PC3 Baseline Edited
PC3 Baseline Edited
PC3 Baseline Edited
Funded by USAID
2005
The urban HIV prevalence rate is 13.7 percent compared to rural rate of 3.7 percent. In
Addis Ababa, the nation’s capital, prevalence is estimated at 15.6 percent. The Ministry
of Health (MOH) states that two-thirds of deaths among 20-54 year olds are AIDS
related, with women becoming infected more often than men. Projections suggest that
as many as 5.25 to 6 million Ethiopians will have died from the pandemic by 2014.
Current sources indicate that Ethiopia’s total estimated orphan population is four million,
or roughly 13 percent of the country’s almost 31 million children. An estimated 1.2
million children (or 4 percent of all children or 30% of all orphans) are children orphaned
by AIDS.2. According to the Ministry of Health, “The issue of AIDS orphans is the worst
crises as compared to all crises caused by deaths of adults because of AIDS. In view of
the prevailing AIDS related general and specific death rates, the number of orphans in
Ethiopia would increase from 1.2 million to 1.8 million by 2007 and to 2.5 million in
2014” (AIDS In Ethiopia, Ministry of Health, 2002).
The multifaceted impact of this disease across generations, households and communities
is profound, resulting in complex causes and consequences of vulnerability3.
1
USAID, Bureau for Global Health; Country Profile- Ethiopia, Dec. 2003;
http://www.usaid.gov/our_work/global_health/aids/Countries/africa/ethiopia.html
2
Children on the Brink. A Joint Report on Orphan Estimates and Program Strategies. USAID, UNAIDS
and UNICEF, 2002.
3
Most information on OVC is contained in program document (Positive Change: Children,
Communities and Care (PC3) Program, Technical Approach), although the efficacy of interventions
targeted at OVC has yet to be examined in the future.
Owing to this situation a five-year program plan, starting 2005, entitled “Positive
Change: Children, Communities and Care (PC3)” has been initiated by Save the
Children /USA, CARE, Family Health International, World Vision and World Learning
with US$20 million budget promised by USAID as part of PEPFAR (Presidential
Emergency Plan for AIDS Relief.)
The overall objective (Goal) of the program is to improve the well being of orphans and
other vulnerable children and families affected by HIV/AIDS, specifically with respect to
education, health and nutrition, livelihoods, psychosocial and legal support.
The program is implemented in partnership with major actors including community-
based institutions, civil society organizations (CSOs), the government and NGOs. It is
assumed that this approach will ensure the most effective community based response and
further the reach (about 500,000 OVC) of the program PC3 builds on the existing
traditional social networks and institutions, and capacity building of communities to
increase availability, quality and consistency of community-based support services as
well as to create a more supportive environment for OVC and families affected by
HIV/AIDS are the key results of the program.
The base line survey assesses the situation of OVC and their families specifically with
respect to education, health and nutrition, livelihoods, psychosocial and legal support.
The results of the baseline survey will be used to establish the status of OVC as compared
to non-OVC in terms of key outcome and impact indicators, and as source of information
for programming as implementation progress.
particular intervention. Problems with stigma, as well as resentment and conflicts within
households and communities, occur when very specific definitions established for
quantitative purposes are used for program targeting or eligibility criteria in policy and
program implementation. The quantitative process must have clear boundaries and allow
for absolute distinctions. In contrast, developing and implementing programs and
services must take into account local variations in the factors that cause or constitute
vulnerability. For programming and service delivery, no one specific definition will
suffice for every context.
While defining ‘orphan’ for national study purposes is fairly easy, establishing a
measurable definition of ‘vulnerable’ is a bigger challenge. It is recommended that the
concept of ‘vulnerability’ be captured with a minimum number of variables to strictly
specify the population measured and ensure replicability.
Following a review of recent quantitative studies that included different concepts of
vulnerability and a series of consultations on the findings, the UNAIDS Monitoring and
Evaluation Reference Group recommends, for household surveys, the following
definition of orphans and other children made vulnerable by HIV/AIDS.
An orphan is a child below the age of 18 who has lost one or both parents.
A child made vulnerable by HIV/AIDS is below the age of 18 and:
i) has lost one or both parents, or
ii) has a chronically ill parent (regardless of whether the parent lives in the same
household as the child), or
iii) lives in a household where in the past 12 months at least one adult died and
was sick for 3 of the 12 months before he/she died, or
iv) lives in a household where at least one adult was seriously ill for at least 3
months in the past 12 months
Although the focus of the baseline survey is to measure the characteristics of children
affected by HIV/AIDS, it is not feasible to know the HIV status of the adults or children
in most circumstances due to the fact that most hide the status due to stigma and others
do not know their HIV status as they have not been tested. So proxy measures have to be
used to identify and estimate children who are affected by high adult mortality, morbidity
and inability to meet the needs of a child. Therefore, after a series of consultations with
the client we defined OVC for the purposes of this study as follows:
OVC is a child below the age of 18 who:
1. Has lost one or both parents, or
2. Has a chronically ill parent or lives in a household where at least one adult was
chronically ill (at least 3 months in the past 12 months), or
3. Lives in a household where in the past 12 months at least one adult (18-60) died, or
4. Lives in a household headed by child (under 18 person), or
5. Lives in a household headed by a single (female), or
6. Lives in a household headed by an elderly (grand parent)
In a setting with high HIV prevalence, this definition is likely to be a fairly reliable proxy
for children and households affected by HIV/AIDS, plus other causes.
Previous studies depicted that children affected by HIV/AIDS often face immense
obstacles as they attempt to thrive in what is already a difficult environment.
The magnitude of the scale of impact HIV/AIDS has had on children has been the focus
of several studies. According to Bicego et al (2003), maternal orphan prevalence ranges
from <2.5 percent to 4.5 percent in sub-Saharan Africa, paternal orphan prevalence
ranges from 4 percent to 8 percent; and double orphan prevalence ranges from <.05
percent to 1.5 percent. Similar results have been reported by Ainsworth and Filmer
(2002) who found that in West Africa, 4 to 10 percent of school-aged children are
paternal orphans which is roughly twice the proportion who are maternal orphans, with
double orphans representing 1.6 percent or less in this setting. In East and South Africa,
numbers of paternal orphans are higher (6 to 13 percent), while the proportions of
maternal orphans are similar to West Africa. Both studies found a positive correlation
between orphan rates and HIV prevalence but point out that the impact of HIV on orphan
status depends on the maturity of the epidemic in a certain country, therefore the scope of
the problem is likely to change over time.
Some research studies have focused on documenting the disadvantages faced by OVC.
The studies have compared health, economic and social status between OVC and non-
OVC. Poverty, health and education are generally the key characteristics of interest in
these literatures. Research in this area is still developing and consistent patterns of results
have not emerged from the evidence available to date.
The paragraphs below summarize what is known, from other studies, about the
circumstances of OVC with respect to socioeconomic status, school enrollment, health
and psychosocial status.
Studies on socio-economic status of households with OVC versus those without OVC
have found mixed results. Several smaller studies have found households with OVC to be
worse off than non-OVC households, whereas other studies have not found this to be the
case. For example, Ainsworth and Filmer (2002) found no statistically significant
difference in the prevalence of orphanhood between poor and non-poor households. In
some countries the poorest households were found to be less likely to have orphans. This
could be because households with the most resources were more likely to take in
orphaned children or this finding could reflect the socioeconomic distribution of HIV
infection. This is contrary to the findings of a survey jointly conducted by
MOLSA/UNICEF/Italian Cooperation.
proportion of ‘non-AIDS orphans’ forced to beg is one out of 50. The probability of
‘AIDS orphans’ engaging in economic activity considerably increases after the death of
the mother as compared to the death of a father. But this study didn’t take into account
other children made vulnerable due to HIV/AIDS without being orphaned. Moreover, it
did not indicate socio-economic status of households with OVC versus those without
OVC.
On the other hand, Deininger et al. (2001) examined the socio-economic impact of
orphan hood at the micro and macro economic level. Their findings indicate that at the
micro-economic level, receiving an orphan into the household has a large negative impact
not only on the household’s consumption but also on their capital accumulation in the
long term. Since fostering reduces savings and investment at the household level, this can
have an indirect impact on aggregate savings at the macro level.
