Household livelihood trajectories and survival dynamics in rural Uganda

aboriginalconspiracyUrban and Civil

Nov 16, 2013 (3 years and 4 months ago)



Jovita Amurwon_concept paper_110216

Household livelihood trajectories and su
rvival dynamics in rural Uganda


Earlier studies on the impact of the epidemic in developing economies predicted that Human
Immunodeficiency Virus /

uired Immunodeficiency Syndrome (
illness and death would have a devastating impact on household composition, labour
availability and incomes (Barnett et al., 1992; Seeley, 1993; Rugalema, 1999). Over almost
three decades of the epidemic, knowledge has been generated on the causes,

the course and
the impact of the epidemic. In low income countries it is time to assess whether the
predictions made earlier have come to pass. An emerging literature argues that the predicted
outcomes have not been realized (Barnett and Whiteside 2007;
eegle and De Weerdt 2008;

et al., 2010). What has led to these different assessments of impacts? Why on the one
hand are ther

assessments of the impact of HIV/AIDS which are deeply pessimistic about its
impacts while there is also emerging evidence

that the impacts have not been as nearly severe
as anticipated: the rural economy has not collapsed and rural populations are rising. This
raises questions about how causal links between infection to households and wider impacts
have been assessed and un
derstood. What methods have informed these studies and what
evidence has been used to support their claims?

This study aims to investigate this apparent contradiction between predicted impacts of the
HIV/AIDS and its effects on specific households. It
will draw on evidence on the trajectories
of rural households using empirical household
level data from Uganda. Of key interest will be
the differences in social and economic resources and outcomes of households. This paper is
constructed in 5 sections. T
he first section provides a primer on HIV/AIDS, which includes;
the history; methods of transmission; treatment, including current status; effects, both medical
and non
medical related; and, definition of some key terms used in disease occurrence. The
nd section presents trajectories of four household case studies from the south
western part
of Uganda. It describes the characteristics of the households (demographic, composition, asset
ownership); key shocks (sickness, death, food
shortage, property loss
); household responses
and the current household status in terms of social and economic and identifies some key
themes. In the third section these themes are used to explore key aspects of the literature on
HIV/AIDS impact. This section is engaged in teasi
ng out the possible reasons why there are
mixed views on the impact of the epidemic, which include, the methods used, concepts and
how they are misused, location and,
the use of the
as a unit of analysis
. The fourth
section highlights the theori


concepts that are relevant for this study and how they will

Jovita Amurwon_concept paper_110216

be used. It uses the evidence from the case studies. Particularly the theories of agency and
structure highlight the flexibility and the constraints that households face in decision making
as they strive to meet their consumption and income needs. It also shows how the resilience
and adaptation concepts are relevant using the evidence from the case studies. The methods
that will be used in the study are discussed in the fifth section. The st
udy will go beyond the
individual and household level data and collect information on political, social, economic,
cultural and historical contexts in which these groups of people are situated. This creates an
understanding on the contextual factors of cau
sality that run across society creating
circumstances under which HIV is spread or prevented. The sixth section provides
background information of the data sources, the organization that has collected the data; the
data content, the data shortcomings and t
he additional data requirements. The final section
provides the research plan including duration, deliverables and the thesis format.

Box 1.
Basic facts about HIV/AIDS

History and responses to HIV/AIDS, the “Slims Disease”

About the Human Immunodeficiency Virus /

Acquired Immunodeficiency Syndrome

The human immunodeficiency virus (
HIV) causes AIDS. The HIV virus destroys a type of defence cell in the body
called a CD4 helper lymphocyte. These lymphocytes are part of the body’s immune system. As HIV destroys these
lymphocytes, the infected person becomes susceptible to infections tha
t wouldn’t normally occur. These may include
tuberculosis, lung infections like pneumocystis carinii pneumonia (PCP), and blood vessel tumors like Kaposi’s
sarcoma (KS). This is due to immune deficiency and it’s this condition that is called AIDS.

HIV is
transmitted from an infected person through body fluids like blood, semen, vaginal fluids and breast milk. The
common modes of transmission include; unprotected sex with infected persons, sharing needles like for narcotic drug
injection; people with other

sexually transmitted infections (STIs) like syphilis, Chlamydia,
, have higher
risk of getting HIV during sex with infected partners; a pregnant woman with HIV can pass it to her baby during the
birth process or from breast feeding; blood trans
fusion with infected blood; sharing skin cutting or piecing objects like
razor blades, circumcision objects may have been used by an infected person.

Currently there is no cure for HIV and AIDS. However as the science community learns more about the natur
e of this
virus, drugs have been discovered that are able to delay and even prevent the progression from HIV positive status to
developing full
blown AIDS

manifested through various clinical symptoms, opportunistic infections such as TB, as
well as visible

in lab tests, such as t
cell counts. This has been done by producing “highly active” combinations of
medications. These medications can limit or slow down the destruction of the immune system, improve the health and
quality of life of people living with H
IV, and, when adhered, reduce their ability to transmit HIV. People with the HIV
virus are now living longer (Murphy

2006), even for decades.

The best way to avoid HIV infection is to avoid practices

that would involve exposure to infected body fluids, i
unprotected sexual intercourse or sharing needles to inject drugs (Marks et al., 2005). If this is not possible

, numerous
health organizations have shown

that the use of latex condoms during vaginal, anal or oral intercourse can significantly
uce the risk of HIV transmission; HIV
positive pregnant women can take medications that can reduce the risk of
HIV transmission to her child; and injection drug users should not share needles or injection equipment.


