines in productivity both on and off the farm. Declining productivity, in turn, leads to
declines in household income through both decreases in the household's own production and
through declines in off-farm income and remittances. An increase in household expenditures on
Draft
medical care results in a decline in savings and the loss of assets through the sale of both
productive and non-productive assets. Thus, the loss of human capital leads directly to declines
in the financial capital of the household. For food insecure households or those slightly above
this threshold, the loss of labour, income and increased expenditures for medical care can push
them further into poverty and food insecurity (Stokes, 2003). In Tanzania studies reveal that, by
2010, the estimated size of the labour force will shrink by 20 percent because of AIDS, and the
mean age of workers will fall from 32 to 28 due to a shift to younger and less experienced
workers (Jackson, H. 1997).
Committee on World Food Security (2001) estimated that approximately 2-person years of
labour are lost by the time one person dies of AIDS, due to their weakening and the time others
spend giving them care. According to Shah et al (2001) in his study on Production Systems, there
was evidence that out of 310 households over 70% of households affected by chronic sickness
experienced a loss of labour, 45% experienced delayed agricultural operations, 23% left the land
fallow, 26% experienced changes in crop mix and 36% experienced changes in source of
livelihood. The timing and duration of the sickness (pre and post harvest), multiple stresses and
relative economic status were found to be the most critical factors determining the intensity of
the impact. Women in patrilocal villages were reported to be more vulnerable than in matrilocal
villages.
Barnett and Blaikie (1992), published research undertaken in the Rakai and Kabale districts of
Uganda to identify the effects of HIV and AIDS on households and farming systems. The
research focused on the effect of labour losses and mapped the relative vulnerability of different
farming systems. The main conclusions were that, some farming systems would contract in areas
cultivated, productivity and range of crops because of labour shortages and that some child
headed households were emerging.
The impact of mortality on household demographics may be much more severe when the adult
death is due to HIV and AIDS than with other causes of death. Two person-years of labor may
be lost because of the weakening of the person and the amount of time spent caring for him or
her before death (FAO 2003a). Adverse dependency ratios were observed in households with the
death of an HIV-positive adult but not with the death of an HIV-negative adult, and there was a
significant association between child-headed households and adult death from AIDS (Menon et
al. 1998). Floyd et al. (2003), in a retrospective cohort study in Malawi, investigated the effect of
HIV on household structure over more than ten years. At the time of the follow-up survey, only
one in five marriages in which one partner was HIV positive at the outset was still intact.
Draft
Children of HIV-positive parents were less likely to be alive and resident in the district than
children of HIV-negative parents.
(b) Gender implications
Gender is the differences between women and men within the same household and within and
between cultures that are socially and culturally constructed. These differences are reflected in:
roles, responsibilities, access to resources, constraints, opportunities, needs, perceptions, views,
etc. held by both women and men. (Moser, 1993)
Women play a major role in all developing countries in the different aspects of agricultural
production i.e. subsistence crops, market gardens, cash crops and animal production (Muchopa
et al 1999). Although their work is largely unacknowledged, women are the major food
producers, accounting for approximately half of the communal farmers. Although men and
women participate in most agricultural tasks, men predominate in land preparation, ploughing
and pest control; women are primarily engaged in watering, planting, fertilising, weeding,
harvesting and marketing, firewood gathering, food processing and preparation, cooking and
domestic work, activities that are typically labour intensive. Hence women in agriculture can be
considered as an untapped source of agricultural growth (Muchopa et al 1999). Various studies
that have been done so far indicate that women spend more labour in the production of food
crops especially those intended for consumption than men. Apart from food production women
are solely responsible for food processing, preservation and storage. Alvord, (1929) and
Holleman, (1952) revealed that women were the major food providers and participants in the
labour force within the communal mode of production. Boserup. E. (1970) also described
Southern Africa as a region of female farming par excellence. In rural Africa, in studies that were
done by Neema (1999) and Quisumbuig et al (1998) revealed that women account for 70% to
80% of food production.
Traditionally, rural women have always had a triple role to play in society. These roles are
differentiated as reproductive, productive and community roles (MOHCW, 2003) The study by
Laver (1995), showed that in comparison to men, most women operate small-scale farm
businesses and earn low income from agriculture; they are severely overburdened and work
longer hours in their triple roles of production, reproduction and community work. Female-
headed households are usually poorer; fewer rural female-headed households own agricultural
productive resources and their household incomes are 40% less than rural male-headed
households. HIV and AIDS therefore exacerbate existing constraints already faced by women
farmers. The advent of HIV and AIDS has further expanded the care giving role as women are
required to or expected to take care of those who are sick with HIV and AIDS related illnesses.
Draft
The primary impact of HIV and AIDS on crop production is reduced yields, because the women
are unable to perform all the crop production activities. Labour intensive activities such as
weeding and harvesting suffer severely. The resulting decline in output may have important
implications on the food security status of the household and income from crop enterprises.
Similarly, livestock enterprises relying heavily on female labour may suffer from low production.
Affected female headed farming households may suffer severely in the short run through AIDS
related production losses in food and income. They might be forced to compromise the longer-
term survival by taking steps to offset emergent consumption needs. Some of which includes:
Sale of assets such as equipment, livestock and household items. This might have
detrimental implications on long-term agricultural production.
Borrowing to meet the short-term household consumption requirements, which again
means that the household is engaging in future debts that might deepen the crisis.
Reduced and diversified consumption when households are forced to cut back the
number of meals eaten each day. This has important implications on the health of the
family and any continued agricultural production.
A study which was done in Kagabiro village in Tanzania revealed that when a household included
someone with HIV and AIDS, 29% of the household labour was spent on AIDS related matters
and in two thirds of the cases women were devoted to nursing duties and in average the total
labour that was lost to households was 43%. This affects yields that are produced since labour
and time that would otherwise have been used productively in the fields or doing agricultural
work is transferred to caring for the sick.
The Committee on World Food Security (2001) research in Tanzania found that women spent
60% less time on agricultural activities taking care of their ill husbands. Yamano and Jayne,
(2004) in a 2-year panel of 1,422 Kenyan households surveyed in 1997 and 2000 revealed that
household size declined by 0.64 persons among households with adult death compared to the
control group indicating partial replenishment. Larger reduction in household size due to female
death than male death was observed. No change was found in total area cropped between
households incurring and not incurring an adult death. Households incurring a male adult loss
reduced land devoted to high-value-added crops. Those incurring the death of a female spouse or
head reduced the size of cultivated area devoted to cereals. The death of household heads and
spouses adversely affected the value of total crop output per acre.
Draft
Both the quantity and quality of farm household labour are reduced through incapacitation or
death. Researchers and scholars generally agree that the infection rates are higher among women.
Since women account for 70% of the agricultural labour supply and as much as 80% of food
production, HIV and AIDS prevalence among women registers negatively on the quantity and
quality of labour and on farm output (Baier, 1997). In addition the care time devoted to the
patient ill due to AIDS by the healthy members of the household robs agriculture of labour.
While this applies to most diseases such as malaria, from which African smallholder farmers
suffer, the effects of HIV and AIDS are more telling because of the its long-term impact. Malaria
may be treated and overcome within days of effective treatment so that the patient returns to his
work, this is not the case with HIV and AIDS, which may linger for several years with or without
treatment, and during which time the infected is perennially incapacitated.
(c) Mobility and disintegration of household members
HIV and AIDS have increasingly becoming a factor influencing migration and mobility in Africa.
HIV and AIDS have become so all-encompassing for individuals, household and communities
that it seems to generate new forms and different mixes of population mobility. One clear focus
of mobility associated with the households of those who die from HIV and AIDS is the prospect
of substantial changes in composition of households: some households lose their cohesion on
the death of the household head and may dissolve, with spouse and children of the deceased
having different social obligations; others may gain labour to replace the loss of one member,
perhaps with the ‘fostering’ of the much increased number of orphans.
Some factors that induce movement include:
HIV positive people commonly returning to live with family members to obtain care and
treatment.
Others migrate in order to provide care to family members living elsewhere (Young and
Ansell, 2003).
Migration by other household members to seek income-earning opportunities especially
after the loss of a household's income though mortality and morbidity of the
breadwinner(s),
People diagnosed with HIV or displaying physical evidence of disease may migrate to
avoid stigmatization by their communities,
Children that are orphaned by AIDS may migrate to live with relatives or to seek their
own income-earning opportunities;
and widows or widowers may migrate upon the death of their partner. The tradition of
“wife inheritance” in some African societies means that a widow becomes the wife of
her late husband's brother, which may require relocation. Other women or men choose
Draft
to move after the death of a spouse, perhaps to join biological family elsewhere. The
death of a husband can lead to wife losing access to land and thus livelihood, forcing her
to move elsewhere to seek a new life.
Research by various scholars done in Southern Africa has found a similar pattern where
households experiencing adult mortality tend to become permanently smaller than other
households (Janjaroen 1998; Menon et al. 1998; Yamano and Jayne 2004) as some household
members leave following an adult death. In Uganda, for example, household size fell by 1.7
members on average in households that experienced death, compared to a decline of 0.1 persons
in other households (Menon et al. 1998). Changes in household size and composition following a
death are sensitive to the age, gender, and position of the deceased adult. In Kenya, when a
female adult died, children were frequently sent to live with relatives, whereas the death of a male
household head often led to daughters leaving the household on marriage (Yamano and Jayne
2004). Similarly, in a recent study in Mozambique, Mather et al. (2004a) found that after a female
prime-age death, it is likely that children will leave the household and a new female adult will
arrive. Some households dissolve after a prime-age adult death. The death of a male head of
household in Tanzania is more likely to cause dissolution of the household than the death of a
female head (Urassa et al. 2001). Hosegood et al. (2003) in rural KwaZulu Natal, South Africa,
found that 5 percent of 10,490 observed households experienced at least one AIDS-related death
during the one-year observation period. These households were three times more likely to
dissolve than other households.
In Uganda, Ntozi (1997), in a retrospective study of the migration of spouses and other
household members, found that 37 percent of widows and 17 percent of widowers migrated
from their original homes (spousal death from AIDS-related causes ranged from about 50 to 60
percent). Younger spouses and those in worse health were more likely to leave. Women were
more likely to leave because they were generally not entitled to inherit the land, and their kin
often lived elsewhere. Even when it does not dissolve the household, death may cause dislocation
of families: among the matrilineal people in rural Zambia, for example, women return with their
children to their own mothers’ villages (Drinkwater 1993). In studies in Kenya and Mozambique,
households were often unable or unwilling to replace their adult members after an adult death
(Yamano and Jayne 2004; Mather et al. 2004a).