Several studies have found that orphans are less likely to be currently attending school
than non-orphans. Other studies have found that this varies by country and by sub-regions
within countries. Case et al. (2003) used DHS data from 10 countries in sub-Saharan
Africa to examine the impact of orphan hood on school enrollment and found that
orphans are significantly less likely than non-orphans to be enrolled in school. Non-
enrollment of orphans is not explained by poverty but by relationship (i.e. closeness) of
the orphan to the household head. The probability of attending school is higher when
orphan is living with close relatives as compared to those living with distant relatives or
non-relatives
Bicego et al. (2003) found that an orphan is less likely to be at his/her proper educational
level than a child who has both parents living. The effect is stronger at younger ages
(ages 6-10) than older ages (11-14). Furthermore, double orphans (defined as children
who have lost both parents) are less likely to be at their proper level than single orphans
(defined as children who have lost one parent). For single orphans, a mother’s death
causes more deterioration in education at the primary school ages, as compared to a
father’s death. Similarly, the Survey of the Prevalence and Characteristics of AIDS
Orphans in Ethiopia, (2003) indicates that the death of the mother significantly affects
the educational development of AIDS orphans. Following the death of the mother, 1
out of every 5 child drops out of school, most often due to poverty.
Most studies on physical well-being of OVC (Lindblade et al, 2003; Hess, 2002) have
focused on children under five years of age. Results from these studies, like those
reviewed above, have been contradictory. In one (Lindblade et al., 2003) the authors
found little difference in physical well being, while in another (Hess 2002) there were
key differences. Deininger et al., (2001) also found the health and nutritional status of
orphans to be worse than non-orphans.
The Survey of the Prevalence and Characteristics of AIDS Orphans in Ethiopia, (2003)
jointly conducted by MOLSA/UNICEF/Italian Cooperation depicted that AIDS orphans
younger than 10 years old have very poor health condition and poor access to healthcare
services. The report also indicated that children orphaned by AIDS are underfed – 1 out
of 2 (~49%) do not receive adequate food , while this figure stands at ~40%for children
made orphan by non-AIDS causes.
Psychosocial
The above summary of the literature review regarding OVC in Ethiopia and sub-Saharan
Africa suggests a number of gaps in our knowledge, some of which the Baseline Survey
described herein seeks to fill.
The main purpose of the baseline survey is to establish benchmark information on the
situation of OVC and households affected by HIV/AIDS in the target areas. Unlike most
of the surveys conducted in this country and elsewhere in Africa, this survey answers the
question ‘to what extent are OVC more disadvantaged than non- OVC?’ by measuring,
comparing and contrasting the situation of OVC with that of non-OVC and situation of
households with OVC to that of those without OVC in terms of variables that adequately
indicate access to services and impact of the program.
Thus this baseline survey addressed a wide array of outcomes within six domains: 1)
prevalence and Characteristics of OVC; 2) the socio-economic characteristics of families
with OVC and the capacity of families to protect and care for children (in terms of
providing basic material needs, adequate food, succession planning), 3) access to
essential services (health, education, recreational, etc.) 4) psychological and emotional
health and existence of awareness and supportive environment for OVC and families; 5)
Extent of stigma and discrimination towards, and abuse and exploitation of the OVC and
families and existing legal protection/ support for OVC, 6) Community-based responses
for OVC and the capacity of CBOs to protect and care for OVC and families.
To produce a comprehensive data base and wide array of information on OVC and non-
OVC that can be used by interested stakeholders for further intervention and also help
serve as a benchmark for future monitoring and evaluation; Quantitative Methods and
Qualitative Inquiry were integrated and used to conduct the survey. Before leaving for
the study areas as well as after field work, the team reviewed all relevant and existing
documents produced by various government institutions (HAPCO, MoH, MoLSA, CSA),
and many by relevant NGOs (FHI, CARE, etc.), bilateral (USAID) and multilateral
organizations (UNICEF, UNAIDS, World Bank, etc.) related to the situation of OVC and
households affected by HIV/AIDS in Ethiopia.
This paragraph is unclear. The principal focus of the study group is orphan and
vulnerable children (OVC) living within the household settings of seven different
towns/cities of Ethiopia. These towns include Addis Ababa, Nazareth, Bahir Dar, Dire
Dawa, Shashemene, Awassa, and Asaita.
The study variables are addressed to collect information for households with OVC and
without OVC as well as OVC and Non- OVC youth of age 12-17. “We formed a subset
of households with youths aged 12-17 since a sample of youth groups administered their
own questionnaire. Although data derived from younger children (<12) would also be
very useful, due to concern for sensitivity and the lack of maturity of younger children,
we have opted not to attempt to collect this data.” Certain key issues are addressed to
compare results of the survey for different characteristics shared by the two household
groups.
The total number of households for each of the target Kebeles in the study sites was used
to adopt two-stage sampling design, a selection of Kebeles as a primary sampling unit
(PSU) with probability proportional to population size (PPS) and then households.
Having selected the Kebeles based on their proportional sizes, a fixed sample of
households is randomly selected with systematic sampling procedure from each selected
Kebeles of each town. The procedure yields an equal probability of selecting household
from the list within each town.
In the absence of readily available sampling frame for sample selection of households
and the huge cost of sampling frame construction, 33 out of 196 identified Kebeles
project sites are selected using PPS. In each town, the number of selected Kebeles ranges
from 2-10 depending on the number of Kebeles included in the project.
Current population data and size of household at the Kebele level were difficult to obtain
for the study towns. To overcome the problem it was decided to use the 1994 Population
and Housing Census result and extrapolate to 2004. For the extrapolation, urban
population growth rate of 5% (Mulugeta, Solomon and McLead, Ruth, 2004) is used in
the population growth formula;
Pt = Po(1+r)t
After data is generated, household sizes of each Kebele were cumulated within each
town/city to make selection of Kebeles possible with probability proportional to size. A
Random Number Table is used to identify and then select the required number of Kebeles
(PSUs) in the city/town if the random number falls within the range of the number of
households residing in the Kebele.
In the absence of a sampling frame, an alternative approach for selecting households and
children of 12 to 17 years of age was adopted. To select the households, a survey team of
enumerators and a supervisor were initially deployed to a selected Kebele to make
reconnaissance survey of “physical” sample selection and identified locations of
households using systematic random sample selection procedure.
Before selection and registering the household as a member of the sample, the
team enquired whether the household has a person under 18 living with it.
Although a particular systematic nth selection may fall on a specific household, it
does not qualify to be included in the sample unless a household has a person
under 18 living with it. Rejection of any non-qualifying households is replaced
with the next nearest qualifying household. i.e., If the selected household does not
meet the selection criteria of having a person with less than 18 years of age then
we try the next (19th) neighboring household. If, for example, the 15 th and the 16th
households do not have a child and the 17 th does, we make a selection of the 17 th
household. Then we retain our initially set interval and go to the 25 th in the line.
It is understood that the procedure introduces some selection bias, but it is one
simple option in the absence of the sampling frame.
The name of the head of the selected households was recorded with the Kebele
and house identification number so that the next visit for the survey is easily
identified.
Selecting the youth respondents from the households selected for administering
the household questionnaire helped to link the household characteristic and
psychosocial sample surveys.
It should also be noted that the youth were selected randomly in two forms.
o Where there was only one child (boy or girl) of that age range in the
household, we automatically selected that boy or girl for the sample
o Where there were more than one child (any gender) of that age range in the
household one child, male or female, was randomly selected using random
number or by Coin flipping. Randomness ensured gender and OVC and non-
OVC balance.
Sample size estimates are initially determined using standard estimation procedures under
the assumption of simple random sampling procedures. With the adoption of multi-stage
sampling of clusters as PSU and then households, the design effect was taken into
consideration.
Sample estimate of the proportion of OVC 0-17 living in households to the total
population is computed to give 0.06 as illustrated by Turner (Turner, Anthony G.; May
2003).
Thus, no= 1.962 *(0.06)(0.94)/ δ2 where δ=0.15*p = 0.009
= 2654
As a result of multi-stage sampling (selection of clusters and then households) a design
effect of about 1.3 is also applied to yield n=2654*1.3=3450. With the assumption of
about 15% non-response as a characteristic of urban surveys, the sample size is further
calculated to give about n=3450/0.85=4059 households for all the study sites.