Jovita Amurwon_concept paper_110216

Between 1983 and 1985, researchers from the Pasteur Institute in France and Dr. Robert Gallo
from the USA

isolated similar retrovirus which the International committee of Scientists later
renamed to be the HIV virus (AVERT 2006). By 2001 it as estimated that 21 million people
had died of complications related to HIV and AIDS, of which 81% of these deaths were

Africa (see box 1 on the basic facts about HIV/AIDS).

In Uganda, Doctors identified a surge in the cases of severe thinning and an increase in fatal
opportunistic infections. They referred to this condition as “slim disease” and in 1982, the
first A
IDS case in Uganda was diagnosed establishing the link between ‘slim disease’ and
AIDS (Serwadda, 1985). By 1986 the prevalence rates (the proportion of a population that is
affected by the disease at a specific time) peaked at 29% in the urban areas of Ug
anda (New
Scientist, 1990). The spread of the disease was perceived to be associated with “promiscuity”
and the “quick solution” was as a prevention message to
e faithful or use
(ABC). The ABC message is said to have contributed to a drop

in the prevalence rates to 5%
in 2001 in Uganda (STD/AIDS Control Programme, 2002).

Since then there has been a

trend of a
n increasing prevalence
in Uganda (AVERT, 2011).
The current prevalence is estimated to be 6.5% among adults (UNAIDS, 2010).
increasing prevalence is associated with; antiretroviral (ART) roll out in 2004, which has
resulted in complacency


and ‘normalisation’ of


in the Ugandan population

os Global AIDS Programme, 2010)
; and, the current, American backed
, prevention
strategy of ‘Abstinence only’

condom use’

reduces condom availability (The
Independent, 2006). This last point will be discussed in more detail in the section on concepts
and theories.

The Ugandan ABC and now A

only HIV/AIDS st
rategy reduces prevention responses to
being wholly concerned with sexual transmission of the virus and a matter of free choices by
autonomous, empowered individuals (Smith, 2004; Hankins, 2010; Henderson et al., 2009).
Renewed attention needs to be given
to the social and economic context in which behaviour
takes place

of which sexuality is part (
Barnett and Whiteside, 2007
; UNAIDS 2010).
ndividual and groups of people are continuously shaped by internal and external factors
which position them in certai
n trajectories of life. These are cultural, economic, political,
gender, and history. The household case studies below

the dynamics that shape the
trajectories that individuals and households take.

One household has an adult ill (25009), one
has ha
d deaths

(some reported to be AIDS

of adult
household members
both resident

Jovita Amurwon_concept paper_110216

and non
resident in the household

two have had no death since 1990 (23221 and

though one
has a son, living close to her, who is currently infected with HIV.


Jovita Amurwon_concept paper_110216

Table.1 Household case


Economic status, capacity to meet household needs
and source of household income

Jane’s Household (15002)

husband died before 1990, after one
year of illness, leaving her with 1 son. She sold coffee to
cover hospital and funeral expenses. Her son did not go to school because the tuition fees could not
be afforded. Now she is a 49
year old widow in poor health
who lives with an 8
year old grandson
whom she adopted in 2005. Until 2000 she and her son lived in the same house farming coffee and
other crops on the 3.0 acres of land they own for food and income. In 2001 the household split when
Jane’s son built his
own house on the same compound and married. They now cultivate separately
with the son owning 2.5 acres of the land but had shared the coffee plantation.

In 2005, Jane’s daughter
law died while delivering the second child. Jane adopted the 3 year
grandson. Her son fell sick and Jane took care of him. She borrowed money from a neighbour to pay
hospital bills, which she paid back by providing her labou
r in the farm .In 2006 her son married
another woman and by 2009 had three children. Jane stopped making use of the coffee farm because
her son’s household had grown and had more food needs. The half acre she was cultivating was not
sufficient to provide f
or her household (she and her grandson). Jane prepares one meal in a day due to
food shortage. She began to work in 2009 as hired labourer for cash and food but her health limits
her working capacity and dry weather led to a low demand for labour. Jane t
up to work as
hired labour with a friend, sharing the proceeds equally. When she was too weak to cultivate and
work as labour, her neighbours, friend and sister provided the food. When her son’s health was better
(he is now on ART), he sometimes cont
ributes food to Jane’s household. Jane also sends her
grandson to his ‘parents’ for a day so that he will have his meals there.

Jane’s house collapsed in 2008 due to poor maintenance and she had to move to an abandoned house
in the neighbourhood.

Economic status

Poor, female headed

In the 1990s

Declining social economic status

Loss of assets (sold coffee for funeral and hospital expenses)

limited physical and human assets (small land size, small household
size of 2 people)

In the 2000s

ll health (son and herself)

asset loss (lost ¾ of land and coffee plantation due to fragmentation
of the land)

food shortage (one meal aday)

poor living conditions (house collapsed)

Capacity to meet household needs is low

food from own farm is not s

she can’t hire out her labour all the time for off
farm income due to
low labour demand caused by the dry season

she has poor health (she falls sick)

Source of food and income

joint off
farm labour

Borrowing money (when the son was ill she
borrowed money)

support from friends, neighbours and sister (when she was too she got
extra support in terms of food)

Maria’s household (17028),

Maria’s husband died in 1990, after a short illness, leaving her with 11 children. She had 3 cows

she sold to pay the hospital bills, funeral expenses and some just sold because they would be
taken away from her. Her late husband’s relatives and sons inherited 90% (18 out of 20 acres) of the
land. Between 1990 and 1993, three of her adult children, wh
o had left her household and built in the
land they inherited, died due to AIDS related illnesses. In the same period 3 of her children left the
Economic status

less poor, female headed

sharp fall in
economic status

loss of human and physical assets (deaths of 4 and out
migration of 5
loss of land and livestock)

Low demand on food and income due to the small size of the household


Jovita Amurwon_concept paper_110216

household to get married, 2 others dropped out of school because their school fees could not be
afforded. This
left the household with Maria and 1 daughter with mental illness. In 1994, an orphan
grandson from one of her late children joined the household but schooling could not be afforded for
him. In 1995 two other grandchildren joined the household to help with
work as they studied. Their
parents paid their school fees. Between 2000 and 2002 two grandchildren were born to her mentally
sick daughter who lived with her. The land that Maria had was insufficient meet all the food needs of
her household. This was wors
ened by the drought. She rented 1 acre of land which she paid for with a
share of the harvest from the land.