In contrast, Rwandan households with an adult death were able to maintain their labour supply
through addition of new members (Donovan et al. 2003). In Mozambique, households were
more likely to hire or share labour after male deaths than after female deaths. Death of a
household head increased the likelihood of the use of child labour. One interesting finding of
Draft
Michigan State University’s comparative study of Kenya, Malawi, Mozambique, Rwanda, and
Zambia was that, in four of the five countries, a majority of deceased prime-age adult deaths were
not household heads or spouses (Mather et al. 2004b). This observation explains the low
household dissolution rates in these countries. The authors suggest that the potential effect of
prime- age mortality on agriculture may thus be less than is predicted by other studies, given that
household heads and spouses tend to be the household members most heavily involved in
agriculture. The demographic group most affected was younger female dependents.
Urassa et al, (2001) in his study on the Impact of AIDS pandemic on mortality and household
mobility in Mwanza Region, Tanzania found out that in 44 percent of households in which the
head died, all members moved out of the household. Hosegood, Herbst and Timaeus, (2003) in
their study also found that, household instability was significantly associated with younger heads,
female heads, and death of a household member. Five per cent of the households experienced at
least one AIDS related death during the period of observation. These households were nearly
three times more likely to dissolve than other households. Child-headed families were found to
be rare. Household size decreased due to both the death of a household member and out-
migration of surviving members. Janjaroen, (1998), in a cross-sectional comparison of 324
households with AIDS related deaths, non-AIDS related deaths, and no deaths found out that
households that experienced an adult death were almost a full individual smaller than they had
been prior to the death. Deaths of adult female had a stronger negative impact on consumption
than deaths of adult male.
Overall, it is apparent that there are new mixes of forms of movement in HIV and AIDS affected
populations. The mobility generated by the excess mortality is clearly of major economic as well
as social importance. However, these outward migrations are not always negative but could be
evidence of still functional social support systems of a society in the face of the catastrophic
economic effects of increased adult mortality.
(d) Impact on agricultural extension services
Agricultural extension is the process of transferring information and technology to farmers for
use in the production process and similarly transferring information from farmers to researchers
to solve the problems of farmers (Swanson, 1984). The extension process has been clearly
expressed as a two way process where extension agents transfer knowledge and ideas to farmers
whilst on the other hand being receptive to farmers’ ideas, suggestions and problems so they can
be incorporated into the extension message (Rukuni and Eicher, 1994). HIV and AIDS takes a
heavy toll on national development staff. Agriculture extension services usually provided to the
farmers by government are disrupted as the staff responsible for these activities become ill and
Draft
die. Provision of care to sick family members, bereavement and compassionate leave and
observation of mourning times further reduce the staff's productive time.
Dry land and irrigation schemes in Zimbabwe have experienced extension staff shortage due to
HIV and AIDS. In a study on the impact of HIV and AIDS on Smallholder Agricultural
Production that was conducted in Gweru by Ncube (1998) revealed that in the whole of
Smallholder Agricultural Production Area (SHAPA), five extension workers died each season
because of the AIDS related illnesses which is about 15 percent of the total number of extension
workers in SHAPA. These deaths left some areas unattended for a long time, negatively affecting
productivity (Ncube, 1998). This led to 79.1hectares of maize crop, 7.9 hectares of sunflowers
and 2.3 hectares of groundnuts remaining uncultivated. In Uganda, disruption in services due to
illness and death has led to an informal reduction of the length of the staff working week
(Topouzis, 1998).
By one estimate, approximately two person-years of labour are lost by the time one person dies
of AIDS. According to FAO, AIDS has killed about 7 million agricultural workers since 1985 in
the 25 hardest-hit countries in Africa, and it could kill 16 million more before 2020. The loss in
the agricultural labour force through AIDS in the nine hardest-hit African countries, for the
period 1985-2020, was projected as follows: Namibia 26 percent; Botswana 23 percent;
Zimbabwe 23 percent; Mozambique 20 percent; South Africa 20 percent; Kenya 17 percent;
Malawi 14 percent; Uganda 14 percent; United Republic of Tanzania 13 percent (FAO, 2001).
A study in Zambia and Uganda (IFAD 2001), discovered that:
a) Increased funerals in the communities were leading to cancellation or postponement of
expensive planned activities and make impossible timely attainment of programme
targets.
b) Families directly affected by AIDS related sickness or death were prevented from
participating in group to organise extension activities through lack of adult
representation or the shift in family priorities towards caring for the sick or searching for
food rather than attend extension meetings.
c) Illness and death of staff and extension contact farmers/ community organisers lead to
loss of expensively attained knowledge and experience and low adoption of technologies
and agricultural innovations. In one district office of Zambia, four of the 22 extension
staff members had died in the one year prior to the mission and three of these, according
to the officer who reported this incident, were AIDS cases and similar high staff
mortality was reported in Uganda.
d) Increasing staff workload through the need to train community workers, group leaders,
Draft
and farmers to replace those trained but ill or dying off or need to attend to AIDS
related eviction of orphans and widows from land and property.
e) Increased domestic pressure, low incomes of staff, trauma and burn out on account of
having to look after sick relatives or attend funerals.
Although systematic surveys are not available, it is probable that agricultural extension workers
have higher levels of HIV than the general population. They are mobile and relatively affluent in
poor communities, which are known risk factors to HIV. It is probable that they are dying more
quickly than they can be trained. In a number of countries, agricultural extension services are on
the verge of collapsing entirely because of the HIV and AIDS pandemic. The pre-existing
weakness of the services predisposed them to this structural crisis. (DFID, 2003)
The high HIV and AIDS related deaths among Ministry of Agriculture (MoA) staff in Eastern
and Southern Africa (ESA) is likely to have negative impacts in the fight against the pandemic. In
Kenya’s MoA, 58 percent of all deaths in the late 1990s were thought to have been AIDS-related
(GTZ: 1999). In Malawi in 1998, at least 16 percent of the staff of the Ministry of Agriculture
and Irrigation (MoAI) were reported to be living with HIV and AIDS, 76 percent had lost at least
one colleague, and 60 percent had lost at least one close relative to AIDS (Bota et al. 1998, cited
in Topouzis 2003). In Zambia, 67 percent (of 155) agricultural extension workers interviewed had
lost at least one co-worker to HIV and AIDS in the three years preceding one study (Alleyne et
al. 2001). Studies have also reported reductions in agricultural extension service time due to HIV
and AIDS. Haslwimmer (1994) in Uganda reported 25 to 50 percent reductions in agriculture
extension time. In the mid-1990s, Haslwimmer (1994) found that up to half of agricultural
extension staff time in one district in Uganda had been lost to HIV and AIDS. Staff members
were frequently absent from work because they had to care for sick relatives or attend funerals,
or were sick themselves. Organizations in areas with a high HIV and AIDS prevalence are
characterized by high absenteeism, high turnover, a loss of institutional memory, and reduced
innovation. As individuals in government and NGOs continue to die, the capacity gap between
what is needed and what can be delivered is becoming an abyss.
2.3.2 Financial Capital
(a) Changes in household financial capital assets
Financial capital refers to the financial resources that people use to achieve their livelihood
objectives, including stocks (savings, convertible assets, including livestock) and flows of income.
The loss of human capital would lead directly to a loss of financial capital. Household incomes
Draft
declines especially if, as stated by Stokes, HIV infections and AIDS related deaths are
disproportionately concentrated in the most productive age groups (15-49 years). This income
decline from farm and off-farm sources further renders rural households vulnerable to food
insecurity. As productive assets are sold off, the household's future livelihood is jeopardized.
Among the financial capital effects thought to be influenced by the pandemic are reductions in
income from farm and off-farm sources, liquidation of savings accounts, seeking of remittances
from family, changes in degree of reliance on off-farm income among male, orphan and female-
headed households, changes in wage earnings among female-headed households, changes in
income-generating activities among female-headed households (Topouzis 2000), sale of stores of
value (jewellery, household goods), borrowing from informal sector (Mutangadura et al. 1999),
borrowing from rural traders or money lenders (often at exorbitant interest rates), exhaustion of
credit resources, sale of livestock, increased expenditure on health care, transport and funerals
and reduced expenditure on agricultural inputs.
It should be noted that the sale of livestock appears under several asset rubrics and illustrates the
difficulty in classifying some effects under only one category. While livestock are generally
thought of as part of natural capital, they also operate as a store of wealth in many societies in
which financial markets are underdeveloped. Moreover, by providing animal traction power, they
also operate much like physical capital assets. Disposal of draught livestock directly affects the
household's productive activities and increases the risk of food insecurity.
(b) Changes in household income, expenditure and consumption patterns
In a cross-sectional comparison of households with and without an HIV-positive individual in
Free State, South Africa, per caput income in AIDS-affected households was 50% to 60% that of
unaffected households (Booysen and Bachmann 2002). In another cross-sectional survey of 680
households in Limpopo province, Oni et al. (2002) made similar observations. In a five-year
retrospective study of 232 urban and 101 rural AIDS-affected families, Nampanya-Serpell (2000)
reported a decline in monthly disposable income of more than 80 percent in more than two
thirds of the AIDS-affected families, with higher losses following a paternal death. But Urassa et
al. (1997) found that in Tanzania, households with orphans did not have a lower economic status
than those without orphans (though this may be a positive selection bias, as households with
greater resources are more likely to foster orphans).
Reducing food consumption quantity or quality may be a highly erosive “coping” strategy, as
nutrient requirements rise following HIV infection. In a panel study in Indonesia, Gertler et al.
(2003) showed a prime-age male death to be associated with a 27% reduction in mean per capita
household consumption, whereas the death of a female had no significant impact. In Mexico,
Draft
they found the death of a prime-age adult household member to reduce per capita consumption
by nearly 8 percent, with no significant gender differences. In Côte d’Ivoire, Bechu (1998)
surveyed 107 households with at least one adult ill with AIDS related illnesses and with one or
more children and interviewed them six times at two-month intervals. The data was compared
with the results of a study conducted in Yopougon in May 1992 and based on a sample of 2,064
households. The study found per capita consumption of AIDS-affeccted households to be half
that of other households. In a cross-sectional survey of 119 households in the Rungwe district of
Tanzania, Mwakalobo (2003) found that households that experienced an AIDS related death
spent substantially less on food than other households. HIV and AIDS related death significantly
increased the probability of a household falling below the poverty datum line. In South Africa,
average monthly per capita food expenditure of affected households was 70 percent to 80
percent less than that of other households (Booysen and Bachmann 2002), but no significant
difference was found in total monthly expenditures, most likely because of rises in health-related
expenditures. Many studies show that households experiencing adult death tend not to recover to
pre-shock levels of consumption (e.g., Yamano and Jayne 2004; Gertler et al. 2003; Bechu 1998).
Such a lack of resilience is likely to increase vulnerability to other shocks to food and nutrition
security.