Based on the total sample size obtained from the above computation, the number of
households to be selected from each Kebele was set to a fixed sample size of about 125.
Selection of 10 kebeles from Addis, the number of sample allocation for Addis Ababa is
set to 1250 while Dire Dawa and Nazareth are both assigned 750 households from 6
Kebeles and 625 households from 5 kebeles, respectively. Awassa and Shashemene are
both allocated 500 households from 4 kebeles each, while Bahir Dar and Assaita are
assigned 250 households from two Kebeles picked as a primary sampling unit from each.
Accordingly, the sample size for the survey of seven cities/towns sums was 4,125
households from 33 Kebeles selected as primary sampling units based on their
proportional sizes.
In addition to the household survey to be conducted using the above sample size, the
procedure was believed to guarantee a reasonable number of OVC below the age of 18
for psychosocial survey of the young age group. Having a total selection of 4,125
households for the study, it is expected to have about 40 OVC households from each
Kebele which yields a total of about 1400 OVC households to be included in the total
sample households.
Once the number of sample households is determined it was possible to compute for the
number of OVC age 0-17, which in turn can be used to roughly guide the proportion of
youth within the age group of 12 to 17 using the following formula (Turner, A.; 2003).
At an average family size of 5.6 for an urban population, a sample of 4,125 households
which were known to have at least one child under the age of 18 gave a minimum total
population of about 23,100 people of various ages. Assuming that both parents were
living and together, this represented a minimum of 14,850 children. Based on the census
data for age group 10-14 and 15-19, it was assumed that the young age group of 12-17
comprises about 9.0% of the total population and thus captures about 2,080 adolescents
in the total sample households. It is possible to assume that a total of this group comes
from a maximum of 2,080 households in the case of at least one child (age 12-17) per
household. Therefore, it looked simply logical to expect a sample of about 1300 from
this group.
Sample selection for the psychosocial survey of young age group also followed random
selection procedure, i.e it is limited to only one young individual, male or female per
household. Statistically speaking random selection procedure will balance type of child to
be included into sample (i.e. gender, orphan hood, etc).
In the absence of sampling frame for this group, determination of the exact total sample
size and its allocation to each town/city was not finalized until “physical” selections of
sample households are carried out. In this particular case where both male and female
youngsters could also be selected from a household, the procedures enables to increase
the size to match the number of sample size calculated using the above formula. Hence, it
was decided to fix the sample size to about 1425 with a minimum of 120 from a
town/city and to a maximum of one-third of the sample household per town/city.
Household Heads and Youth Sample Size allocation by the Study Towns
After incorporating the feedback and comments from the clients the questionnaires were
translated into the local language, Amharic, to ease administration process and pre-tested
to see its suitability in terms of ease of language, flow, length, match between responses
from the respondents and possible answers on the questionnaire, relevance of the
questions etc. Adjustments were made based on the pretest results to ease administration
process. Experienced and trained Enumerators administered the questionnaires to sample
households and youth in the selected towns. (See Annex 2 and 3 respectively for the
Indicators and Questionnaires, respectively)
Different Qualitative Inquiry tools: Focus Group Discussion (FGD) and Key Informants’
Interview (KII) complemented with Matrix Scoring, and Pair wise Ranking were applied.
Beside the core team, additional professionals (four public health experts, a sociologist,
and a Psychologist) were recruited to facilitate the qualitative inquiry. There were 3
teams of qualitative inquirers (2-3 professionals per team) to speed up the data collection
process.
During the actual data collection, field guides were hired to accompany the enumerators.
In some of the survey areas, particularly Dire Dawa and Assaita translators were hired to
ease the communication difficulties encountered by the enumerators.
To enable the field supervisors and enumerators to have better understanding of the
purpose, and content of the questionnaire and skills to administer and supervise
questionnaire, an intensive 7 days training session was organized for enumerators and
supervisors, from September 1- 7, 2005, at Nazareth Refit Valley College. The training
had five modules that cover all aspects of the survey.
Before dispatching the enumerators for interviews, PC3 partners were informed of the
coming of data collectors to their respective towns or sub-city (in the case of Addis
Ababa) and made arrangements were made with Town officials and Kebele leaders to do
the sampling and conduct the interviews.
The actual field fork started on September 15, 2005 in Addis Ababa and Bahir-Dar; on
September 18, 2005, in Nazareth, Awassa and Shashemene; on September 21 in Assaita
and Dire Dawa. This cascading method is intentionally arranged to adequately monitor
and provide technical guidance for the supervisors so as to ensure the proper listing of
sample households and quality of data right from the beginning.
The fieldwork has taken more time than planned due to suspicions of the Kebele
Authorities on the purpose of the survey. It was especially challenging in Addis Ababa
and Dire Dawa, where things were rather expected to be done smoothly.
In spite of some challenges during delineation and sampling, questionnaire administration
went well. Each survey team (each Supervisor and his team of enumerators) were
covering one kebele at a time and moved on to the next to ease supervision. On average
3-4 Questionnaires were administered by enumerators per day. The supervisors
monitored and supported daily activities of the enumerators. To ensure reliability of the
information supervisors rigorously checked all filled questionnaires on a daily basis and
re-interviewed 5% of the respondents.
To ensure reliable data collection, a system that encourages the regional coordinators and
supervisors to monitor the work of enumerators and themselves rigorously was
developed. Each supervisor completed a daily log book, prepared for this purpose, which
detailed the number and name of households re-interviewed, number and name of
enumerators supervised and number of questionnaires checked every day. The regional
coordinators, whose responsibility it is to oversee the survey progress and backstop
supervisors in each town, also filled in the logbook. Moreover, core team members have
visited and supervised the data collection process in all towns continuously. (The
supervisors and coordinators' logbook is attached as Annex 5)
The data entry format was developed as per the codes used for questionnaire design in
Census and Survey Programming (CS-Pro 2.6) - Statistical Software, while field data
collection is in progress. The software provides exact paper-format of the questionnaire
on the screen, and data checking systems such as double entry, skip and range checks
were built in to facilitate data entry and minimize errors at entry stage w.
Data entry clerks having experience with CS-Pro were employed to reduce the training
time as well as to speed up the data entry process. Data entry clerks were oriented on the
format and content of the questionnaire and also on the entry format to make the process
of data transfer as accurate and efficient as possible.
Rigorous data validity and consistency checks were part of the data cleaning exercises
before starting the analytical work. The CS-Pro data management system was used to
minimize erroneous “out of range” data entry by the data entry personnel. A system of
random spot check is employed to determine whether each member of the team is
continuing to operate to the required level of standard.
For more in-depth cleaning, the SPSS data “EXPLORE procedure” was employed for
data screening, outlier identification, description, assumption checking, and
characterizing differences among sub-populations (groups of cases). Standard statistical
descriptive procedures and plots were also used to validate the data sets. As a result of
rigorous cleaning and data verification of all 4002 households interviewed, data from 6
households were rejected and only data from 3996 households were considered as valid
data sources; these were used for analysis and interpretation of the results
Once the cleaning was completed, data analysis was done using Statistical Package for
Social Science (SPSS). As a preliminary data analysis, the use of frequency tables,
percentages and cross tabulation relevant to nominal as well as ordinal type of
measurements were employed. Interval or ratio measurements were analyzed using
several descriptive statistics such as means, standard deviation, minimum, maximum,
range, etc. A combination of nominal/ordinal with interval scales was also employed
where appropriate.
Besides describing the situation, the analysis also investigated comparative characteristic
differences between respondents' survey variables. The analysis focused on the OVC vs
Non-OVC as well as sex and age group differences of relevant study variables.
Overall, 53% (2120) of the households were identified as households that accommodate
children identified as OVC; while about 32% (1280) households were found to be caring
for orphans. The average number of orphaned children living in these households was
1.7 with the range being from 1 to 8.
Relative distribution of OVC, among the survey towns, is nearly the same except for
Dire Dewa where it tended to be lower (36.5% of children are OVC), while higher
proportion of OVC were found in Awassa (51% of children are OVC). Distribution of
orphans follows the same pattern of OVC. Prevalence of orphans ranges from 18%
in Assaita to 33% in Bahir Dar.