Since 2009 one of Maria’s daughters who lives in another town sends her money for household needs
including school fees, labour hiring for farming

and bought Maria a cow. Maria also rears a goat she
received from the animal sharing club where she is a member

she will give it back after it has got
kids which she will keep. Maria is now 72 years and lives with 2 grandchildren (15 and 16 years old),
d a nephew (16 years old). She does not borrow land for cultivation anymore.


improving household social
economic status

members joined the household to provide labour

land was borrowed for cultivation

Capacity to
meet household needs is improved

due to extra land and

From 2009

increasing assets

financial support from external sources (the daughter sends money to
support the household)

livestock recovery (bought a cow and received a goat from animal
ng community club)

There is capacity to meet household needs with increased income

Ntonio’s household (25009)

Ntonio, who is now 74 years old, has had ill health with persistent headache since 2002. In
2009 he got worse and is since then bed
ridden. He moved out of his household to live with
his daughter who is a nurse in another town for care. Ntonio’s household
, which he left, is
currently composed of his wife, two sons, two adopted grandchildren (one lost a father and
the other lost a mother), 1 adopted nephew, 1 non
relative who came for lodging close to
school, and 1 niece who joined it due to ill health. Th
ey sold motorcycle, cows and coffee to
take care of his medical bills. Between 2000 and 2006, 4 of his adult children dropped out of
school to start individual businesses. The business start
up capital was provided by Ntonio
through selling coffee and ‘war
agi’ (locally made beer). In 2006, Ntonio’s son died in an
accident leaving behind 2 children and Ntonio adopted the older child. He had left his
father’s household and was living on the land he had been given by Ntonio.

Ntonio owns 10 acres of land part

of which is in a different location where he rears and
grazes cows. His daughter (not the nurse) recently invested in a piggery project for the
household to generate income. They also sell cassava and beans for cash. His children attend
private schools an
d also pay for extra lessons. His household members attend the private
clinics when they are ill. The household hires labour to for much of the farm work. They
have not had food shortage at any one time.

Economic status

off, male headed


A stable economic status

Before 2000

Accumulated physical
and capital assets (

and bananas
acquired land and
, gave business
up capital to the children for independent businesses



increased income

children got employment and businesses
and sent money to the household, crops are sold to generate

Capacity to meet household needs is due to

diversified income sources (on
farm and non

maintenance of farm out

ed the crops to maintain
good yield, replaced the old coffee and banana trees, applied
pesticides to disease infected crops

support from daughters and sons


Jovita Amurwon_concept paper_110216

Namara’s household (23221)

When Namara completed school in the 1970, he got formal employment

for 9 years, 6 of
which he served in the army. During the political instability in 1979, when he was
imprisoned for 6 months, all his property was looted, and his wife left. He started business
in 1981, of selling hides and skins, using start
up capital
from his brother
law and
purchased land in Kampala city center. He stopped business after 2 years and went back to
the village (his current residence) to invest in cash crop farming on 4 acres of land that he
inherited from his parents. He also married

another woman who died 1 year later after
having 1 child with him. He also has 2 children from other relationships. In 1989 he left his
sister in charge of his farm and went back to the city to start business in agricultural produce
using capital he had

accumulated from selling crops. After 3 years in business, in 1992 he
returned to his rural home and continued farming and has never stopped.

Namara lives in the village on his own. His wife runs his business in the city, one of his sons
is formally empl
oyed after completing university education, 1 son is in the university and
the rest have their homes and are doing business. He hires labour to carry out cultivation and
harvesting of his crops. He sells most of his crops during off
season when there is fo
shortage and the prices are high. He uses the income to buy pesticides, hire labour for the
next season, medical care, buy food

to have variety, and transport to attend political
rallies, funerals. His children also give him support especially towards c
onstructing his
house. Namara is currently applying for his pension from the government, because he once
served in the army.

Economic status

off male headed


A stable economic status

Before 2000

Accumulated physical, human and
capital assets (cash crops,
land, started a business in the city)



increased income (children got employment)

maintained crop yield (mulched crops and sprayed
crops with pesticides

acity to meet household needs

due to

income sources (on
farm and non


support from daughters and sons

maintenance of farm out put


Jovita Amurwon_concept paper_110216

From the case studies,
a number of
key themes
can be identified:

these include
the use of
household as a unit of analysis given

is variable, household
, the
significance of
social networks

expanding the boundaries of the household, the capacity of
individuals and households to respond to event

and the dimension of


The household as a flexible and fluid entity

It is evident that household composition fluctuates considerably over time. In the case of
household 17028, in 1990 it had 11 members; in 1993 it changed to 2 members; from 1994
todate it has 4 members,
3 of whom are not part of the initial members. In household 25009,
in 1990 it had 10 members; between 1993
1999 it had 7 members; between 2000
2005 it had
4 members; between 2006
2008 it had 7 members (3 of whom are not part of the initial
members); in 200
9 it had 10 members (including a niece and her children); and to

date it has 7
members. It’s clear from this that the household size and structure varies overtime,
questioning how robust the household
as a
category and
unit of analysis is over time and

how impact assessments at the household level can be made. The flexibility and fluidity in
household size and composition raises issues for methods


units of analysis. This also has
implications for household consumption (as new needs arise e.g. educat
ion, food), labour
resources, assets (e.g. land to increase cultivation).