AIDS-affected households do tend to incur high health-care expenditures (Tibaijuka 1997 in
Tanzania; Booysen and Bachmann 2002 in South Africa). Bechu (1998) in Côte d’Ivoire found
that health-care costs specific to the person with AIDS accounted for almost 80 percent of the
household health-care budget. In the Rungwe district of Tanzania, rising medical expenses or an
HIV and AIDS-related death significantly increased the probability of a household’s falling below
the poverty datum line (Mwakalobo 2003).
A study by Menon et al (1998) revealed that a decline in household durable goods was only in
households with HIV-related deaths. This effect was more pronounced in agricultural than in
trading villages. Bechu, (1998), in his study on the Changes in household expenditures and
consumption due to illness and death of a household member from HIV and AIDS in Cote
d’Ivoire found out that, consumption among the AIDS affected households was only half that of
the comparison group representative of the general population. The portion of the budget spent
on health care in AIDS related households was almost double that of the households in
Yopougon. Although a slight increase in consumption was observed after the initial shock of
death, the households did not return to earlier levels of consumption. Health costs for the sick in
the AIDS affected households accounted for almost 80 percent of the health budget. But once
the illness was identified and there was access to health care, most of the sick seemed to distance
themselves gradually from the structures of modern care – and to a lesser extent from traditional
Draft
medicine – by cutting back on hospitalisation, medical consultations and medicine. Oni et al,
(2002) revealed that HIV and AIDS affected households had lower annual income, were smaller
in size, had lower savings and spent more on transportation, funerals and health care but less on
housing, remittances and education than unaffected households. The coping strategies adopted
by affected households included sale of household assets, withdrawal of children from schools
and joining community support groups. Nampanya-Serpell, (2000) found out that a decline in
monthly disposable income of more than 80 percent was observed in more than two thirds of the
AIDS affected families, especially following paternal death. According to Booysen and
Bachmann, (2002) household members, mainly those who were unemployed, spent an average of
7.5 hours per day providing care during the fatal illness of the deceased. Per capita adult
equivalent income in affected households was only between 50-60 percent of the levels of
income in non-affected households. The average monthly food expenditure of affected
households was 70-80 percent less than of that of the unaffected households, although no
significant difference was found in total monthly expenditures.
Therefore, from the empirical evidence presented it can be seen that taking care of a person sick
from AIDS related illnesses is not only an emotional strain for household members, but also a
major strain on household resources. Loss of income, additional care-related expenses, the
reduced ability of caregivers to work, and mounting medical fees push affected households
deeper into poverty. It is estimated that, on average, HIV-related care can absorb one-third of a
household’s monthly income (Steinberg M. et al. October 2002).
(c) Changes in household investment choices
Loss of income and unplanned additional care-related expenses adversely affects household
investment decisions such as investments into human capital e.g. education, agriculture, financial
assets e.g. savings, physical assets, etc.
Many studies have reported the negative effect of HIV and AIDS on children’s schooling (Urassa
et al. 1997; Gilborn et al. 2001; Deininger et al. 2003; Gertler et al. 2003). Deininger et al. (2003),
in an analysis of a panel data set of 1,300 households included in surveys conducted in 1992 and
2000, showed that foster children were at a distinct disadvantage in both primary and secondary-
school attendance before the introduction of universal primary education. In Uganda, in a
descriptive analysis of a baseline survey of 353 HIV-positive parents, 495 children of People
Living With HIV and AIDS(PLWHAs) 233 orphans, and 326 guardians, Gilborn et al. (2001)
found declining school attendance among 28 percent of the older children of PLWHAs, whereas
it was improving for 21 percent of older foster children. In a recent study, Case et al. (2004) using
19 Demographic and Health Surveys (DHS) conducted between 1992 and 2000 in 10 sub-
Saharan African countries, showed that although poorer children were less likely to attend
Draft
school, poverty alone did not account for the lower enrollment of orphans. Orphans were less
likely to be enrolled than are non-orphans with whom they live. The lower enrollment of orphans
was largely explained by the greater tendency of orphans to live with distant relatives or unrelated
care-givers.Yamano and Jayne (forthcoming) found the negative impact of adult mortality on
school attendance in Kenya to be more severe in poor households, as did Nampanya-Serpell
(2000) in urban but not in rural areas of Zambia. Yamano and Jayne (forthcoming) also found
out that adult mortality was negatively affecting schooling even in the period directly before the
death, most likely because children were sharing the burden of care giving. Again, the type of
orphan hood seems to matter. In Indonesia only maternal death resulted in delayed school entry,
whereas paternal death increased the dropout rate.Exactly the reverse occurred in both cases in
Mexico (Gertler et al. 2003).
(d) Sustainability of household food and nutrition security (food accessibility and utilization)
Food security is the availability of food, access to food and the absence of risk related to either
availability or access. Most people in the SADC region derive their household food security from
crop and livestock agriculture. About 70 percent of the SADC population is engaged in crop and
livestock production. As a result, increases in farm output and productivity enhance food and
income security. But the adverse effects of HIV and AIDS have resulted in labour deficits and
the sale of agricultural output to meet household medical costs, leaving the household food
insecure. According to Barnett and Rugamela (2001), households are said to be food secure if
four factors are in balance. These are food availability, equal access to food, stability of food
supplies and quality of food. HIV and AIDS affect all these factors thus reducing food security.
IFAD (1996) describes household food security as ‘the capacity of households to procure a
sustainable and stable basket of adequate food’.
Food storage and processing activities are impaired when a household is affected by HIV and
AIDS. This also impacts on the availability of seed for subsequent cropping. Households
normally spend some time engaging in off-farm activities that enable them to earn some income.
In the case of urban areas most households meet their food needs through purchase of food
using income earned from formal employment. HIV and AIDS reduce this income given that the
infected become less productive and the affected are forced to tend to the sick. In essence this
reduces household’s purchasing power leading to a vicious cycle of food insecurity and poverty.
At the micro level HIV and AIDS reduces the ability of the household to produce and buy food,
depletes assets and reduces the household income and purchasing power thus reducing labour,
management of farm resources and skills and reducing the productivity of current workers thus
affecting food security.
Draft
Barnett and Rugalema (2001), Mutangadura et al (1999; 2000) cites the serious depletion of
human resources as one of the major impacts of the disease on agriculture. HIV and AIDS
results in the loss of experienced agricultural workers, which affect both individual households
and communities, resulting in labour shortages and declines in productivity both on and off the
farm. A decline in productivity leads to declines in household income through both decreases in
the household's own production and through declines in off-farm income and remittances. As is
clear from the preceding points, a decline in the quality and quantity of food can often be
expected. The incidence of increases of orphans and the food consumption of all surviving
household members often declines when an adult dies. In addition to these effects, which are
owing to the loss of labour, household food security can also be reduced through an increase in
the number of mouths to feed arising from the fostering of children or the hosting and caring of
sick relatives. Mwakalobo (2003) discovered that households experiencing AIDS death spent
substantially less on food than those not experiencing death. Logistics regression results revealed
that HIV and AIDS related death significantly increases the probability of a households falling
below the poverty datum line.
In households coping with HIV and AIDS, food consumption generally decreases. The family
may lack food and the time and the means to prepare some meals, especially when the mother
dies. Research in Tanzania showed that per capita food consumption decreased 15 percent in the
poorest households when an adult died. A study carried out in Uganda showed that food
insecurity and malnutrition were foremost among the immediate problems faced by female-
headed AIDS-affected households. For the patient, malnutrition and HIV and AIDS can form a
vicious cycle whereby under-nutrition increases the susceptibility to infections and consequently
worsens the severity of the HIV and AIDS condition which in turn results in a further
deterioration of nutritional status. Even when a person does not yet show symptoms, infection
with HIV may impair nutritional status. The person may lose their appetite, be unable to absorb
nutrients and become wasted. According to Mason et al (2003) the 2002 drought in Southern
Africa interacted with HIV and AIDS in high prevalence areas to bring about rapid deterioration
in child nutrition. However, because these effects were seen mainly in areas that previously had
better child nutrition, the effect is not obvious from averages. Underweight children increased
very substantially, for example, from 5 to 20 percent in Maputo between 1997 and 2002; from 17
to 32 percent in Copperbelt, Zambia from 1999 to 2001-2; and from 11 to 26 percent in
Midlands province, Zimbabwe from 1999 to 2002. Changes were much smaller during non-
drought periods and in areas of low HIV prevalence. These trends may be explained by direct
effects of pediatric AIDS (growth failure is occurring at younger ages) but the larger effect is
probably indirect, as drought and HIV hasten destitution in affected families. Traditionally worse
Draft
off areas appeared protected perhaps because of food assistance. Though this impact remains to
be determined.
(e) Household Market Access
Rural households have diverse livelihood strategies and one of them is engaging in non-
agricultural activities, including micro enterprises (agro-processing, trading and other off-farm
occupations). Through these various activities, households seek both to ensure their food
requirements and to generate the income they require to satisfy their immediate consumption
needs, social purposes and investments (IFAD, 2003). Interacting with agricultural markets is
thus an important aspect of the livelihood strategies of many rural households, rich and poor
alike. Markets are where, as producers, they buy their agricultural inputs and sell their products;
and where, as consumers, they use their income from the sale of crops, or from their non-
agricultural activities, to buy their food requirements and consumption goods. Virtually all
households in rural areas are, by preference, both producers and consumers, buyers and sellers;
and many sell agricultural produce and buy their food at different times of the year. However,
rural households that, for one reason or another, are unable to interact with these markets are
prevented from adopting these diverse livelihood strategies; and indeed, in many parts of the
world, rural poor people often say that one reason they cannot improve their living standards is
that they face difficulties in accessing markets.
The issue of market access may usefully be considered according to three dimensions:
Physical access to markets: Distance to markets and lack of roads to get to them or roads that are
impassable at certain times of the year is a central concern for rural communities throughout the
developing world. It undermines the ability of producers to buy their inputs and sell their crops;
it results in high transportation costs and high transaction costs, both to buyers and sellers; and it
leads to uncompetitive, monopolistic markets. Difficult market access restricts opportunities for
income-generation whilst remoteness increases uncertainty and reduces choice resulting in more-
limited marketing opportunities, reduced farm-gate prices and increased input costs. It also
exacerbates the problem of post-harvest losses, which can reach as high as 50 percent in some
areas. In doing so, it weakens incentives to participate in the monetized economy, and results in
subsistence rather than market oriented production systems (IFAD, 2003).
Market structure: Rural markets are characterized by extreme asymmetry of relations between, on
the one hand, large numbers of small producers/consumers, and on the other, a few market
intermediaries. Such market relations are characteristically uncompetitive, unpredictable and
highly inequitable. Rural producers who face difficulties in reaching markets often become
dependent on traders coming to the village to buy their agricultural produce and to sell those
Draft
inputs and consumer goods. However, especially in remote areas, a trader may not arrive reliably
or at all, and producers are often faced with little choice but to accept the first offer of the first
trader who shows up, however unfavorable it might be. Such a situation is exacerbated when the
trader is also the only source of information on prices and other relevant markets (IFAD, 2003).