Table 3.1 shows the distribution of OVC by the specific criteria of vulnerability. The
largest proportion of OVC are those living in households headed by single (divorced
&widowed) females (20.8%), followed by children living in HH with critically ill adult
(i.e. a non-parent adult?) (17.6 %), Proportion of children living with critically ill parent
and Children living in HH Headed by Elderly/ Grand parent are nearly the same (12%).
Only 5% of children are found to be living in HH where at least one adult died -12
months before the survey. Among the OVC group, children living in child-headed
HHs are the least prevalent (0.4%) (Table 3.1). This implies that most of the children
who become double orphans are either moved to households headed by adults or that
adult caregivers into the houses of children who lost both parents. This is supported
by the fact that about 60% of guardians, who have taken over the responsibility of
household head, had become head of the household during the last 5 years. And
about 30% of OVC have joined another households due to death of parents, unmet
basic needs and inability to get education (see Table-3.6: Child Mobility and Family
Cohesion)
4
Children on the Brink. A Joint Report on Orphan Estimates and Program Strategies. USAID, UNAIDS
and UNICEF, 2002.
Addis Ababa Nazareth Shashemene Awassa Assaita Dire Dawa Bahir Dar Overall
n % n % N % n % n % n % n % n %
Prevalence of Orphan and Other Vulnerable Children Among all Children (ages under 18) recorded by the survey
Valid N= 2691 1500 1467 1350 560 1839 528 9935
Orphans and Vulnerable children 1234 45.9 720 48.0 664 45.3 688 51.0 224 40.0 671 36.5 239 45.3 4440 44.7
(OVC) overall
Orphaned Children 579 22.1 355 24.2 326 22.6 298 22.6 100 18.0 338 18.6 175 33.3 2171 22.3
Children who don't know where- 84 3.1 42 2.8 36 2.5 52 3.9 3 0.5 26 1.4 2 0.4 245 2.5
abouts of their Mother and/or Fa-
ther
Children with critically ill parent 224 11.4 159 15.2 172 15.6 178 17.6 29 6.4 112 7.7 34 9.3 908 12.3
Children living in HH with criti- 335 15.1 235 18.3 261 20.5 318 28.3 102 19.5 187 11.5 61 12.6 1499 17.6
cally ill adult
Children living in HH where at 92 3.4 77 5.1 99 6.7 72 5.3 42 7.5 94 5.1 28 5.3 504 5.1
least one adult died -12 months be-
fore the survey
Children living in HH Headed by 23 0.9 5 0.3 4 0.3 3 0.5 3 0.2 38 0.4
Child
Children living in HH Headed by 600 22.3 389 26.0 299 20.4 204 15.1 86 15.4 322 17.5 161 30.5 2061 20.8
Single Female (divorced &wid-
owed)
Children living in HH Headed by 394 14.6 180 12.0 146 10.0 178 13.2 17 3.0 197 10.7 36 6.8 1148 11.6
Elderly/ Grand parent
Prevalence of OVC Among Sample households
Households with OVC 1213 608 491 481 245 722 236 3996
649 53.4 345 56.7 279 56.8 276 57.4 121 49.4 327 45.3 123 52.1 2120 53.0
Households with orphan 1196 593 487 476 244 715 234 3945
375 31.4 198 33.4 182 37.4 173 36.3 65 26.6 189 26.4 98 41.9 1280 32.4
Average No. of orphans in the HH 1.5 1.8 1.8 1.7 1.5 1.8 1.8 1.7
with orphans
3.2 Prevalence and Type of Orphan hood
As indicated in Table 3.1 22% (2171) of children in surveyed households were
identified as orphans. Further analysis of type of orphans (table 3.2) depicts that
the proportion of paternal orphans is more than twice of that of maternal orphans
throughout the study towns. The proportion of paternal orphans ranges from
14.4% in Assaita to 29.3% in Bahir-Dar as compared to maternal orphans, which
ranges from 6.8% in Assaita to 12.7% in Bahir Dar. The average proportion of
children who were double orphans was 4.4% (436) across the study towns. On
average, 11% (229) of orphans (or about 3% of all children) had become orphans
within 12 months prior to the survey.
Findings of the survey reveals that the relative proportion of maternal, paternal
and double orphans are consistent with those of Bicego et al (2003) and Ainsworth
and Filmer (2002), though the overall prevalence of orphans is higher in our study,
reflecting the high prevalence of HIV and other causes of adult mortality in
Ethiopia. For example to Bicego et al (2003) found that in sub-Saharan Africa
generally, maternal orphan prevalence (<2.5 percent to 4.5 percent) was roughly
half that of paternal orphan prevalence (4 percent to 8 percent); and that double
orphan prevalence was much lower (from <.05 percent to 1.5 percent). Ainsworth
and Filmer (2002) found that in West Africa, 4 to 10 percent of school-aged children
are paternal orphans, roughly twice the proportion who are maternal orphans,
with double orphans representing 1.6 percent or less in this setting.
Table 3.2 also indicates average length of orphanhood. According to the survey
result, the average length of orphanhood (number of years a child has been
orphaned), was about 6 and half years with the range varies from lessthan 1 year
to 17 years. This was fairly consistent across study towns. This indicates that the
orphan crisis is a serious and deep-rooted problem, and not a new phenomenon.
Positive Change: Children, Communities and Care (PC3) - Baseline Survey Report
Fig-3.1 Prevalence Orphanhood by Type Among all Children in Sample Households: Overall
Result of PC- 3 Baseline Survey- 2005, Ethiopia
Table 3.2 Prevalence of Orphan by Type Among Children (Under 18) and length of orphanhood Across
the Seven Selected Towns of Ethiopia PC- 3 Baseline Survey, 2005
The survey indicated that OVC were, on average, older than non-OVC, with 57% of OVC
being age 10 and above, while only 47% of non-OVC were 10 years or older. This
phenomenon holds true for most of the surveyed towns. This is primarily a reflection of the
passage of time. The older a child is, the more likely s/he is to have lost one or both
parents.
Fig-3.2 Age Distribution of Children in Sample Households- OVC vs. Non-OVC: PC- 3
Baseline Survey Result, 2005- Ethiopia.
Analysis of sex distribution also depicts that higher proportion of Female fall under OVC
category relative to the non-OVC counterpart. OVC Female-to-Male ratio was significantly
above 1 for most of the study towns (X 2 =6.2, p=0.013), but closer to 1 for Non-OVCs
group(Table 3.3).
Table 3.3. Age and Sex Distribution of Children in the Sample Households- OVC vs. Non-OVC
Across the Seven Selected Towns of Ethiopia PC- 3 Baseline Survey, 2005
Shashemen Dire Bahir
Main Addis Ababa Nazareth e Awassa Assaita Dawa Dar Overall
Characteristics n % n % n % n % N % n % n % n %
Valid N= 2691 1500 1467 1350 560 1839 528 9935
Age Distribution of Children Covered in Households Survey
OVC (nN=) 1234 720 664 688 224 671 239 4440
Under 5 212 17.2 14420.0 97 14.612117.6 56 25 124 18.5 28 11.7 78217.6
5-9 293 23.7 18225.3 193 29.118426.7 4218.8 191 28.5 51 21.3113625.6
10-14 413 33.5 24433.9 226 34.023834.6 7935.3 213 31.7 89 37.2150233.8
15-17 316 25.6 15020.8 148 22.314521.1 47 21 143 21.3 71 29.7102023.0
Mean age 10.3 9.8 10.1 9.8 9.2 9.6 11.3 10.0
NON OVC (nN=) 1457 780 803 662 336 1168 289 5495
Under 5 290 19.9 19825.4 215 26.815723.712336.6 348 29.8 62 21.5139325.4
5-9 375 25.7 20526.3 245 30.519829.9 8625.6 311 26.6 81 28150127.3
10-14 514 35.3 23229.7 225 2820831.4 7321.7 329 28.2 97 33.6167830.5
15-17 278 19.1 14518.6 118 14.7 99 15 5416.1 180 15.4 49 17 92316.8
Mean age 9.7 8.9 8.4 8.7 7.7 8.4 9.2 8.8
Sex Distribution of Children
OVC (nN) 1234 720 664 688 224 671 239 4440
Male 529 42.9 34047.2 294 44.333949.311049.1 333 49.6 108 45.2205346.2
Female 705 57.1 38052.8 370 55.734950.711450.9 338 50.4 131 54.8238753.8
F/M ratio 1.33 1.12 1.26 1.03 1.04 1.02 1.21 1.16
NON OVC (nN) 1457 780 803 662 336 1168 289 5495
Male 707 48.5 37247.7 400 49.830646.217552.1 593 50.8 126 43.6267948.8
Female 750 51.5 40852.3 403 50.235653.816147.9 575 49.2 163 56.4281651.2
F/M ratio 1.06 1.10 1.01 1.16 0.92 0.97 1.29 1.05
Data on the living arrangements of orphaned children was obtained from 2168 out
of the 2171 orphans (Table-3.4). Roughly half of orphans (48%) live with the
surviving parent, while more than 52% live with non-parents household heads, for
example, grandparents (15%), Aunts /Uncles (12%), Brothers/Sisters (~8%), or
Other Relatives (11.5%), (Fig.3.3). About 4% of orphan children are living with
non-relatives, while 1.5% of total orphans (2.2% of female orphans) are, reportedly,
living with their employer as helper (Figure 3.3, Table 3.4).