Then two questions arise;

How do households shift their composition? They do it by members leaving the household
and others joining.

What factors determined this? Members grow and
leave the household or move out to look for
alternative ways of getting resources to meet their needs, and if need be, for the household.
The decision when to move and who should move depends on the availability of resources,
and need for resources.

The H
ouseholds adjusted the size consciously as a response to a need to cover a fall in income
or unconsciously as a strategy to maintain household resource flows. In household 17028,
after the death of the household head and loss of productive assets, they wit
hdrew from school
and married off the adult girl
children. New members joined the household almost a year later
from other relatives and also orphans from the sons who had died. They came seeking for
support in terms of housing, food and education, but at
the same there was a need for
household labour at that time. The return was realised about ten years later when the children

Jovita Amurwon_concept paper_110216

who left the household earlier started supporting the household and helping it through
providing both money and physical assets lik
e a cow.

In household 25009, when the household head fell ill, the adult children started independent
business activities and living separately. This reduced not only the consumption demands on
the household but also increased its income from diverse sour
ces. In addition to that the care
of the father was spread across the network of households. The ill man was taken to the
daughter’s home in another town for better care, while another daughter and sons provided
support towards the other needs of the house
hold. However, this flexibility and agency is
determined by a number of other factors which restrict or increase the choices available.

These case studies show the household as a flexible social unit for production

farming, non
farm and off
farm provision

of food and income for other needs; consumption

that is
meeting the household current and future needs as they arise; and investment

planning for
future household and individual needs. These functions are met jointly by allocating and re
allocating functi
ons over time among household members who may or may not live in the
residential unit.

What factors drive or constrain changes in household composition?

Structural determinants of household responses

The availability of resources

land, labour, in the hous
ehold is one of the factors that
determine its flexibility. For, example, a household with large land holdings is more likely to
have more income from diversified farming activities. With 4 or more acres of land
households 25009 and 23221 have both cash c
rops, food crops and livestock. Household
15002 is limited by choice when her 0.5 acres of land is unproductive especially during the
dry season. However, having assets alone is not enough but having power and control over the
use and allocation of the ass
ets. Household 17028 lost most of the land and all the cash crops
(coffee and cotton), when her husband died, through inheritance to sons and other relatives,
leaving the household with no fall
back position. This is also an issue of gender and power
ions. The policies on property ownership and control in Uganda don’t favour women
Zziwa, 1995) or were not in their favour in 1990 when the widow in household 17028
lost most of her land.


Jovita Amurwon_concept paper_110216

Social support networks

The social interactions

hips of trust, in the case studies display an important role to
livelihood outcomes and life chances. The extended family relations in all the households
presented, and membership in local associations for some of the households like 17028
contribute to th
eir well being. Some of the households, household 15002, sometimes
depended solely on friends, the son, neighbours and a sister to meet basic needs like food.
Support from sons and daughters in households 25009 and 23221 boosted household income.
from a daughter and membership in the livestock
sharing club in 17028 has helped
the household realise a positive trend in its trajectory. Due to a loan from a brother
household 23221, after losing all property to looters, was able to re
its asset capital
and have a positive trajectory. However the social support networks varied by household.
Some households had strong social support, this could have been due to reciprocation from
initial investments. For example the sons who had capital t
o start business in household 25009
were in a better financial position to send support to the household when it was needed.

What effects do changes in household composition have?

The changes in household composition resulted in new needs for the househol
ds as well as
filled the labour gaps in the households. In household 17028, all the children, except one with
a mental incapacity, left the household to look for work and to get married. This resulted in
labour shortage for cultivation. A new group of join
ers composed of a niece, grand children

some of them orphans, joined the household. This group provided for farming but also had
additional needs some of which the household couldn’t meet. An additional amount of food
was needed which the household solved
by renting land. However the need for school fees
could not be afforded. Household 15002 had a grandson join after the son had left. This did
not only create new needs of paying for education but also food shortage. This was because
the household head was
sickly much of the time and the grandson was too young to
participate in cultivation.

How does the timing of events affect impact

Tracking impacts of events that are long
wave in nature is challenging




with time. The impact of ill health in household 15002 was visible

the household
facing food shortage. The impact of deaths in the early 1990s in household 17028 was

the household acquired a cow and all the children were attending school. In

Jovita Amurwon_concept paper_110216

old 23221 there was no evidence that the household had ever lost all its property
during the war in the 1980s.

Assessing the impact of HIV/AIDS on households is complicated. The line between affected
and unaffected households is blurred.
The risk associa
ted with sickness and death of
individuals, including care for the sick, is

spread out

among the social network of the
The ability of a household to recover and the length of time it takes to recover
from stressful events is also unpredictable.

There are other drivers of change in the rural
economy against which HIV/AIDS impacts have to be assessed. The fluctuating prices for
output and inputs;
the long dry spells
and pests
that affect crop

changing preferences of the ru
ral populations caused by education

the young people migrate
to look for non
farm sources of livelihood
the wider community interventions that
impact on
household life trajectories like education
, availability of
; and many others.

of the

wider literature on impacts of HIV/AIDS on rural economies brings up

Literature review

There are divergent views on the impact of HIV/AIDS on rural households and economies.
Some argue that there has been a severe effect on rural economies as suggested when the term
‘New Variant Famine’ (de Waal 2004) was coined. De Waal predicted that popu
burdened by HIV/AIDS apply extreme coping mechanisms including withdrawing children
from school, disposing
off physical assets like land and livestock, etc which make it difficult
for them to recover from shocks.
Another study suggested that all as
pects of food production
are affected
(Clover, 2003).