Lack of skills, organization and information: In their participation in agricultural markets, poor
producers find themselves at a major disadvantage. Many have a poor understanding of the
market, how it works and why prices fluctuate; they have little or no information on market
conditions, prices and the quality of goods; they lack the collective organization that can give
them the power they require to interact on equal terms with other, generally larger and stronger,
market intermediaries; and they have no experience of negotiation and little appreciation of their
own capacity to influence the terms and conditions upon which they trade. With little experience,
no information and no organization, they have no basis upon which either to plan a market-
oriented production system or to negotiate market prices and conditions. Ultimately, their lack of
knowledge means that they are passive, rather than active, players in the market; that they can be
exploited by those with whom they have market relations; and that they fail to realize the full
value of their production (IFAD, 2003).
For these reasons, improved market access is not an issue of consequence only to better-off
producers, and it is not relevant only to cash crop, rather than food crop, production. It is of
importance to all rural households including HIV and AIDS affected households, and assisting
such households in improving their access to markets must be a critical element of any strategy to
enable them to enhance their food security and increase their incomes (IFAD, 2003).
2.3.3 Physical Capital
(a) Changes in household productive physical capital assets
Physical capital comprises the basic infrastructure and producer goods needed to support
livelihoods e.g. buildings, roads/ transport, water supply, and other communications. The
productive equipment that the households use in pursuit of their livelihoods also comes under
threat with HIV and AIDS (Gillespie and Haddad, 2001). Households' physical capital refers to
those tangible assets and producer goods other than their natural capital, i.e., housing, household
goods, furniture, tools and equipment, as well as livestock. Once savings and credit resources
have been exhausted and liquid assets have been disposed of, households resort to selling of
other assets (Stokes, 2003). The disposal of physical assets and equipment needed for agricultural
production means that rural households' ability to generate income and sustain their families in
the short term is reduced.
Draft
Donovan et al, (2003) observed that attempts by households to make up for lost income included
distress sales of household assets and livestock, increased dependency on natural resources for
food and income, and the establishment of clubs for group income generation. Rugalema (1999)
also noted that AIDS-induced illness consumes cash, productive assets, and social claims,
particularly the use of external labour. Funerals deplete resources of afflicted and affected
households.
Livestock activities might be jeopardized by loss of time to take care of the livestock, resulting in
death of livestock due to poor management. Frequent slaughters and sale of livestock to finance
medical care for AIDS related illnesses also lowers livestock productivity and has detrimental
effects on crop production (Engh et al. 2000; Haslwimmer 1994). For example in Rakai, Uganda
65 percent of 752 households with AIDS related deaths reported selling property to cover
medical treatment and burial costs (Konde-Lule et al., 1996). In another research in Zimbabwe,
Kwaramba (1999) indicated that AIDS affected households experienced a decrease in cattle and
goat numbers of 26 percent and 3 percent respectively due to livestock sales to meet treatment
costs. Given the importance of draft cattle to agricultural production, this severely compromises
the smallholder farmer’s ability to sustain agricultural production. The effect in agricultural
production could be further exercebated by the fact that after a male death, widows and children
are left to take care of the livestock and they often lack the management skills to do so
effectively, further eroding this asset (Engh et al, 2000 and Haslwimmer, 1994).
(b) Changes in optimal farm-household production decisions
Many studies in sub-Saharan Africa, particularly over the last three to four years, showed the
vulnerability of subsistence agriculture to the impacts of AIDS. These include reductions in the
area of land under cultivation and crop diversity, abandonment of specific activities and crops,
shifts to less labor intensive mono-cultivation, use of minimum-tillage techniques, and reduced
livestock use (e.g., FASAZ/FAO 2003; NAADS 2003; Drimie 2003). Below is a summary of the
main findings from different studies:
In a retrospective study from Kenya, Yamano and Jayne (2004), using a panel of 1,422 Kenyan
households surveyed in 1997 and 2000, found the death of a prime-age male household head to
be associated with a 68 percent reduction in per capita household crop production value. Adult
female mortality caused a greater decline in cereal area cultivated, whereas prime-age male adult
death resulted in a greater decline in cash crops such as coffee, tea, and sugar and nonfarm
income. Shah et al. (2001) made similar observations in Malawi.
Draft
In her study Black-Michaud, (1997), compared Burkina Faso and Cote d’Ivoire examining the
links between savannah and forest zone systems and migration. In Burkina Faso she found a
reduction in cultivated area and changes to the agricultural calendar apparently because of
reduced remittances due to the illness or death of a migrant worker, rather than due to local
illness and death among smallholder farmers themselves. In Cote d’Ivoire, however, illness and
death in village farm households were having a serious impact. Common to both countries were
the findings that cash crops were reduced before food crops and that the total area under
cultivation declined. In Rwanda, Donovan et al. (2003) in their descriptive analysis of a cross-
sectional study with a four-year recall of household mortality information also found that 60
percent to 80 percent of rural-study households suffering illness or death reported reduced farm
labor and land cultivation following death of a male head.
In Swaziland, Muwanga (2002) found a reduction of 54 percent in maize production following
the death of the household head. In Malawi, Shah et al. (2001) found 70 percent of the
households affected by chronic sickness to be suffering from labor shortages, with 45 percent
delaying agricultural operations and 25 percent leaving land fallow or changing the crop mix.
Empirical research in Zimbabwe indicate higher decreases in cash crop production such as
cotton compared to traditional crops such as maize and groundnuts for households which
experienced a death of a bread winner due to AIDS (Kwaramba, 1997).
In a reconnaissance survey of 220 households, followed by an in-depth study of ten households
affected by AIDS, Tibaijuka (1997) reported significant losses in agricultural production due to
labor loss, and reallocation of labor to nurse the ill. In Uganda, in an in-depth qualitative survey
in three communities at different stages of HIV and AIDS impact and different farming systems,
Barnett et al. (1995) found a progressive decline in production and socioeconomic status among
affected communities. In Mozambique, a nationally representative survey with recall data on the
deaths, departures, and arrivals of household members between 1999 and 2002 found that
affected households had smaller total and cultivated land areas, particularly following the death of
a male household head. But cultivated area per adult equivalent of the households experiencing
death was similar to that of unaffected households because of net out migration (Mather et al.
2004a)
Thus the overall observation from the literature is that HIV and AIDS related losses in labour
may have an overall negative impact on crop production in terms of area cultivated, yield levels
achieved and working capital devoted to agriculture leading to major changes in the cropping
patterns of the affected households.
Draft
2.3.4 Natural Capital
(a) Environmental Degradation and Changes in Household productive natural capital asset base
Natural capital encompass the natural resource stocks from which resource flows and services
useful for livelihoods are derived e.g. lands, trees, water sources. The loss of human and financial
capital can have important effects on a household's use and preservation of its natural capital.
HIV and AIDS can result in serious deterioration in the natural capital of households as the
declines in their human and financial capital could limit their ability to invest in maintaining and
improving their land base. Despite the fact that land is the most important primary natural asset
that rural households possess, adversely affected families end up selling or disposing off their
land. Stokes (2003), identified the following natural capital assets that can decline in the presence
of HIV and AIDS:
Reductions in soil fertility,
Declines in on-farm conservation and/or irrigation practices,
Decreased biodiversity due to asset stripping, selling of firewood, increased harvesting of
wild food, game, etc,
Renting or leasing out portions of the household's landholdings,
Appropriation of land by relatives (taken from widows, orphans) and
Sale of land
Each of the declines in natural capital reduces the household's ability to cope with the effects of
HIV and AIDS.
According to Odenya (2003), weeding and other inter-cultivation measures may be neglected as a
result of labour and input shortages. Some families may abandon traditional practices such as
mulching which replenish the soil, or may sell livestock which would otherwise provide manure,
thus reducing soil fertility. About 60 percent of respondents indicated that soil fertility had
declined, and greater exploitation of fuel wood for sale and wild foods for home consumption
were resulting in increased deforestation and the increased scarcity of wild foods.
In a study by Zakhe Hlanze et al (2005) titled: Impact of HIV and AIDS and Drought on Local
Knowledge Systems for Agro Biodiversity and Food Security, the results showed that in Swaziland
increase in morbidity results in indiscriminate exploitation of natural medicinal plants to cure
diseases and HIV and AIDS. Women and children are sometimes deprived of access to natural
resources upon the death of the man e.g. when a man dies land is reallocated by the chief.
Draft
2.3.5 Social Capital
(a) Social Support networks
Social capital refers to the social resources upon which people draw in pursuit of their livelihood
objectives, including networks, membership of formal and informal groups, and relationships of
trust and reciprocity.
The illness and death of household members can disrupt a household's links to their extended
family and the larger community (Stokes, 2003). On the communities the death of a male of a
household could seriously impair a household's ability to access community resources or even
receive family support. Some of the most common social impacts of the pandemic include:
Increased reliance on extended family and formal and informal community
organizations for agricultural production, housework, child care, and fostering
Increased reliance on community willingness to support educational and nutritional
needs of orphaned children (e.g. school fees, uniforms, and supplemental feeding)
Less time to participate in social and cultural activities
Possible disintegration of household (Gillespie et al. 2001; Stokes 2002; Harvey 2004).
At a household level, morbidity and mortality due to HIV and AIDS affect household livelihood
resources and assets, resulting in a reduction of the ability of the household to generate livelihood
and adjust to future shocks. Loss of human resource (labour), other resources and assets hinders
household participation in various social networks and groups. In areas where cultural practices
limit women's participation in formal organizations outside the home, the death of a male
breadwinner can seriously impair a household's ability to access community resources or even
receive family support. Nonetheless, studies indicate that households affected by HIV and AIDS
draw their support primarily from family, neighbors, community institutions and informal
organizations (Mutangadura et al. 1999). Thus, the social capital of households operating through
their relationships with extended kin and the community is critical to their ability to recover from
the illness and/or death of a household member due to HIV and AIDS.
At a community level social capital can have major impacts on mitigating the effects of HIV and
AIDS. Communities with high levels of social capital can provide affected households with a
variety of social support activities that permits families to adjust to the illness or loss of members.
Conversely, communities with low levels of social trust and solidarity can leave households and
families to fend for themselves or even to isolate and ostracize those households afflicted with
HIV and AIDS. According to Bernett and Whiteside, (2000), susceptibility and vulnerability are
Draft
determined by two variables, that is., the degree of social cohesion and the overall level of wealth
of the society. Four broad types of society were distinguished, each with a distinctive pattern of
HIV prevalence. Social cohesion can be strengthened in four areas, that is, altering social norms
and standards, improving the status of women, improving the performance of social institutions
and improving the quality of controlled social environments (e.g. improving housing and social
support for migrant-labor camps.