Fig-3.3 Living Arrangements of Orphaned Children: Result of PC- 3 Baseline Survey in Seven
Selected Towns of Ethiopia, 2005
The distribution of living arrangements is similar across study towns, except that
orphans in Assaita are less likely to live with a grandparent and that orphans in
Bahir Dahar are more likely to be living with an employer (7%) and non-relatives
(6%) – suggesting that orphans in this town may be more vulnerable to household
labor. (Table 3.5).
Table-3.5: Living Arrangements of Orphaned Children Across the Seven Selected Towns of
Ethiopia: PC-3 Baseline Survey Result -2005
Addis Ababa Nazareth. Shashemene Awassa Assaita Dire Dawa Bahir Dar
Living With n % n % n % n % n % n % n %
Valid N= 575 355 326 298 100 338 175
Surviving parent 266 46.3 174 49.0 145 44.5 136 45.6 58 58.0 173 51.2 81 46.3
Grand parent 90 15.7 79 22.3 48 14.7 35 11.7 5 5.0 46 13.6 24 13.7
Uncle/Anunt 78 13.6 35 9.9 31 9.5 35 11.7 14 14.0 49 14.5 21 12.0
Brother/Sister 54 9.4 21 5.9 33 10.1 19 6.4 4 4.0 20 5.9 19 10.9
Other Relatives 68 11.8 35 9.86 43 13.2 48 16.1 11 11 38 11.2 7 4
Non Relatives 11 1.9 11 3.1 24 7.4 21 7.0 2 2.0 12 3.6 11 6.3
Employer (as Helper) 8 1.4 2 0.6 4 1.3 6 6.0 12 6.9
As illustrated in Table 3.6, child mobility is higher both among individual OVC and
within households with OVC. On the average, about 15% of households with OVC
reported at least one child had moved out of the household within 12 months prior
to the survey, while only 11% of non-OVC households reported this. This is most
common in Assaita, Awassa and Shashemene towns.
Nearly 30% of children identified as OVC had joined (moved into) the family they
are currently living with from another family/household, while this figure stood at
only 10% for non-OVC group. The proportion of OVC who had joined their current
household from another household ranged from 22% in Assaita to 40% in Awassa.
Among OVC who had moved from one household to another, 80% had lived in one
household prior to the current household, and 20% had lived in two or more.
In Assaita, Shashemene and Awassa children are especially likely to have lived in
multiple households.
Table-3.6a: Child Mobility among Children and Households in Seven Selected Towns of
Ethiopia: PC- 3 Baseline Survey 2005
Addis Ababa Nazareth Shashemene Awassa Assaita Dire Dawa Bahir Dar Overall
n % n % n % n % n % n % n % n %
“Mobility of Children in Households”
HHs with OVC
Valid N= 647 345 279 276 121 327 123 2118
Reported Cases 57 8.8 42 12.2 52 18.6 71 25.7 34 28.1 44 13.5 14 11.4 314 14.8
HHs without OVC
Valid N= 566 263 212 205 124 395 113 1878
Reported Cases 30 5.3 25 9.5 31 14.6 37 18.0 28 22.6 40 10.1 9 8.0 200 10.5
Mobility of Individual Children
OVC
Valid N= 1218 710 661 634 223 671 231 4348
Reported Cases 334 27.417324.37 259 39.2201 31.7 51 22.9 204 30.4 79 34.2 1301 29.9
non-OVC
Valid N= 1448 773 796 609 330 1167 289 5412
Reported Cases 152 10.5 72 9.31 107 13.4 76 12.5 21 6.36 93 7.97 24 8.3 545 10.1
Among Mobile Children, Number of Households Lived In
OVC
Valid N= 334 173 259 201 51 204 79 1301
1 275 82.3140 80.9 192 74.1153 76.1 29 56.9 189 92.6 61 77.2 1039 79.9
2 42 12.6 32 18.5 67 25.9 47 23.4 22 43.1 15 7.4 14 17.7 239 18.4
3& More 17 5.1 1 0.6 0 0.0 1 0.5 0 0.0 0 0.0 4 5.1 23 1.8
non-OVC
Valid N 152 72 107 76 21 93 24 545
1 130 85.5 64 88.9 63 58.9 58 76.3 11 52.4 90 96.8 19 79.2 435 79.8
2 15 9.9 8 11.1 41 38.3 16 21.1 10 47.6 3 3.2 4 16.7 97 17.8
3& More 7 4.6 0 0.0 3 2.8 2 2.6 0 0.0 0 0.0 1 4.2 13 2.4
Looking into family cohesion, the result of the survey depicts that only about a
quarter of mobile OVC and mobile orphans are living with all siblings (Table-3.6b).
This implies that extent of family disintegration on OVC in general and Orphan
children in particular are very.
The major reasons for OVC to move from former household into the current
household are:- death of parent/ guardian (~35%), unmet basic needs (~33%),
inability to go to school (while living with the former household), i.e. to have access
to school (~20%) and other (~13%); Table 3.7a. The inability to go to school while
living with the former household is more pronounced on girl child than the male , and
on older children age 10 –17 than the young once. On the other hand unmet basic needs as
a deriving force to move the OVC from one household to another is more important for
children of young age bracket (table 3.7b,and 3.7c) .
Table-3.6b: Extent of Family cohesion among the mobile children in Seven Selected
Towns of Ethiopia: PC- 3 Baseline Survey 2005
Addis Ababa Nazareth Shashemene Awassa Assaita Dire Dawa Bahir Dar Overall
n % n % n % n % n % n % n % n %
Proportion of Mobile OVC Currently Living with All Siblings
OVC in general
Valid N= 334 170 253 201 52 204 70 1290
Reported Cases 77 23.1 38 22.4 62 24.5 3918.8 14 26.9 63 30.9 11 15.7 304 23.6
Orphaned Child
Valid N= 197 114 164 134 35 128 58 830
Reported Cases 45 22.8 30 26.3 36 22.0 2518.7 5 14.3 46 35.9 8 13.8 195 23.5
Table-3.7a: Causes for the OVC to move into the current household from previous one
in seven selected Towns of Ethiopia:- PC3 Baseline Survey-2005.
Reasons for moving out Addis Nazaret Shashemen Awass Assait Dire Bahir
of Ababa h e a a Dawa Dar Overall
the former HH n % n % n % n % n % n % n % n %
Valid N= 332 170 257 202 51 200 72 1284
Death of Parent's/ 44 34.
Guardian 90 27.1 52 30.6 89 34.6 81 40.1 18 35.3 86 43.0 2636.1 2 4
41 32.
Unmet basic needs 99 29.8 67 39.4 106 41.2 58 28.7 21 41.2 47 23.5 2027.8 8 6
25 19.
Inability to go to School 94 28.3 30 17.6 33 12.8 27 13.4 8 15.7 44 22.0 1622.2 2 6
17 13.
Other reasons 49 14.8 21 12.4 29 11.3 36 17.8 4 7.8 23 11.5 1013.9 2 4
Table-3.7b: Causes for the OVC to move into the current household from previous one
disaggregated by gender, average of the seven study towns of Ethiopia:- PC3 Baseline
Survey-2005.