However there are those that think the impact of HIV/AIDS has been exaggerated
(Patel and
Scott, 2003),

leading to a risk of resources being diverted from addressing more complicated
problems like pove

Ellis points out that communities learn and adapt to the environment in
which they live and know the risks and stresses and these have been responded to through
appropriate livelihood strategies
(Ellis et al., 2003).

Survival is central to rural livel
ihoods as
individual’s are constantly involved in complex and dynamic strategies to meet their
consumption and income needs

they are always ‘doing something’ as Marschke
and group

call it (Marschke
et al.,
2006). The case studies in table 1 above show hous
eholds take action
to maintain production and meet household consumption and other income needs.

study in
Uganda found no differences in socio
economic outcomes between households afflicted

Jovita Amurwon_concept paper_110216

(adult death occurrence) and non
afflicted (
Seeley et al.
, 2010)

2008) found an
increase in household size by an average of 1.2 persons following a single adult death in a
study in Malawi. Others point out that it’s too early to know about the epidemic and its impact
since it is the first of its kind. Barnett a
rgues that given the nature of the epidemic, a ‘long
wave’ event, it will take years and decades to spread through society and generations before
its full impact and responses can be understood
Barnett, 2006

What explains the differences between these two divergent views on HIV/AIDS impact?

One factor may be the methods used. A literature review by Booysen and others, (2003)
indicates that most studies assessing impact of HIV/AIDS at household level used cros
sectional methods as opposed to longitudinal methods. Cross
sectional interviews

which is
usually one encounter with the respondent, examine mainly correlation and cannot handle
complex time bound cause effect relations (

and Christopher, 2006), a
s household
trajectories presented above illustrate, and therefore fail to detect capacities of households to
respond to events. Longitudinal studies have the advantage of observing events during the
study period and other changes and events in the wider
community which also affect the
households. The

(1995) advices a minimum of 3 years for conducting longitudinal
studies, while Barnett warns that it will be upto 30 years before we can gain knowledge of the

suggests using ‘mixed
methods’ (Cresswell, 2003) that combines
participatory approaches with numeric survey data. This provides a rich portrait of how
households are experiencing and handling the epidemic (

2004). Lucia Da Corta adds,
the cause
effect analysis should not be

limited to the agents understanding and definition of
events, but should be broadened to capture events external to the group under study
(Lucia Da
, 2009
). These include climatic conditions, market situation for labour and agricultural
products, po
litical situation, etc.

There is also a central problem, as the case studies illustrate, on the use of the ‘household’ as
the unit of analysis to recognise, record and analyse the impact. Many studies have used
uncritically the household as the unit of an
alysis (
, 2007; Bachmann et al., 2003;
Sengozi 2009, Beegle, 2005
, Casale et al., 2009).

The growing evidence indicates that a
household is not the best unit of analysis to assess impact and wellbeing (
Haddad et el. 1997
Hosegood et al. 2007). The d
efinition of a household is clear (Yanagisako, 1979)

eat in one
pot, share resources, activities and sleeps under one roof, but it does not fit into the complex
realities as shown in the case studies above and it does not mean that it is fixed. The externa
links that the household maintain with relatives, other household members or part of the

Jovita Amurwon_concept paper_110216

family influence the trajectory it takes and the social and economic outcomes. In the same
way, the impact of an event of death, ill health or other event that thre
atens the income flow
of the household can be shared or absorbed in the wider network of the group to mitigate its
effects. The head of household 25009 has been ill for the last 10 years, but his financial,
medical and physical care has been shared out amo
ng the adult household members. As
Drinkwater notes, in areas of high prevalence, almost everyone is affected in one way or
another, but few households will experience adverse effects like declining food production,
loss of assets, a dec
line in income (Dri
nkwater et al., 2006

There is also a problem in definition and use of concepts. The concept of vulnerability is
commonly used in studies on the HIV prevention and impact assessment (Delor and Hurbet,
2000) and defined and interpreted in different ways ac
ross and within disciplines (
Casale et

al., 2009). Common to all these definitions is that it concerns exposure to potential harm or
oss (Ellis, 1998;
Weichselgartner, 2001). Rural based households are known to be vulnerable
to shocks and stresses caused

by economic crises, health shocks, unexpected changes in
household composition, climatic extremes, and many others. However this depicts a group
that is passive and on the receiving side of whatever nature has to offer, which is not the case
from the hous
ehold case studies presented above. Vulnerability is a potential state to a
potential risk and becomes manifest when it manifests itself as impact. It can only be
measured in relation to the event that has caused it. But society is complex, the outcomes th
are observed are as a result of interplay among various components that may not add up to the
‘vulnerability’ observed. Food and income shortage in household 17028 was not caused by
the death of the household members, but due to chains of causality an
d interactions of layers
of ‘vulnerability, including low investment in education of the children, loss of assets and
income sources after the death event, long dry seasons causing poor harvests, and many
others. This is one reason why vulnerability is di
fficult to measure. (Theoretical Notes on
Vulnerability to Disaster, 2009). In addition to that, the concept is commonly used in disaster
studies to assess risk, for example, to assess the vulnerability of certain groups of people as
result of climate. A
study in Thailand discovered that responses in this the case of climate
change are different compared to health and economic shocks (
Songporne et al., 2010

A further factor could be differences in location or context. It is widely known that assessing
he impact of HIV/AIDS is best done at micro

that is individual, household and
society level. And the findings are specific to the group under study and to the context in
which the study is carried out. Transferability or generalisation of findings is

not possible due

Jovita Amurwon_concept paper_110216

to differences in individuals, communities, countries. Even in small groups of people there are
differences as in the case studies above. Households 25009 and 17028 increased household
income by having the adult children go to look for no
farm work. In the same way, the
impacts of an epidemic are particular to specific individuals or groups of people in different
locations, just as the responses that work in one place to specific groups of people may not be
the same in others.