(b) Access to behavioral change information on HIV and AIDS and Agriculture
Access to information by household members as far as HIV and AIDS and agriculture affects the
vulnerability of a household to the impacts of HIV and AIDS.
Behavior change information on HIV and AIDS promotes and sustains risk-reducing behavior in
individuals and communities by distributing tailored health messages in a variety of
communication channels. Households that have access to HIV and AIDS related information are
better informed on how to deal with the condition. Understanding the pandemic influences the
kind of coping strategies that are adopted by a household in the fight against HIV and AIDS.
Access to HIV and AIDS information allows households that have sick members to adopt
recommended eating practices and sexual behaviour. This reduces frequency of illness and
subsequently medical cost incurred by the household.
Another important aspect is household’s access to agricultural information. Agricultural
information affects management practices employed by the rural households. Households that
are not well informed usually adopt farm management practices that are unsustainable and make
the households more vulnerable. Information that relates to crop mixes, input usage, market
information, etc, affects agricultural activities and the productivity and efficiency of agricultural
activities. This has a bearing on how the households deal with the impacts of HIV and AIDS.
Chapter 3: Theoretical Framework to HVI Development
Draft
3.1 Approaches to HVI Index Development
Two major types of approaches were used in the development of the HVI, i.e., principal
component analysis and the Costa’s fussy set approach to multidimensional analysis of poverty
given composite indicators. These approaches though different, they compliment each other as
far as model development is concerned. Whilst the principal component analysis served as a
dimension reduction tool that was used to reduce a large set of variables to a smaller set that still
contained most of the information in the large set to facilitate HVI computation, the fussy set
approach attached a score on these indicators given the extent, nature and severity of HIV and
AIDS impact and calculated a compounded index to describe the level of vulnerability of each
household.
3.1.1 Principal Component Analysis
This analysis was used solely for refining indicators or variables, i.e., sieving out important
variables to be included in the HVI model. In this way Principal Component Analysis (PCA) was
applied more as a data reduction method to reduce the number of variables used to measure each
impact area, than as a structure (relationship) detection method.
Data from the 2004 study was stored and analyzed using a statistical package known as Statistical
Package for the Social Sciences (SPSS). PCA was done using the same statistical package. After
going through PCA of all identified variables to measure each impact area an output table known
as the Total Variance Explained appeared in the SPSS output file. This table gave the initial
eigenvalues for each component or variable. The table always appears as shown in the example
below.
Component Initial Eigenvalues Extraction Sums of Squared Loadings
Total % of
Variance Cumulative % Total
% of variance
Cumulative
1 1.3 65 65 1.3 65 65
2 0.7 35 100 0.7 35 100
Extraction Method: Principal Component Analysis
The eigenvalues indicated the amount of variance accounted for by each of the
components/variables. For example in the table above the first principal component explains the
maximum variance in all the variables -- a variance of 1.30. The second principal component
explains the maximum amount of the remaining variance -- a variance of 0.70. The two
components explain all the variance in the original variables (1.30 + 0.70 = 2). The proportion of
variance accounted for by the first principal component equals 0.65. The proportion of variance
accounted for by the second principal component equals 0.35.
Draft
A method suggested by Kaiser (1960) was used to reduce the number of variables in the data set
by finding the smallest possible set of principal components which explain most of the variance
in the data set. Kaiser suggests that only those factors whose eigenvalues are greater than 1 are
retained. Hence for the HVI model, components with the highest eigenvalues among the
components or variables suggested in each impact area were used in the computations.
3.1.2 Fussy Set Approach to Household Vulnerability Index Analysis
a) Theoretical framework
The theory proposed for the construction of the HVI largely took on from the work originally
proposed by Costa1. The quest for the exercise, as was the case in the work by Costa, was to
quantify the multi-dimension aspects of the impacts of a health problem on a household. Our
specific quest was to assess at the household level, the impact of HIV and AIDS on agriculture.
The Fussy Set approach was used to analyse the data. The following definitions help clarify how
the approach was used:
One can state that for the population N made up of n households i.e. (N={hh1, hh2, hh3
…hhn}, V is a subset of v households that have some degree of vulnerability to HIV and
AIDS- hence impacted by the epidemic. Thus v≤n and v=0 implies that there are no
vulnerable households, and v=n implies that all households are vulnerable.
One can also break down the vulnerability X into m specific dimensions of impact, and
give a corresponding weight (wi , i=1,…,m)to each dimension. The weights can be
predetermined, or developed using an appropriate function.
The vulnerability of any given household hhi i=1…n to the jth j=1,…m dimension of
impact can be expressed as Xij, and set to take values between 0 and 1 such that 0=no
impact and 1 full impact. A specific formula for calculating Xij is discussed later. Thus
each Xij denotes the degree of vulnerability of household i to the jth dimension of
impact, and Xijwi will be the corresponding weighted vulnerability.
The sum of the weighted vulnerabilities across all dimensions will give the particular
household’s total vulnerability Vhhi to HIV and AIDS, that is:
1 Costa, M. (2002). A Multidimensional Approach to the Measurement of Poverty: An Integrated Research
Infrastructure in the Socio-Economic Sciences IRISS Working Paper Series No. 2002-05; and Costa, M. (2003). A Comparison Between One-dimensional and Multidimensional Approaches to the Measurement of Poverty An Integrated Research Infrastructure in the Socio-Economic Sciences IRISS Working Paper Series No. 2003-02.
Draft
ij VhhwXm
j
wj
m
j
11
/
It is also possible to sum down the dimensions and calculate the particular dimension’s
contribution to vulnerability to HIV and AIDS.
For the HVI, the sum of the weights were conveniently set to 1001
m
j
jw . The
weights were preset as discussed below.
The Household vulnerability index was calculated by applying the theory discussed above to data
collected by household questionnaires, observing a number of steps as shown in section 3.2.
b) Setting the HVI Dimension Weights
Weights or scores for each of the five HVI dimensions are not necessarily the same across every
community, district, province or country but depend on the livelihood strategies of the
community being investigated. The weights are preset after taking into consideration the
importance of each livelihood asset in the lives of the target community. For example a
community that is well networked to a number of markets and that is actively participating in
those markets would put more weight on their financial capital compared to natural capital. On
the other hand a community that is near a rich forest and survives mainly from harvesting that
forest would also put more weight on the natural capital asset. So the contribution of the
particular livelihood asset to the community’s way of life is of importance in presetting the
weights.
3.2 Step taken in designing HVI Computations
Step 1: Developed overall framework for HVI – This involved critical reviewing of different theories
supporting vulnerability analysis and methods that have been used in the past. This formed the
foundation of the HVI conceptual thinking and proposed methodology for quantifying
household vulnerability due to an external shock.
Step 2: Linked theory to practice – Involved identification of data sources and variables that can be
used in HVI analysis. This was done by conducting a detailed literature review on the impact of
HIV and AIDS on the 15 selected impact areas. The review mainly focused on empirical
evidence that was observed through different studies in the Sub Saharan Africa. Through the
Draft
literature review several variables were identified as indicators that would be used to test each
impact area.
Step 3: Defined dimensions and identified impact areas (hypotheses) that could be used to test vulnerability - Five
dimensions of impact that is, the five livelihood assets that are affected by HIV and AIDS were
identified using results from the 2004 FANRPAN impact study. The study was able to reveal that
HIV and AIDS affect the entire livelihood of rural households such that any effort to address the
pandemic should consider all the five livelihoods assets, i.e., human, financial, physical, natural
and social capitals. These were taken as dimensions of impact. These dimensions are defined by
different impact areas that were also identified during that study and are outlined in Chapter 4.
The impacts areas help explain how vulnerable households are to each of the dimensions.
Step 4: Assigned weights and transformations to impact areas using evidence from other statistical models and
previous studies - The weights given to each impact area were determined as mentioned in section
3.1.2 (b). According to the HVI model , the sum of the weights are set at 100 so that the
individual HVIs take values between 0 and 100, with 100 being full impact on the basis of
selected dimensions. The higher the value the more vulnerable is the dimension and hence
defining impact areas within the dimension.
Transformation of selected variables was done by setting an appropriate scale so that each
variable falls between 0 and 1. This process allowed for use of a similar scale which made
comparisons of results possible. The transformation approach used depended on the variable of
interest and how information on that variable is collected. A very simple approach has been used:
an attribute of 1 is set whenever impact is felt 100%, and 0 if not. For dummy variables, that is,
simple yes or no answers this can be achieved by setting 0 and 1 according to the direction of
impact of that variable. Other values between 0 and 1 are set according to relative severity of
impact. Other more robust techniques were also used. For instance for dependency ratio, a value
Y for a "standard normal dependency ratio for the particular community say, from an unaffected
household can be used in the formula X-Y/(Max X -Y). Another approach that was used to
transform the data to 0, 1 without losing its general distribution, was by dividing by the range.
Table 3.3 shows all the transformations that were used for each variable in the HVI computation.
See Annex 1.
Step 5: Determined the contribution of each dimension/impact area to the HVI - This was computed using a
simple formula. This is given as:
Draft
Weight of an impact area to a given dimension = Extent to which a given impact area
determines a given dimension (impact coefficient) x normalized value of the variable selected to test the
impact area x the total impact score or weight for that dimension
The impact score or weight and the impact coefficients were derived through a process described
in Step 4 (b).
Step 6: Calculated, the weighted vulnerability to a given impact area (per household), and the total vulnerability for
each household, and for the community - The weight of each dimension of a given household is a
function of the summation of the impact area scores in that dimension. All the impact scores for
the impact areas defining each dimension are aggregated. Then the total score is computed by
aggregating the scores for all the five dimensions per household. The HVI is then computed by
dividing the calculated total scores per household with the total possible score when there is full
impact which was set at 100.
Draft
Chapter 4: HVI Methodology
4.1 HVI Data Requirements
Data required for HVI computation is no more than most data that is collected in baseline or
impact surveys by development organizations and governments through its different
departments. This data can be classified as primary and or secondary data.
a) Primary Data for HVI Computation
Most data used in HVI generation is regarded as primary data which should be collected on a
regular basis. This data is mostly on the livelihoods of households and communities under
investigation. For example during the development of the model, data used was largely drawn
from that collected in the 2004 FANRPAN study focusing on the identified 15 impact areas as
shown in the table below. The table also shows corresponding indicators that were tracked for
each impact area and used in HVI computation. A closer look at this table shows that most
organizations already collect most of the information used in the computation of the HVI.