Table-3.7C: Causes for the OVC to move into the current household from previous one -
disaggregated by age group, average of the seven study towns of Ethiopia:- PC3 Baseline
Survey-2005.
Age Group of Children in the Household Survey
Reasons for moving out of Under 5 5-9 10-14 15-17 Total
the former household n % n % n % n % n %
Valid N= 105 285 522 372 1284
Unmet basic needs 44 41.9 94 33 159 30.5 121 32.5 418 32.6
Inability to go to School 11 10.5 41 14.4 112 21.5 88 23.7 252 19.6
Death of Parent's/
Guardian 17 16.2 108 37.9 202 38.7 115 30.9 442 34.4
Other 33 31.4 42 14.7 49 9.39 48 12.9 172 13.4
The age and sex distribution data presented in table 4.1 shows that more than half of
household heads interviewed were in age range of 30-49 years. The age distribution of
household heads, in general, follows a normal distribution curve; (fig-4.1). The mean age
of household heads (those with OVC and without OVC combined ) is ~43 years with Std.
Deviation of 13.3 and range of 16-98 years.
Households with OVC are generally headed by more elderly people. About 40% of heads of
households with OVC are 50 years and over (ranging from 15% in Assaita to ~47% in
Addis Ababa), whereas, only ~17% of heads of households without OVC belong to age
category of 50 and over. This implies OVC are more concentrated in households headed by
elderly than non-OVC.
Overall, more than twice as many households are headed by males rather than females (64%
versus 36%.) However the majority (52%) of OVC households are headed by females
except in Awassa and Assaita ; while most ( ~82%) Non-OVC households are headed by
males.
A little more than half (~51%) of heads of OVC households are out of the wedlock
(including unmarried, widows, divorcee) while 86% of HH without OVC are headed by a
person, curretly, in the wedlock (i.e married). The large proportion (38%) of HH with OVC
headed by a widow(er) indicates the high level of vulnerability of children in these
households. The proportion of OVC households headed by a widow(er) ranges from 28%
in Assaita to 46% in Nazareth. In Addis Ababa it is ~41% (Annex 4.1)
Table- 4.1: Sex and Age Distribution of Household Heads Classified into With and
Without OVC- in Seven Selected Towns of Ethiopia- PC3 Baseline Survey-2005.
Table 4.2: Marital status and relationship of the head of household to the family
Classified into With and Without OVC: PC3 Baseline Survey in Seven Selected Towns of
Ethiopia -2005
Analysis of religious affiliation of household heads depicts that the majority (~71%) of HH
with OVC belongs to Orthodox Christian (ranging from 54% in Dire Dawa to 87% in Addis
Ababa and 94% in Bahir Dar), except for in Assaita where 77% of OVC HH are Muslims,
which is expected, given the religious practice of the overall population in this area. Overall
it is observed that households with OVC are more likely to be Orthodox and less likely to be
Muslim than households without OVC. This implies the need to engage Orthodox churches
and church leaders in care and support of OVC in all possible dimensions, including
spiritual services to address this segment of the society.
Table- 4.3: Literacy status (ability to read and write) and Religion Affiliation of Household
Heads Classified into With and Without OVC: PC3 Baseline Survey in seven selected Towns
of Ethiopia: -2005 ax
ax
See also Annex 4.1 for data by study town
Table-4.4: Employment Status of Household Heads Classified into OVC and Non-OVC
Households PC-3 Baseline Survey Result of Seven Selected Towns of Ethiopia -2005 ax
ax
See also Annex 4.2 for data by study town
Most common sources of income for HHs are regular employment (salary), Micro
Business run by Adults, and Daily Labor, Rental Income, Pension Funds,
Assistance/Gifts or Remittances and Micro Business/Street vendoring run by
Children. Micro Business/Street vendoring -run by Children is a clear indication
of child labor exploitation especially for OVC household. (See section 7.2 for
detailed analysis on this)
Relative to non-OVC households, those households with OVC are more likely to
report- Daily Labor, Rental Income, Pension Fund, Assistance/Gift or Remittances
and Micro Business/Street vendoring -run by Children and Micro Business -run by
adult- as their main source of income. On the other hand OVC households are less
likely to report income from regular employment. This implies that livelihoods of
OVC are under severe threat as compared to non-OVC (i.e. income is less
predictable and reliable to sustain the livelihoods of the family).
Table-4.5 Sources of Livelihood (Income) for OVC and Non-OVC Households: PC-3
Baseline Survey of Seven Selected Towns of Ethiopia -2005
Regarding ownership of houses, 39.5% of respondents live in their own or their parents’
houses. Around a third (34.1%) reside in government owned houses. A little less than a
quarter (~23%) live in privately owned houses while the remaining 3.7% live in relatives’
houses and houses owned by others. OVC households are more likely to be in government
owned houses and less likely to rent their homes than are non-OVC households.
Table 4.6a House ownership among OVC and non-OVC Households: PC-3 Baseline
Survey of Seven Selected Towns of Ethiopia -2005
The findings on the housing conditions in terms of type of floor, ceiling, roof cover and
number of rooms is presented in the table below (Table 4.6b). Floor type of the house for
nearly 50% of the households, both HHs with OVC and HHs without OVC, is cement (lean
concrete) floor followed by mud (dirt) floor (~34%). However, there is still a disparity
between HHs with OVC and HHs without OVC in the type of floor. Proportion of HHs
with OVC living in a house with mud floor is ~38% as compared to HHs without OVC,
which stand at 28%.
As to the type of ceiling in the house, the common type of ceiling for both HHs with OVC
and HHs without OVC are ‘Abujadied’ Cloth (33%) and Polyethylene Sheet (~24%)), while
about 32% of HHs live in a house without ceiling. The pattern and Proportion is nearly the
same for both HHs without OVC and HHs with OVC. The type of roofing for the majority
(~97%) of both category of HHs is ‘Corrugated Iron Sheet’, except in Assayita and Dire
Dawa where ~91% and 5% are living in house roofed by wood and mud, which is
associated with the climatic condition of these areas. In terms of rooms over 60% of HHs,
both category, are living in house with less than or equal to 2 rooms. But it has to be noted
that a little over a quarter of sample households are living in a house with only one room
Overall, data on the housing conditions reveals that there were few differences between
OVC and non-OVC households with respect to the types of ceiling and average number of
rooms. However, OVC households were more likely to have a mud rather than a concrete
floor, indicating these households may be poorer.
Table 4.6b Housing conditions in terms of type of floor, ceiling, roof cover and number of
rooms among OVC and non-OVC Households: PC-3 Baseline Survey of Seven Selected Towns
of Ethiopia -2005
The most frequently reported household assets include tape recorders, TV sets, sofa chairs,
foam/spring mattress and house. Non-OVC households tended to report more household
possessions, except with respect to ownership of house/building (Table 4.7). This means
that OVC households have fewer assets that could be liquidated to help cope with a
shock, such as an adult death.
Table 4.7. Asset Ownership of Households OVC vs non-OVC: PC-3 Baseline Survey of Seven
Selected Towns of Ethiopia -2005
The following Figure (Fig 4.2) depicts that the dependency ratio is highest among
households headed by the elderly; OVC households headed by women; and OVC
households generally, in all study towns. Highest dependency ratios, all households in
general and OVC household in particular, were observed in Shashemene and Dire Dawa
towns (see annex 4.2b for details).
Fig 4.2 Mean Dependency Ratio Among Different Category of HHs: PC3 Baseline
Survey in Seven Study Towns of Ethiopia: -2005
To assess the capacity of families to provide basic material needs for children under their
care, questions on the availability of three minimum basic material needs were posed,
namely: Clothing- two set per year, Shoes- two pairs per year, and Bedding (Blanket and
or bed sheets, depending on the local norms) were asked for each, 5–17 years old, child in
the household. This indicator is a proxy measure of the number of children in need of
services and support such as school uniforms.
Overall, there is significant difference (X2 =48.2 , p<.001) between OVC and non OVC in
terms of adequate clothing. It is only 18.2 % of the OVC had two sets of clothing and two
pairs of shoes per year, while this stands at 25.3% for the non-OVC. With regards to
bedding (Blanket and/or bed sheets) there are no considerable differences between OVC
and non-OVC.