From the
household case studies and from the literature reviewed, the kind of question for
further investigation is not clear.

A number of theories and concepts
on human behaviour and protective processes
will be used
in this study
which will facilitate
understanding of rural household

Theories and

This research is based on an understanding that micro
level research on livelihood

at the household and individual level, supported with a larger cross sectional
panel set of data (Beeg
le and De Weerdt, 2008), can build a more detailed understanding of
cause effect relations of HIV/AIDs impacts at the household level.

In the first place the study will explore the complexity of household composition in the
context of Uganda and in the co
ntext of this community in south
western Uganda. As made
clear in the case studies households are located in a wider social network, providing access to
social and other resources in times of need, an observation supported by Seeley (2008) on a
study carri
ed out in the same population in south
western Uganda.

The theories of “agency” and “structure” can contribute to understanding on behaviour of
individuals and groups of people. "Agency" refers to the capacity of individuals to act
independently and to
make their own free choices. It implies that choices are not a product of
causal chains, but are significantly free or undetermined. "Structure", by contrast, refers to the
recurrent patterned arrangements (like social class, religion, gender, ethnicity, c
ulture) which
seem to influence or limit the choices and opportunities that individuals or groups of people
posses (Baker, 2005). From the case studies (table 1) some households (household 25009)
display agency in choice of strategy that will sustain house
hold resources. The structural
constraints in household 17028 limit its flexibility in choice of strategies to meet household
consumption and income needs. In this case one of them is gender and power relations. This

Jovita Amurwon_concept paper_110216

study will take an approach that tries

to understand the both the agent on their own and the
situated agent limited by policies, culture, power relations that determine capabilities.

Some policies that attempt to control the spread of the epidemic have assumed the agency of
individuals in ch
oice of behaviour. For example the ABC strategy in Uganda, as mentioned
earlier, is fighting the epidemic using an individualistic approach (Barnett and Parkhust,
2005). In this case the agent is blamed for the outcomes, which is contracting or spreading o
the virus, as opposed to circumstances that influence individual behaviour

like economic
pressures. The choices that individuals make to meet life events are influenced by processes
that are protective, or unprotective.

The concepts of resilience and ad
aptation are processes that involve response of an agent to
the environmental and socio
economic changes. Resilience is the capacity of a system to
absorb disturbance and reorganize while undergoing change so as to still retain essentially the
same funct
ion, structure, identity, and feedbacks (Walker et al., 2004). In the context of rural
livelihoods, and from the case studies in table 1 above, it indicates a capacity to resist decline
in functioning even when the individual or unit appears to be falling
in a trough. It is a
process, and not an attribute, which is constructed through adaptation, another concept that I
find useful for this study. This study will attempt to define and identify traits of resilience in
the context of the population under study
. Adaptation, refers to adjustments of a system
(individually or as a group) to deal with new experiences in a constantly changing
environment. It can be either reactive (stimulus

response) or proactive (anticipation

response), or both as can see in t
he case studies above. However the line between reactive or
proactive behaviour or changes is blurred. This study will establish the differences in
strategies that households employ, consciously and unconsciously, as they take advantage of
opportunities in

the environment or as they struggle to meet their needs.

This study will also investigate the historical experiences of the individuals, households and
community the

like price stability of farm products, market situation, climatic conditions,
and the political environment, so as to create understanding of the h
ousehold trajectories. This
information will be collected both at individual and household level and at community level.


Jovita Amurwon_concept paper_110216

Methods and the study area

The Medical Research Council (MRC)
and the General P

The Medical Research Council/Ugand
a Virus Research Institute Uganda Research Unit
(MRC/UVRI) on AIDS was established in 1988 based on a bilateral agreement between the
Ugandan and British Governments. It is a multi
disciplinary research programme for the study
HIV infection in Uganda
. T
he General Population Cohort (GPC) study was established in
1989 in 15 rural villages (expanded to 25 villages in 2000) in a sub
county in Masaka district
in south western Uganda.

This is an open cohort covering a total population, with no age
limit, of ab
out 20,000 people. The GPC’s main objectives are to describe the dynamics of
HIV infection within a rural population, to identify the major risk factors for contracting HIV
and to quantify the impact of mortality and fertility and to study treatment seekin
g behaviour.
Every year since its inception, the GPC has conducted household censuses of the resident
population collecting social and economic data among others. These include age, sex,
education, relationship of household members to the household head,
and an adult medical
This dataset will be used for my study. The first step will be to explore
the criteria
used for recruiting these

households into the GPC

Basing on these findings, a

of a hundred households will be
drawn and used t
among, others,
changes in
household composition
, movement of individuals in and out of the households and assets

These will be used to compare with the information from the case studies.
might have
. I
t uses 1
990 as the base year, e
vents prior to that time might be
nt in explaining current household experiences and social
economic status


property acquisition, ownership and loss; income sources and changes; household
composition; social networks. Th
e other problem is that not all variables are included
in the
annual census
ousehold e
conomic data is captured every four years, cropping has been
captured only at two points, in 1990 and 2008.
The household definition might be

it is based on

the house structure.
The evidence
e case studies presented
above shows that t
he household can relocate

when the structure collapses.