Impact areas and variables to be used to measure household vulnerability
Impact areas Indicators
1 Optimal farm-household production decisions
% cash cropping Changes in input use (especially fertilizer)
2 Changes in household demographic structure and labour availability
household size dependency ratio No sick members
3 Changes in household productive physical capital assets
farm implements number livestock index
4 Sustainability of household food and nutrition security (food accessibility and utilization)
Number of meals per day Regular household income Household nutrition diversity Number of regular food sources
5 Impact on household Market Access
Net revenue from asset transactions Net revenue from crop transaction Distance from nearest market place
6 Agricultural extension services Access to extension services Absenteeism from extension meetings
7 Changes in household income and expenditure patterns
% expenditure on health care % expenditure on food
8 Changes in household productive financial capital assets
Savings withdrawals and deposits made Amount of credits received and level of interests charged
9 Impact on household investment choices
Farm equipment purchases and sales Livestock purchases and sales Land purchases/improvements Types of crops grown
Draft
10 Access to behavioral change information
Sources and quality of information on HIV and AIDS Sources and quality of information on agriculture
11 Changes in household productive natural capital assets
Access to land % land utilization
12 Mobility of household members Household disintegration due to HIV and AIDS
13 Gender implications % of female-headed households % of child-headed households
14 Support networks
Number and type of support from government, NGOs and local community Remittances from relatives Number of social networks
15 Environmental degradation Use of forest products Household access to water resources
However most important is the fact that the design of the HVI approach allows for flexibility in
the type of indicators used in the computations. The methodology allows for the use of different
indicators depending on the scope of the assessment. Thus data on new questions can be
collected and used in HVI analysis.
Generic HVI questionnaire (Annex 2) - Data for HVI computations is mostly collected
through semi-structured interviews with respective households. There already exists a generic
questionnaire and as indicated in the section above the questionnaire collects information on
basic livelihood of a household focusing mainly on the five livelihood assets. The questionnaire is
4 pages long and can be administered at most in 30 minutes. It is divided into 7 sections as
follows:
A. Household data - This section collects information on household demographics such as age,
gender, education, health, migration, etc.
B. Asset ownership – Collects information on physical and financial assets with the household such
as land, inventory e.g. fertilizer and seeds, quantity of food available, farm implements,
infrastructure e.g. water source, livestock assets, etc
C. Use and management of the environment – Collects information on use and management of natural
resources and the environment by the household.
D. Use of financial resources – Collects information on the use of financial resources within the
household
E. Household nutrition diversity – Information is related to food diversity within the household.
Usually the recall period for this section ranges from 1-7 days before the survey.
F. Sources and quality of information on HIV and AIDS – This collects information related to
household sources of information on HIV and AIDS
Draft
G. Social support networks – Information in this section is related to household’s access to social
support systems in their community
b) Secondary Data
Secondary data sources are also important when using the HVI approach. Information about the
communities where the study will be conducted is very important in HVI computation.
Secondary sources will mainly provide information regarding rainfall patterns, soil quality, prices,
etc. Although this information is not collected regularly, it should be updated every season.
There already exists a semi structured questionnaire for collecting community related information
that can be administered in less than 15 minutes.
4.2 HVI Database
The HVI has been built into a programmed database which allows HVI generation on a click of a
button. The database is user friendly and comes with a manual that users can refer to whenever
they have issues they want to clarify. Currently the HVI database is a Microsoft Access database
with SQL properties that allow for stability and handling of large records of data.
The HVI database has two interfaces i.e., the user interface that allows for viewing of existing
data and has provisions for data entry, view and print HVI and other reports such as distribution
lists, etc . The other interface is for administrative purposes. It allows a user with administrative
rights to basically manage the database i.e., to change certain factors within the HVI programme,
add or delete users, set user rights, process and update the HVI.
4.3 HVI Web Portal
FANRPAN has established an HVI web portal which acts as an online resource allowing
stakeholders to interact with each other on the HVI through the use of a discussion an e-forum.
The portal is a centralized facility that people can access from anywhere in the world whenever
they have access to the internet. The HVI web portal can be accessed on the following link
http://www.developmentdata.info/fanrpan/hvi.
Through the web portal stakeholders interested in the HVI can get regular updates about the
HVI through automated emails. The portal has provisions that allow stakeholders to subscribe to
a mailing list to allow them to receive the updates. The HVI website offers an online resource for
access to background information and development process of the HVI. Some published
reports on the HVI including workshop proceedings can also be posted on the website for easy
access by those interested. There also exist a comprehensive search facility to allow easy access to
Most importantly the HVI website has provisions for online computation of the HVI by
stakeholders. There is an online form provided on the website and thus by completing this form
and submitting their details, stakeholders can view their HVI online on a click of a button.
Draft
Chapter 5: Pilot Testing of the HVI
5.1 Pilot Testing of the HVI
Pilot testing of the HVI was conducted in Lesotho, Swaziland and Zimbabwe. The major
objective of this exercise was to access the applicability of the HVI under different conditions
and compare how the index would perform compared to traditional vulnerability assessment
methods.
A total of 460 households were interviewed and data analyzed using the HVI model. In
Zimbabwe the pilot test was conducted on an on-going HBC programme in Marange and Seke
communities. Affected and less affected households were defined by programme managers as
those with and without a person living openly with HIV respectively, plus those with orphans.
The study aimed at a 50:50 ratio of “affected” and “less affected” households. The survey was
successful in achieving this target. A total of 235 households were sampled. Affected and less
affected households were defined by programme managers as those with individuals openly living
with HIV and or AIDS and those with orphans. Approximately 49.8% of the households in the
sample were characterized as HIV and AIDS affected, while 50.2% of the households are less
affected by the pandemic (Fig 3).
Fig 3: Distribution of affected and less affected households by district
A vulnerability analysis was conducted on data collected in the two districts. Results in table
below show that approximately 51.3% of households in Seke district were at the coping level,
while 48.7% fell in the acute level of vulnerability. In Marange district, 29.5% of sampled
households were in the coping level, while 70.5% were in the acute level of vulnerability.
However, there were no households lying in the emergency level in both districts. On average,
49.2%
49.4%
49.6%
49.8%
50.0%
50.2%
50.4%
Seke Mutare Total
Affected
Less Affected
Draft
most of the households in the survey sample fell under the acute level of vulnerability i.e. 60% of
the households, while 40% are classified as coping.
Model Parameters used in Zimbabwe
HVI Level HVI Range Situation of households
Seke Marange Total
Vulnerability Level 1 HVI =<0.5 Coping level 58
(51.3%) 36
(29.5%) 94
(40%)
Vulnerability Level 2 0.5<HVI<0.75 Acute Level 55
(48.7%) 86
(70.5%) 141
(60%)
Vulnerability Level 3 HVI => 0.75 Emergency Level
0 0 0
Total (n) 113 122 235
However further analysis showed that it is possible to use the HVI to analyze inclusion and
exclusion errors in programming. This is particularly important in light of livelihood and
development interventions. Fig 4 below show that there were some households that were
included in both the HBC programmes in Marange and Seke communities yet their vulnerabilities
show that they should not have been included. They are also some households that were
excluded from the programme yet they were most deserving.
Fig 4: Exclusion and Inclusion Errors in studied HBC programming
0.00
0.10
0.20
0.30
0.40
0.50
0.60
0.70
0.80
0 1 2 3
1=affected 2=less affected
HV
I
Coping level
HHds
Acute level
HHds
Emergency Level
HHds
These households were included but should not
These households were excluded but should be included
Draft
Results from the analysis also managed to show that it is possible to use the HVI analysis to
compare vulnerability levels across households, interventions, communities, regions and even
countries (Fig 5). This is particularly important in vulnerability assessments.
Fig 5a: Marange District Fig 5b: Seke District
5.2 Issues to be addressed in order to reduce vulnerability
The pilot testing also gave insights of issues that need to be addressed in order to reduce
vulnerability of households to the impacts of HIV and AIDS on agriculture and food security. In
light of the adverse impacts of HIV and AIDS on household assets, households need to manage
their assets efficiently so as to reduce their vulnerability levels. In addition to efficient
management of assets, households must be able to protect their asset base. This include
improved community based natural resources management, transfers of improved mechanisms
for resource protection and management, and mitigation measures for protection from risks.
Protection of human capital would focus on health and nutrition through provision of health
services. Ensuring access to micro credit such that the borrower is not made vulnerable by
indebtedness can protect financial assets.
In addition policies to expand asset base of the vulnerable are necessary. For example, land
reform where distribution of land is inequitable and granting of tenure rights to groups squatting
on public lands in urban areas or provision of credit loans to squatters to help them own the land
that they live on. Human capital can also be improved though provision of free or subsidized
education and health services as well as by ensuring sufficient levels of food and nutrition.
No. of house
hold
s
.75.69.63.56.50.44.38.31.25
50
40
30
20
10
0
Std. Dev = .08
Mean = .53
N = 122.00
Household Vulnerability indices
No. of house
hold
s
.75.70.65.60.55.50.45.40.35.30
40
30
20
10
0
Std. Dev = .08
Mean = .49
N = 113.00
Household Vulnerability indices
Draft
The quality of the expanded asset base is important to ensure the resilience of a community once
relieved of their vulnerability. Hence there is need to improve the quality of the expanded asset
base. Natural assets such as land have to be made more productive through provision of
agricultural extension services in rural areas and skills training in urban areas for home-based
activities. Physical infrastructure is best maintained when the community co-invests in its
provision and maintains the facilities by themselves.
5.3 Major Findings and Recommendations from Pilot Testing
The pilot testing produced the following major findings:
HVI proposes a new and robust way of tracking and analyzing vulnerability. However
adapting the model for wider community participation and a cost benefit analysis for
adopting the methodology in a programme is necessary.
The flexibility within the methodology allows for equitable use of limited resources by
beginning at the tail end of vulnerable households, i.e., with the most vulnerable.
The following conclusions and recommendations were drawn from the pilot testing:
a) Policymakers and development specialists can use the HVI to design, plan and
implement comprehensive and well targeted public funded social protection
programmes that specifically reduce household vulnerability and improve food security
among HIV affected families.
b) There is potential value addition from further research to provide practical solutions to
developmental problems. Researchers working on different areas such as HIV and
AIDS, agriculture, food security, and health can utilize the model to come up with
practical solutions in their studies. Such research could also help improve on the model
which will be a positive development.
c) The model will not be useful without support from policymakers in the SADC region
and funding partners who provide the financial resources for its application. Efforts
should be made to ensure that there is a continued dialogue especially with the
responsible policymakers so that there is unwavering government support in the
promotion of the model in government and other development partners’ programmes.
d) There is a genuine need to build upon existing opportunities to form strategic
partnership with organizations willing to take the HVI to another level. This will help
promote the wide use of the HVI in the development community. The partnerships will
provide an opportunity to apply the HVI on a wider spectrum and maybe in different
environments. This is important as it then assists in evaluating the universality of the
model.
Draft
e) A major challenge is to target the benefits of social protection transfers to the most
needy – especially in the context of HIV and AIDS. Vulnerability should not be used
synonymously with need as it should reflect the likelihood of a particular outcome
arising for a defined group in the future.
f) There is need to operationalize the HVI within the context of work conducted by
development oriented organizations. Such an opportunity will ensure further evaluation
of the model and customization and development of an operational guideline on how to
apply the HVI in development work.