The inter–town comparison reveals that the situation of OVC relative to non-OVC is
especially poor in Awassa and Bahir Dar (where 11.3% and 15.5%, respectively, of OVC
have two sets of clothing and two sets of shoes). A ratio was developed to show the relative
material well-being of OVC versus non-OVC. The OVC to non-OVC ratio, in these towns,
is about 0.5 (i.e. OVC are about half as likely as non-OVC to have adequate clothing.) In
general, most children had adequate bedding and there were fewer OVC vs. non-OVC
differenced with respect to bedding.
Table-4.8: Level of minimum basic material (clothing and bedding) needs met for OVC vs Non-
OVC: PC3 Baseline Survey in Seven Selected Towns of Ethiopia:-2005, Ethiopia.
Addis Ababa Nazareth Shashemene Awassa Assaita Dire Dawa Bahir Dar Overall
n % n % n % n % n % n % n % n %
Adequacy of day clothing
OVC (N=) 910 467 473 487 138 475 168 3118
Adequate 147 16.2 71 15.2 77 16.3 5511.3 9367.4 99 20.8 26 15.5 56818.2
NON OVC (N=) 1044 468 518 422 186 706 180 3524
Adequate 242 23.2 101 21.6 106 20.5 9121.615683.9 143 20.3 52 28.9 89125.3
OVC vs. non-OVC
Ratio 0.70 0.70 0.80 0.52 0.80 1.03 0.54 0.72
Adequacy of bedding 5
OVC CHILD (N=) 1051 578 564 562 167 547 210 3679
Adequate 1008 95.9 463 80.1 486 86.248486.112776.0 527 96.3 172 81.9 326788.8
NON OVC CHILDN= 1178 583 597 510 216 812 224 4120
Adequate 1150 97.6 450 77.2 552 92.544887.816475.9 775 95.4 196 87.5 373590.7
OVC vs. NOVC Ratio 0.98 1.04 0.93 0.98 1.00 1.01 0.94 0.98
5
For Diredawa and Asaita Sheet and plus; for other towns Blanket and plus considered as adequate)
To determine how widespread food insecurity is among the population and compare the
food security status of households with OVC to the food security status of households
without OVC, heads of the household were asked a series of questions that characterize
household difficulty meeting their food needs. A composite variable was developed to
identify food secure and food insecure household. The variable incorporates the ability
household to provide children and adults three meals per day throughout the year, the
quality of the food (relative to local standards), and worries that the food might run out.
Close to 40% of the respondent households have reported that they were able to provide
children and adults with three or more meals of medium and higher quality per day, while
close to half of the non-OVC households (47.1%) were able to provide the same. OVC
households who responded to provide children and adults with more than three meals per
day of any quality were 83.5% while it was 91% among the non-OVC households. This
indicates that OVC households experience more food insecurity than non-OVC households,
possibly as a result of death of breadwinners or illness of an adult family member. This has
a direct relationship to the decrease in the amount of family income as well as the high cost
of healthcare expenses for the sick members.
Coping mechanisms at times of household food shortages include reducing the amount of
food served at meals, reducing the number of meals per day or even skipping meals for an
entire day depending on the severity of shortage. Accordingly, among the respondent OVC
households 4.8% have reported to “mostly” decrease the amount of meals served to children
1-5 years of age, while this was reported only by 2.6% of the non-OVC households showing
a two-fold occurrence. Moreover, it was observed that this measure of “mostly” reducing
the amount of meals among OVC households to be more prevalent in age group of 5 to 15
years which is at 15.2% while this was 6.8% in the same age group of children from the
non-OVC households.
The proportion of households who reported reduction of meals “mostly” among children of
the age group 15-18 years, however, were relatively lower, 6.2% and 2.1% in OVC and
non-OVC households respectively. Reduction of meals served to adult women was reported
to be “mostly” practiced at a proportion of 36.1% among OVC households whereas it was
reported by 27.7% of the non-OVC household respondents. Meals reduction was also
reported as “mostly” practiced among adult men by 34% and 26.3% of the OVC and non-
OVC households respectively. From these findings, one can clearly notice that reduction of
amount of food served is a function of age. Reduction of mount of food served is less
practiced among young children and proportionally more elder children are subject to
reduction of amount of meals served and this gets even more frequent among youth and
adults. Among adults women are more subject to reduction of meals served during times of
household food shortage. This could be partly due to the norm and values to care for
orphans especially in their early childhood. Studies came out of sub-Sahara African
countries showed that malnutrition to be more prevalent in OVC under 5, while this is
equivocal in other studies at the same time.
Table 4.9: Proportion of people subject to different measures taken by households to cope
up with food shortages
Groups of HH members
Children Children Children Adult Adult Men
1-5 yrs 5-15 yrs 15-18 yrs Women
Percentage of HHs reported to ‘ reduce 3.9 11.6 4.3 33.0 30.7
meals’
Percentage of HHs reported to “ skip 2.1 7.3 3.2 26 23.8
some meals’
Percentage of HHs reported to “ skip 0.2 0.9 3.4 4.6 3.2
meals for an entire day’
Percentage of HHs reported to ‘weight 7.5 18.1 5.1 36.9 33.9
loss’
In the case of skipping some meals of the day by children 1-5 years of age, there was no
major difference observed on the proportions among OVC and non-OVC households.
Among the age group of 5 to15 years, skipping some meals of the day was reported by 9.9%
of the OVC households while it was reported only by 3.9% of the non-OVC households.
Among the age group 15-18 years, 4.8% of the OVC households have reported to skip some
meals of the day compared to 1.5% of the non-OVC households. A higher proportion of
skipping of meals ‘mostly’ was reported among adult women and men, 29% and 27.9%
respectively in the OVC households respectively; 21% and 18.4% in the non-OVC
households respectively.
Regarding skipping meals for an entire day, among the age group 1-5 the proportion of
households, both OVC and non-OVC, who reported skipping meals for a day were very
small, below 1%. In the age group 5-15 years, the percentage of respondents who replied
skipping meals ‘mostly’ for entire days were 1.5% among the OVC households while it was
only 0.5% among the non-OVC group. Skipping meals for an entire day was even worse a
practice among the age group 15-18 years, with 4% and 1.9% in the OVC and non-OVC
groups respectively. Likewise, 5.8% of the OVC households have reported that skipping
meals for a day was a practice ‘mostly’ while it was reported by 2.6% of the non-OVC
households. As regards adult men skipping meals for a day was reported as ‘mostly’ by
3.9% of the OVC households and 2.3% of the non-OCV households.
Generally, several studies showed that where the situation of household food scarcity is
becoming severe, weight loss of family members would happen. In this survey, 6.9% of
respondents from OVC households have reported that there was weight loss among children
1-5 years of age, while about 8.2% of the non-OVC households have reported the same. In
the age group of children 5-15 years, the proportion of OVC households reported weight
loss was 22.8% while in the non-OVC counterpart of this age group the proportion of
weight loss was at 12%. Percent of households who reported weight loss to have happened
among children of age 15-18 years were 7.6% and 2.4% in OVC and non-OVC households
respectively. In addition to this, majority of the FG discussants and key informants argued
strongly that lack of food at household level is the major challenge of HH with OVC.
Regarding weight loss among adults, 39.4% of the OVC households have reported weight
loss to occur among women while the proportion of weight loss was reported by 32.7% of
the non-OVC households. Regarding weight loss among adult men, 38.1% of the OVC
households and 28.3% of the non-OVC households respectively have reported weight loss
have happened.
A total of 1339 episodes of illnesses (out of 9935 children were reported to occur among children
under 5 in the two weeks prior to the survey. The majority of these episodes (1079) were from the
non-OVC group.
The most prevalent illnesses included common cold (47.8%), diarrhea (9.4%), malaria (8.4%) and
non-specific abdominal pain (4.1%). More than a third (37%) were taken to government clinics or
hospitals for care. Another 29.8% were taken to holy water, while (24.1%) were taken to private
clinics and hospitals. Other sources of were traditional treatment (6.3%), village doctors (0.7%) and
self-treatment by buying medications (0.9%).
Focus group discussants and key informants in different cities reported that OVC are exposed to
types of health problems. Some and major problems were predisposition of the children to sexual
assault and harassment that often end up in rape with devastating consequences like unwanted
pregnancies, contracting RTIs and even HIV, and lasting mental health problems. These OVC are
also prone to substance abuse that made them more vulnerable to contract the aforementioned
diseases.