Other studies nested within the GPC have collected other data that can be used in this study.
In 1991/2, the MRC/UVRI conducted an ethnographic study of household coping
mechanisms with the aim of getting in/depth understanding on household survival dyn
and how the epidemic had influenced their life outcomes. This was based on the experience of
27 households selected from three GPC study villages (Seeley JA., 1993). Over a period of
one year interviewers paid monthly visits to the study households t
o record changes in

Jovita Amurwon_concept paper_110216

different aspects of the household’s daily life such as composition, employment, health, food
consumption and social networks. In 2006/7 a follow
up study was carried out (Seeley JA.,
2008) with the aim of assessing the impact of HIV
fection on households over a 15
period (1991
These data
provide details on the dynamics of household survival,
the drivers of change and household responses, which

will need for my study.
there are


with these

he baseline is limited

that of the GPC, which is
1990; the definition of infected versus uninfected households changes overtime

as households
respond in various ways to mitigate impacts of sickness and death

this is evident from the
case studies

events external to the household are n
ot collected

prices of labour, crops,


policies that have impacted on rural livelihoods,
historical events (war,
the ba
n of Asian businesses


Data external to the househol
d will be collected from the national database. These include; the
Uganda statistical database for information on economic performance, rainfall, seasonality,
crop performance, prices; the Uganda Bureau of Statistics established in 1998 has data on

economic and demographic statistics; the Uganda meteorological department has
rainfall and temperature data. Other national databases will be identified and consulted for
more information.

New data will be collected using in
depth interviews to include
old on

and off
life histories, for as far as individuals can remember

and details of specific events
and household responses to them

Additional data to capture events external to the household

but affect household outcomes

like pr
ices, seasonality, crop changes, and others that have
affected the households will be collected.

sample of 20 households

are female
, 10 are
uninfected (had no AIDS related death in
the last 20 years) are being

up to observe
ltural practices, income sources, social ne
tworks, and related livelihood.

In addition to this, 6 focus group discussions with household heads (3 groups with females)
were carried out in July and August


to get the community perspective and interpretat
of events related to livelihood activities, factors affecting household availability of food and
income, agricultural practices and
changes that have occurred

in the community


A sample of a hundred households will be drawn from the GP
C database to explore changes
in household composition and movement of individuals over time. This will be used to
triangulate with the findings from the household case studies.

Study plan

(separate attachment)


Jovita Amurwon_concept paper_110216


titles for the articles


households from the

in Masaka


Risk analysis of household susceptibility to HIV infection


Household livelihood trajectories in Masaka


Analysing HIV prevention and impact mitigation strategies for rural communities and

relevant policies for sustained HIV

prevention in Uganda, in the 21



Amartya S. (1985).
Being, Agency and Freedom
: The Dewey Lectures 1984. Journal of
Philosophy 82 (4):169

Bachmann, M.O.;
Booysen Fle R.

Health and economic impact of HIV/AIDS on South
African households: a cohort study
. BMC Pub. Health 2003, 3, 14.

Barker, C. (2005).
Cultural Studies: Theory and Practice
. London: Sage. ISBN 0
8 p448

Barnett, T. (2006)
A long
wave event. HIV/AIDS, politics, governance and ‘security’:

the intergenerational bond?

International Affairs







and Blaikie


AIDS in Africa, its Present and Future Impact
. Guilford
Press, London.

Barnett, T

and Whiteside, A

(2007) Poverty and HIV: impact, coping and mitigation policy.
In: Cornia, Giovanni Andrea, (ed.)
AIDS, public policy and child well
Innocenti Research Centre, Florence, Italy, pp. 209

Beegle K, De Weerdt J.

Methodological issues
in the study of the socioeconomic
consequences of HIV/AIDS
. AIDS;22 (Suppl 1):S89


Beegle, K.

Labor effects of adult mortality in Tanzanian households
. Econ. Dev.
Cultural Change, 53, 655

Booysen Fle R.
, Arntz T. (2003)
The methodology of HI
V/AIDS impact studies, A review of
current practices
. Social Science & Medicine. Jun;56


hapoto, A. and Jayne,

T.S. (2008)

Impact of AIDS
Related Deaths on Farm Households’
Welfare in Zambia
. Economic Development and Cultural Change, 56 (2):


Clover, J


Food Security and Humanitarian Emergencies in Southern Africa
. African
Security Review 12 (1):10.

Cresswell, J

Qualitative, Quantitative, and Mixed research Approaches
. Thousand
Oaks, Calif.: Sage Publications

Corta, Lucia (2009) “
The Political Economy of Agrarian Change:

Dinosaur or Phoenix?

in B. Harriss

White and J. Heyer (eds.) The Comparative Political Economy of

Africa and Asia, New York: Palgrave Macmillan

De Waal, A

and Whiteside


New Variant Famine: AIDS and Food Crisis in
Southern Africa
. The Lancet 362:1234

Delor, F., and Hubert, M.

Revisiting the Concept of ‘Vulnerability

Social Science &
Medicine. Volume 50, Issue 11, June 2000, Pages 1557


Jovita Amurwon_concept paper_110216

Drinkwater, M., McEwan, M., & Samuels, F. (2006).
The effects of HIV/AIDS on agricultural
production systems in Zambia

A restudy 1993

2005. Lusaka, Zambia:

Ellis, F. (1998).
Household strategies and rural livelihood diversification
Journal of
Development Studies, 35 (1): 1






Livelihoods and rural poverty reduction in Uganda
. World
Development 31, 997



HIV/AIDS, gender Inequality and rural Livelihoods. The Impact of HIV/AIDS on
Livelihoods in Northern Zambia
. Sustainable Development Department, FAO:

Fauci, A. S. (2006).
five years of HIV/AIDS
. Science, 313(5786), 409.