Draft
Chapter 5: Application of the HVI
5.1 Where is the HVI applicable?
The HVI is applicable in development work meant to improve social protection systems for the
poor and most vulnerable groups in African societies. FANRPAN believes the HVI can be
applied in programming and targeting of development interventions, vulnerability assessments
and monitoring and evaluation of development programmes.
5.1.1 Programming and targeting- The HVI is expected to improve targeting of food
aid and other development interventions. The HVI improves programming and
targeting of mitigation responses in three ways:
It makes it possible to classify households according to their level of
vulnerability thereby allowing the targeting of the most affected households
first. This is particularly important where they are limited resources available for
interventions.
It makes it possible to identify the source of vulnerability within a household’s
livelihood thereby making it possible to come up with specific intervention
programmes targeted at addressing these problems. Where a relief agency has a
package of interventions e.g. for food distribution, income generation or
agriculture production, it is possible to use the HVI to allocate which
households qualify for which intervention, within the same community. It is also
possible during implementation to then check if a household has graduated
from a given level of vulnerability and thus no longer qualify for that particular
intervention, and assign it to another intervention if available. The reverse it
also true. The HVI makes it possible for relief agencies and development
organizations to start from the communities and then design appropriate
programmes that suit community and household needs.
The HVI makes management of programmes easier and enjoyable. The
approach makes it possible to deal with pipeline breaks that are always
happening in programming by providing an objective way of scaling up and
down of programmes. If operational research proves the HVI to be an
acceptable measure of vulnerability by communities then the approach has the
potential to reduce social tensions that are a result of these pipeline breaks.
Draft
5.1.2 Vulnerability Assessments –Use of HVI will improve vulnerability assessments in
the SADC region and beyond. Previously, vulnerability assessments have not been
quantitative and therefore made effective targeting of interventions difficult.
Therefore the HVI offers an opportunity for quantitative inclusion of HIV and
AIDS and other vulnerability shocks into regional vulnerability assessments
conducted by Vulnerability Assessment Committees (VACs). Categorising
household vulnerability is important as households affected by HIV and AIDS are
not at the same level of need, neither is that need synonymous with vulnerability.
Empirical evidence shows that not all HIV-affected households are food insecure
and that many unaffected households are actually food insecure. This runs contrary
to generalised labelling of AIDS-affected households as vulnerable and in need of
food security.
5.1.3 Monitoring and Evaluation – The HVI will make monitoring and evaluation of
development interventions easier. With the growing call for accountability given
limited food and financial resources there is a need to demonstrate how resources
have been used and what impact they had on the lives of the target population. The
HVI approach makes it possible to track improvements in the lives of the affected
communities as a result of the intervention. Furthermore when HVI data is collected
over time using prescribed methods, it is possible to compare across communities
and check the trends over time. This is important to track the effect of any
interventions in the region.
Draft
Chapter 6: Reception of the HVI in the SADC Region
6.1 Introduction of HVI to regional development partners
The HVI concept was introduced to development partners in the region through various forums
such as the Agricultural Coordination Working Group convened by FAO in Zimbabwe and
Vulnerability Assessment Committees in Lesotho and Swaziland. A regional workshop on the
HVI was held on the 4th of September 2008 in Zambia and was attended by more that 200
participants from across Africa and beyond. The general consensus among stakeholders was that
the HVI had potential to be a useful development model. Most civil society organizations
including UN agencies felt that the HVI offered an opportunity for improving targeting of
interventions meant to reduce the impacts of HIV and AIDS on the rural society. The Zimbabwe
Vulnerability Assessment Committee, for example, felt that the index could go a long way in
improving the quality of vulnerability assessments. In the initial stages some organizations such
as Food Security Network of Zimbabwe (FOSENET), World Vision, World Food Programme
(WFP), and Food and Agriculture Organization (FAO) indicated their willingness to put the HVI
to test within their programmes.
6.2 HVI Regional Policy Dialogue Workshop
As indicated in Section 6.1 the regional workshop was held in Zambia from the 4th to the 6th of
September 2007. The theme for the workshop was Policy “Triggers” for Agricultural Growth in
Southern Africa. The workshop had two main purposes: 1) to share major findings and policy
recommendations emerging from recent research in Lesotho, Swaziland and Zimbabwe on the
impact of HIV and AIDS on Food Security in Southern Africa and how the Household
Vulnerability Index (HVI) can be used to shed light on the different degrees and levels of
household vulnerability introduced by the pandemic and 2) to obtain recommendations for
future work in this area by FANRPAN.
Several emerging policy issues around research on the impacts of HIV and AIDS on agriculture
and how the HVI can be utilized in this context were proposed. These were as follows:
Policy makers and development planners can use HVI to design and implement targeted
interventions: The current challenge is to develop the framework on how this can be
applied
Funding for the integration of HVI into new and existing interventions is required. The
identified need for operational research on the HVI requires resources for the
implementation.
Draft
HVI should be packaged for three identified groups of end users of the tool in the
region. These groups are individual researchers, civil society organizations and
international agencies.
The workshop concluded that further operational research and wider application of the
methodology in at least three countries should be done in order to ensure that all the operational
constraints of using the HVI are dealt with in preparing of a wider rollout of the tool.
FANRPAN was also tasked with spearheading a regional adoption and implementation plan for
the HVI. This will then form the foundation upon which further work on the HVI in the region
will be based. This adoption and implementation plan should have time frames which will guide
activities and the achievement of set objectives.
6.3 Use of the HVI by FOSENET and World Vision International
Although a number of organizations have shown interest in the HVI only two have taken it upon
themselves to put the HVI to test. These are the Food Security Network of Zimbabwe and
World Vision International. These organizations in partnership with FANRPAN availed some
resources for use and operationalising of the HVI within their context.
a) Food Security Network of Zimbabwe (FOSENET)
They were the first organization to put the HVI to test in 2007. FOSENET commissioned a
study for assessing the impacts of HIV and AIDS on agriculture and food security in two
districts in Zimbabwe, using the HVI model with a view of coming up with intervention
strategies that best suit these two districts. Results from this showed that the HVI was able to
quantify vulnerability of studied households with relative accuracy of 80-90% when compared
with other methods such as community ranking. Were deviations were found these were mainly
due to errors in data collection and misrepresentation of information. The study was able to
categorize households into three levels of vulnerability and proposed interventions were derived
from the HVI analysis.
b) World Vision International
In March 2008 World Vision International commissioned an operational research in Swaziland,
Lesotho and Zimbabwe meant to use and evaluate the applicability of the HVI in its food
transfer programmes. The study aims at developing an information management system for
programme implementation in the pilot sites; and determining the HVI’s effectiveness in
identifying the most vulnerable households appropriate for a specific intervention. The specific
objectives are to assess the effectiveness of the HVI as a targeting/ management tool in dealing
with pipeline breaks and project transition i.e. scaling down, up or changing intervention types; to
determine the cost effectiveness of using the HVI relative to the current practice of targeting; to
Draft
determine the appropriate frequency of updating information in order to avoid inclusion and
exclusion errors in programmes; and to assess the level of acceptability and satisfaction of the
HVI as an objective targeting tool by both communities and field staff.
The operational research is longitudinal study spanning a minimum of two years; the study covers
a minimum of one WV operational Area Development Programme in each of the study
countries. The study is divided into two phases with the first phase being a systematic effort to
develop data collection and management systems for the selected areas. The second phase then
focuses on critical factors for continued use of the HVI, i.e. the frequency of updating
information which is intricately connected to the cost. WV as an organization seeks to empower
communities; as such it endeavours to have effective community participation in the refinement
of the HVI.
6.4 Other partners in the SADC region that would find the HVI useful
Several other partners in the SADC region could find the HVI useful in their programmes. These
include:
Non Governmental Organizations (NGO) implementing impact mitigation programmes but not using
a universally acceptable approach to targeting and monitoring of their programmes could use the
HVI for targeting, monitoring and evaluation. The HVI provides a basis upon which
organizations can effectively target and monitor the effectiveness of those programmes. The
model also provides an efficient way of deriving appropriate response packages for identified
vulnerable communities.
Famine Early Warning Systems Network (FEWSNET) – This is an international network whose
mandate is to strengthen the abilities of African countries and regional organizations to manage
risk of food insecurity through the provision of timely and analytical early warning and
vulnerability information. The network uses the livelihood framework to food security analysis. It
is in this analysis where FEWSNET can integrate the HVI to come up with refined results that
efficiently identify food secure and insecure households, zones, countries or regions. This is
important for decision making especially emergency relief planning purposes.
Regional Vulnerability Action Committee (RVAC) - these committees are responsible for conducting
vulnerability assessment in all the SADC countries. Of late there has been a disgruntlement from
the civil society organizations in the region on the lack of a significant consideration of HIV and
AIDS as a factor that affects food security. The HVI provides a method by which the RVAC
could quantitatively include impacts of HIV and AIDS in its analysis. In this way the RVAC
would be able to determine the extent to which HIV and AIDS is affecting food security in the
SADC region.
Draft
Annex 1: VARIABLE TRANSFORMATION PROCESS
Dimension Hypothesis tested Variables for testing hypothesis Transformation
Natural Capital
Soil fertility declines for vulnerable households as application of natural fertilizers declines.
Proportion of field fertilized by natural means. What proportion X of the fields is fertilized by natural means?
2X; CLH:50-100%=0, ALH:0-50%=1,ELH:none=2
Barriers to access to land for agriculture increase vulnerability
Barred from use of land that you used to cultivate No=0 Yes =1
Households revert to the environment for "free" products such as wood when vulnerable. HIV and AIDS affected households rely more on the forest for their livelihoods.
Tree cutting or wood selling as a means of survival, wild fruits collection, environmental management in the presence of sickness or death, quality of water used by household, participation in water or environmental management
2X/5; CLH:answer yes to at most 1 question=0; ALH:answer yes to 1-3 environment questions=1; ELH:answer yes to at least 4 environment questions
Affected households have difficulties in fully utilizing their land due to limited labour and draft power availability. Vulnerable households do not fully utilize their existing land
% of land not utilized due to sickness (X) What is the total land under cultivation (A)? What land is available but not cultivated due to illness or death in the last season (B) ?
Affected households are vulnerable when they have sick members, and the more the number of sick members, the more the vulnerability. Also worse if the sick member is the head of the household.
Proportion of sick members (X). What is the total Household size (Y)? How many members are sick regularly (have been bedridden for at least three different times in the last year, with each bout extending to up to a week? Or have been diagnosed with any of TB, Meningitis, Caporsi Sarcoma, Hepatitis, Pneumonia (Z)?)
X=Z/Y
Households that have productive sick members are more vulnerable.