Dominguez K
, Kalish



, Donohoe

, Brooks J


Practice of feeding premasticated food to infants: a potential risk
factor for HIV transmission
. Pediatrics
;124 (2):659


L., Hoddinott


and H. Alderman

Intrahousehold resource allocation in
developing countries
, Johns Hopkins University Press, Baltimore, MD


, de Zalduondo


Combination prevention: a deeper understanding
of effective HIV prevention

AIDS, 24: suppl




Henderson, K., Worth, H., Aggleton, P. & Kippax, S. (2009)
ancing HIV prevention
requires addressing the complex relationship between prevention and treatment
. Global
Public Health, 4, 2, 117

Hosegooda, V.,

Whytec, E., Buszaa, J., Moitsed, S.,

Timaeusa, I.M.

the full extent of households’ experiences of HIV and AIDS in rural South Africa
Sci. Med., 65, 1249

Marisa Casale, Scott Drimie, Timothy Quinlan and Gina Ziervogel . 2009.
vulnerability in southern Africa: comparative f
indings using a multiple
approach in South Africa and Malawi.
Regional Environmental Change 10(2): 157


, Crepaz

, Senterfitt


analysis of high
risk sexual behavior in
persons aware and unaware they are infected with
HIV in the United States: implications
for HIV prevention programs
. J
ournal of


Immune Defic



Marschke, M., Berkes, F. (2006)

Exploring strategies that build livelihood resilience: a case
from Cambodia
. Ecology and
Society 11 (1), 42.

Murphy, L

related morbidity, mortality and the environment

Note prepared
for Population Environment Research Network cyber

seminar on HIV/A
IDS and
Environment, April 2006.

Narayan, D.,

Voices of the poor: Povert
y and social capital in Tanzania.

Environmentally and Socially Sustainable Development Studies and Monograph Series
20, World Bank, Washington DC.

Zziwa, A. (1995).
Gender perspectives on land ownership and inheritance in Uganda

Seeley J., Biraro S
., Shafer L
A., Nasirumbi P., Foster S., Whitworth J. and Grosskurth H.
Using in
depth qualitative data to enhance our understanding of quantitative
results regarding the impact of HIV and AIDS on households in rural Uganda
. Social
Science and Medic
ine 67: 1434


Jovita Amurwon_concept paper_110216



Searching for Indicators of Vulnerability, a Study of Household Coping
Strategies in Rural South West Uganda
. MRC/ODA, London.

Seeley, J., Ekoru, K., Bukenya, D., Lutalo, T. and Pain, A. (2010) The longitudinal impact o
HIV and AIDS on rural livelihoods in East Africa. Research Report. Food and
Agriculture Organisation (FAO


, Mugerwa

, Sewankambo

, Lwegaba

, Carswell JW, et al. (1985)
Slim disease: A new disease in Uganda

and its association with HTLV
III infection
Lancet 2: 849


Smith AM, Mahe
r J, Simmons J, Dolan M. (2004)

HIV prevention from the perspective of a
based development agency
. London, CAFOD

Ssengonzi, R.
The impact of HIV/AIDS on the living a
rrangements and well
being of
elderly caregivers in rural Uganda
. AIDS Care 2009, 21, 309

Songporne Tongruksawattana, & Waibel, Hermann & Schmidt, Erich, 2010.

"How Do Rural
Households Cope With Shocks? Evidence from Northeast Thailand

Proceedings of

German Development Economics Conference, Hannover 2010 53, Research Committee
Development Economics.

Yanagisako S.J. (1979). Family and household: the analysis of domestic groups. Annual
Review of Anthropology 8: 161

Other Sources

AVERT (2006)

“History of AIDS 1981


Accessed 05th January 2011

AVERT (2011) “
The history of AIDS in Uganda
” Source:
uganda.htm. Accessed 05th January 2011

New Scientist (1990) ‘AIDS epidemic moves south through Africa’.
africa.html Accessed on 2

December 201

os Global AIDS Programme.

Communication Challenges in HIV Prevention: Multiple
Concurrent Partnerships and Medical Male
Circumcision ‘Source

accessed on 9

February 2011

Patel, D.
and Scott, G., 2003, ‘
Just another day in Paradise
: Morris and Lewis and cutting the

unpublished manuscript

STD/AIDS Control Programme (2002) ‘Trends in HIV prevalence and sexual behaviour
2000) in Uganda’

The Independent (2006) “Public health

and religion: Aids, America, abstinence...” Source
480593.html Accessed on
05th January 2011

Theoretical Notes on Vulnerability to Disaster, 2009. Disaster
planning and emergency
management. http://emergency

planning. theoretical
to.html. Accessed on 12th December. 2010.

UNAIDS (1995). Report from a consultation on the socio
economic impact of HIV/AIDS on
ds. Available:

UNAIDS (2010) 'UNAIDS report on the global AIDS epidemic'


Jovita Amurwon_concept paper_110216

Weichselgartner, J. 2001. Disaster mitigation: the concept of vulnerability revisited. Disaster
Prevention and Management 10(2): 85

Towards universal access: scaling up priority HIV/AIDS
interventions in the health sector
. P
rogress report 2010.
WHO Library Cataloguing
Publication Data.