Who is regularly sick None=0 dependent =1 productive adult = 2 Spouse=3 HH head = 4
Highest possible score
Affected households have a greater number of dependents due to the increasing number of orphans in such households
Dependency ratio (economic burden)X :Number of dependants ({0-15}+{>65} +{bedridden or disabled})/Number of economically active.
Female headed and/or child headed households are less able to cope with shocks, compared to male headed households
Age and gender of household head CLH=0 ALH=3 ELH=6
HIV and AIDS has caused disintegration in affected households
Household members who have moved away due to sickness or death
CLH: 0; ALH:2 CLH: 2
Physical Capital
Vulnerability especially to food insecurity increases with less use of fertilizers
Nitrogen fertilizer use for staple crop(X). What is your land size Y in ha? What is the weight Z of top dressing fertilizer used in the last season in Kg?
X=Z/400Y CLH:X>0.5; ALH:0.25<X<0.5; ELH: X<0.25
Affected households have reduced harvests due to limited labour and draft power
Staple cereal output per capita (X). What is the total household size (Y)? How kgs of Maize were harvested (Z)? X=Z/Y
X=Z/150Y CLH:X>0.5; ALH:0.25<X<0.5; ELH: X<0.25
Households that do not own an ox drawn plough or cart are likely to face difficulties in cultivation, planting and other farming operations.
Ownership of a plough or ox drawn cart Owns a plough and cart = 0, plough only = 1 cart only =2 none =3
Draft
Households that do not own or own fewer cattle and other livestock are more vulnerable due to limited access to draft power and alternative sources of income and nutritious food.
Productive livestock index X = 3c+ G+S+2D. How many Cattle do you own (C)? Goats (G)? Sheep (S)? Donkeys (D)?
CLH: X>6; ALH: 6>X.>3; ELH: 3>X
Affected households adopt unsustainable short term coping strategies which might include the selling of assets such as livestock and farm
Livestock sales index X = (3c+g+s+2d)/(3C+ G+S+2D) How many Cattle do you own (C)? Goats (G)? Sheep (S)? Donkeys (D)? How many Z (=3c+g+s+2d) of each were sold in the last year?
CLH: X<0.2; ALH: 0.2<X<0.5; ELH: X >0.5
Affected households have limited access to extension services due to ill health and inadequate time to devote to such activities.
Access to extension services
Used both = 0; used crop only = 1Used livestock only = 2; do not even know = 3
Financial
Households with little or no savings are more vulnerable
Reliance on bank savings
Every month = 0 In crises only = 1 Do not have many in the bank anymore=2 Do not own a bank account =3
Affected households have fewer sources of regular income due to unavailability or limited number of formally employed members in a household
Regular sources of financial resources Salary (S), Crop Sales (Cs); Livestock Sales (Ls); Remittance from HH member (Rm), No regular source (Ns)
S=0; Rm=1 Cs=2; Ls=1; Ns=3
Affected households have limited access to credit loans due to increased risks and lack of collateral associated with such households
Access to credit loans
Household is part of a community or formal credit scheme= 0 borrow from extended family/neighbour = 1 no access to credit loans at all = 3,
Households with unpaid debts are most vulnerable.
Presence of unpaid debts No=0, Yes= 3
Affected households experience increased expenditure on health care due to the presence of more ill members in the household
Expenditure patterns. Food (F), Non-food basic goods (nF), Health (H), Savings (S), Transport to Clinics (Tc), Transport to Work (Tw), Farming inputs/implements (FI), Do not prioritize/plan (Nm) Other (o) , Beer and recreation (B), School Fees (SF),
FI/S/o=0, Tw/B/SF/F/nF=1, H/Tc/=2, Nm=1
Use of additional resources indicate choices under vulnerability
Expenditure of additional financial resources Food (F), Non-food basic goods (nF), Health (H), Savings (S), Transport to Clinics (Tc), Transport to Work Tw), Farming inputs/implements (FI), Other (o) , Beer and recreation (B), School Fees (SF), Income generating projects (Pr)
FI/Tw/o/B/S/Pr=0, Tw/nF=1, Tc/SF=1, F/H=2
Purpose for selling harvests indicates levels of vulnerability.
Use of revenue from crop sales Food (F), Non-food basic goods (nF), Health (H), Savings (S), Transport to Clinics (Tc), Transport to Work (Tw), Farming inputs including Veterinary (FI), Do not get enough to sell (Nm) Other (o) , Beer and recreation (B), School Fees (SF), Income generating
FI/S/o=0, Tw/B/SF/F/nF=1, Tc/=1, H/Nm=2
Affected households eat less per day due to inadequate food availability
Meals per day
Breakfast, Lunch, Dinner, give 0 for each taken ie 1 meal= 3, 2 meals=2; 3 meals=0;
Draft
Affected households eat less variety per day due to inadequate food availability
Describe the typical food stuffs in meals taken in your household? Maize (porridge/sadza/samp) (St), tea St, sorgum brew St, green vegetable V, wild fruit F, bananas/oranges/apple F, sugar cane St, pumkins V, groundnuts Pr sweet potatoes St, meat Pr, fish Pr,
Give 1 for each category taken, CLH:X>3 ALH: 2<x<3; ELH: X<2
Social Capital
The lesser the number and quality of support channels from external sources, the greater the vulnerability
What support was obtained from Government, NGOs, community and other external support networks in the last 3 months? Give the commonest 2. Food (F), Non-food basic goods (nF), Health (H), Savings (S), Transport to Clinics (Tc), Transport to Work (Tw), Farming inputs including Veterinary (FI), Do not get support(Ns) Other (o) , Beer and recreation (B), School Fees
Tc/H=0; Tw/B/SF/F/nF=1, FI/S/o=2, Ns=4
The lesser the volume of support from external sources the greater the vulnerability
In which areas did support from Government, NGOs, community and other external support networks completely meet households' requirements? Food (F), Health (H), Transport to Clinics (Tc), Farming inputs including Veterinary (FI), Do not get support(Ns) School Fees (SF),
none=2 ; F/H=1; else =2
The more informed a household is, the less vulnerable the household
1. Do you have adequate knowledge to cope with AIDS related illnesses for family members?, 2. Do you have adequate knowledge on type of crops to grow, and when to. 3. In any given season, do you know- in advance- the weather forecasts and use this forecasts
count of "No" answers
Draft
ANNEX 2: GENERIC HVI QUESTIONNAIRE
HOUSEHOLD DATA CARD
Date of data collection: _____/______/_________ Village: _________________________________ Name of respondent: ______________________________ Day / Month / Year
A. Household Data
Full Name National/WVI ID NO
(format: XX-XXXXXXX-X-XX)
Year of birth
Sex M or F
Education2 Employment3 Health4
1 HH Head
2
3
4
5
6
7
85
Record Household members that have relocated within the last three months
9
10
2 Illiterate=0, Some Primary School=1, Completed Primary School=2, Some Secondary School=3, Completed Basic Secondary School=4, Completed Advanced Secondary/Pre-University School=5, Professional College
certificate=6, University Education=7, Other = 8 3 Unemployed/Homemaker/NA=0, Subsistence farmer = 1, School child=2, Artisan/Skilled Tradesman/woman = 3, Petty Trade = 4, Formal Employment=5, Harvesting natural resources(wood, panning etc) 4 Good=0, Regularly (sick at least once every three months) Sick=1, Bedridden=2 Then record the sickness: good health=0, TB/neumonia/Kaposi/=1, Malaria=2, Headache/whole body=3, stomach ache=4 5 For household members exceeding 8, attach additional form.
Household Identification Number: _______/______/______/________/__________ Enumerator: ____________________________________ Country District ADP Village ID
Draft
11
Total in household:
Draft
B. Asset Ownership
Number/Size State6 Asset increased within the last year
Asset reduced within the last year
Land size in acres
Land area under staple crop in acres
Area under staple crop
Amount of fertilizer (AN) used for staple crop in kgs
Land areas forcibly taken from household (barred by relatives, the chief or other community members)
Staple crop available right now in kgs
Meals taken per day
Plough
Cattle
Goats
Donkeys
Sheep
Pigs
Chickens
Agriculture extension services for livestock
Agriculture extension services for crops
Bank account (approximate balance now)
Remittance
community or formal credit scheme
unpaid debt
extended family/neighbour support
Land (acres) fertilized by natural means in the past season
Land (acres) available but not cultivated due to illness or death in the last season?
Source of drinking water
Type of Latrine
C. Use and management of the environment Yes No
Have you ever resorted to cutting down trees and selling wood as a means of survival?
Have you ever resorted to collecting wild fruits because you do not have enough food for three meals a day?
Has sickness or death of a family member prevented you from managing your environment e.g. gully filling, manure collection etc?
Has sickness or death affected the amount or quality of water used by your household? E.g. resorting to collecting water from nearer but unsafe sources or failing to pay for piped water or failing to pay for repairs to safe water sources?
Has sickness or death prevented your household from ever developing or participating in planned water or environmental management projects? E.g. borehole drilling, tree planting, participation in community initiatives?
D. Use of financial resources
Where does the HH spend most of its financial resources (give the most common)
Where would you spend any additional financial resources if they were availed to you: (give the most important)
For what purpose did the household use financial resources from sale of crops from the last season? (give the most important)
1= Food 2= Non –food basic goods 3= Health 4=Savings 5=Transport to work 6=Transport to clinics 7=Burial expenses 8=Farming inputs 9= Beer and recreation 10=School fees 11= IGP 12=other E. Household Nutrition Diversity Has household consumed food within the specified food group in the past week before the survey?
Food Group (1)Yes (2) No
Grains (maize, sorghum, millet, rice, wheat, etc)
Beverages (tea, etc)
Sugars and sugar cane
Oils (avocado pear, nuts and seeds, etc)
Vitamin C- rich vegetables e.g. spinach, green leaf vegetables, potatoes, cauliflower, cabbage
Tubers (potatoes, sweet potatoes, cassava, etc)
Meat bean and fish
Vitamin A – rich vegetables e.g. spinach, carrots, peas, pumpkins, sweet potatoes, squash, butternut
F. Sources and Quality of Information on HIV and/or AIDS Yes No
Do you have adequate knowledge to cope with AIDS related illnesses for family members?
Do you have adequate knowledge on type of crops to grow, and when to.
In any given season, do you know- in advance- the weather forecasts and use this for farming planning?
Do you have access to projects or interventions that can raise income for your household?
G. Social support networks
Part A: Indicate the most useful support your HH got from government, NGO community or other external networks in the last 3 months by writing the appropriate type in the relevant box. (Choose from list below the table) Part B: In which areas did support from government, NGOs, community or other external networks completely meet your households’ requirements
Source of support
Government
NGO Community Other external support
Part A
Part B
1- Food 2- Non –food basic goods 3- Health Savings 4- Transport to work 5- Transport to clinics 6-Burial expenses 7- Farming inputs 8- Beer and recreation 9- School fees 10- IGP 11- Other