Fig 5.1
45.0
40.0
35.0
Gov't Clinic/Hospital
30.0 Private Clinic/Hospital
Percentages
Village Doctor
25.0 Traditional Treatmenmt
20.0 Holy Water
Nothing
15.0 Bought & took medicine
Others
10.0
5.0
0.0
Orphans Non-orphans Total
Categories of children
Both OVC and non-OVC (about 86%) reported that their parents should talk about their health
status including HIV/AIDS. About 1.2 % of the OVC perceived that PLWHA faced discrimination
and stigma while only half (0.6%) of the non –OVC believed the same. The difference is statistically
significant. This could reflect the day today experience of OVC from their counterparts living and
playing with and partly their sensitiveness to their surroundings might contribute.
In this study, the prevalence of children who have started sex was 2.4%. Though different studies in
Ethiopia has shown that the sexual debut below 18 both in towns and country sides was above 10%,
this study couldn’t show relative figure for the age at first sexual contact. This could be due to the
method used to elicit such kind of sensitive responses. This figure is very high in most African
countries and it was found out that the age is earlier for OVC than non – OVC.
Looking into the ratio of (OVC to non-OVC) incidence of illnesses during the two week period
ahead of the survey, it was reported that relatively more OVC to be sick than the non-OVC. A
higher ratio of OVC to non-OVC was reported for sickness in the fortnight period in Nazareth and
Assaita, 1.83 and 1.27 respectively.
Table 5.1 History of illness and access to Formal Health Services- OVC vs.Non-OVC: PC3
Baseline Survey in seven selected towns of Ethiopia-2005
Addis Dire Bahir
Ababa Nazareth Shashemene Awassa Assaita Dawa Dar Overall
n % n % n % n % n % n % n % n %
Episode of illness in the last 2 weeks before the survey
OVC
Valid N= 1232 715 663 686 224 670 238 4428
13. 14. 23. 17. 14.
Frequency 161 1 99 13.8 75 11.3 102 9 53 7 120 9 24 10.1 634 3
NON OVC
Valid N= 1453 778 802 661 334 1161 284 5473
12. 13. 18. 21. 13.
Frequency 177 2 59 7.6 84 10.5 86 0 62 6 250 5 30 10.6 748 7
Illness Ratio 1.07 1.83 1.08 1.14 1.27 0.83 0.95 1.05
Access to Formal Healthcare services
OVC
Valid N= 158 97 70 103 53 117 24 622
46. 52. 71. 48. 53.
Frequency 73 2 56 57.7 42 60 54 4 38 7 57 7 14 58.3 334 7
NON OVC
Valid N= 173 59 83 86 62 247 30 740
The FG discussants and key informants reported that OVC have relatively poor access to healthcare
and even if they have access there is delay in seeking medical care which often complicate the
course of the illnesses.
The overall access to private or communal toilets among respondents was 90.5%. Access
was very similar for OVC and non-OVC households, but varied by location. Households in
Assaita have the least access to toilets, while those in Awassa have the greatest access.
Most households have access to private toilets.
Fig 5.2
105.0
100.0
Percentage
85.0
80.0
Study sites
The result of the survey indicates, overall, ~93% of all school age (5-17 years) children
have been at school at some point in time, while the proportion of school age children currently
attending school stand at 88.5%. The survey reveals that significantly lower proportion of
school age OVC is attending school as compared to non-OVC (86.4 vs. 90.4, P<.000)
(Table 5.3a).
Moreover, current school attendance varies between the study towns. The disparity of
current school attendance between OVC and non-OVC is significantly very high in
Shashemene (~78% vs ~87%, P<.000), followed by Awasa (P < .005) and Addis Ababa (p<
.001). The only exception is in Dire-Dawa, where OVC vs Non-OVC Current School
Attendance Ratio is closer to 1, indicating that the disparity is very minimal (Table 5.3a).
Table-5.3a: Percentage of School age children attending School-Classified into OVC and non-
OVC: PC3 Baseline Survey in Seven Selected Towns of Ethiopia –2005
Addis BahirDa
Ababa NazarethShashemene Awassa Assiata Dire Dawa r Overall
n % n % n % n % n % n % n % n %
OVC
Valid N= 972 554 563 522 154 536 205 3506
Ever been in
school 921 94.8 527 95.1 492 87.448893.513990.3 487 90.9 181 88.3323592.3
Currently Attending
School 870 89.5 490 88.4 441 78.545587.213285.7 470 87.7 170 82.9302886.4
non-OVC
Valid N= 1089 553 581 477 186 802 219 3907
Ever been in
school 1054 96.8 518 93.7 535 92.145795.817091.4 734 91.5 195 89.0366393.8
Currently Attending
School 1018 93.5 502 90.8 506 87.144292.716588.7 707 88.2 191 87.2353190.4
OVC to non-OVC
Current School
Attendance Ratio 0.96 0.97 0.90 0.94 0.97 0.99 0.95 0.96
Further analysis of school attendance by gender reveals that proportion of Female school
age children ‘ever been to school’ as well as ‘currently attending school’ is significantly
lower than that of male school age children. Although the pattern seems the same among
OVC and non-OVC, the gender difference is more pronounced among OVC group, and
OVC to non-OVC current school attendance ratio is lower for Female as compared to that
of male, implying that Female OVC need more attention to improve school attendance
(Table 5.3 b). On the other hand, analysis of school attendance by age group depicts the
proportion of OVC currently attending school is getting lower and lower as compared to
non-OVC as the age increase. This variation is statistically significant (Table 5.3c)
Table-5.3b: Percentage of OVC and non-OVC School age children r attending School-
Classified by Gender: PC3 Baseline Survey in Seven Selected Towns of Ethiopia –2005
Table-5.3b: Percentage of OVC and non-OVC School age children attending School-Classified
by age group: PC3 Baseline Survey in Seven Selected Towns of Ethiopia –2005
With regards to the type of school parents/guardians are sending the children for learning,
the survey result reveals that a larger proportion (~72%) of OVC are using government
school than the non-OVC counterpart (~61%). The inverse is true for private school (32.4%
of non-OVC group vs. 22.5% of OVC group). Utilization rate of other types of schools is
similar among OVC and non-OVC.
Table-5.4: Types of school attended by school age (5-17) children classified by OVC and
Non-OVC: PC3 Baseline Survey in Seven Selected Towns of Ethiopia –2005
Bahir
Addis Ababa NazarethShashemene Awassa Assaita Dire Dawa Dar Overall
n % n % n % n % n % n % n % n %
Type of School being Attended
OVC CHILD
Valid N= 928 507 446 489 143 481 172 3166
Government 543 58.5 375 74.0 306 68.637075.7143 100 375 78.0 154 89.5226671.6
Private 275 29.6 101 19.9 130 29.110521.5 84 17.5 16 9.3 71122.5
Community 94 10.1 24 4.7 8 1.8 6 1.2 132 4.2
Others 16 1.7 7 1.4 2 0.4 8 1.6 0 0.0 22 4.6 2 1.2 57 1.8
NON OVC CHILD
Valid N= 1097 530 515 467 180 721 195 3705
Government 528 48.1 330 62.3 297 57.726256.117798.3 540 74.9 137 70.3227161.3
Private 405 36.9 174 32.8 209 40.619441.5 2 1.1 161 22.3 55 28.2120032.4
Community 149 13.6 23 4.3 9 1.7 8 1.7 2 1.0 191 5.2
Others 15 1.4 3 0.6 0 0.0 3 0.6 1 0.6 20 2.8 1 0.5 43 1.2
For this survey educational wastage is defined and measured as ‘absenteeism from school
for more than one term; i.e ‘ any child who is absent from the school for more than one
term is counted, whether he/she is currently in school or out of school’. Educational
wastage among OVC was found to be significantly higher (P< .001) than non-OVC (9.1%
versus, 5.2%, and OVC vs. non-OVC ratio=1.8) (Table 5.5a).
The differences in educational wastage- OVC vs. non-OVC were more pronounced in
certain towns, such as Bahir Dar, (OVC vs. Non-OVC school wastage ratio reached 3.3.
Things appear more equitable in Assaita with respect to this and other education indicators.
This may be related to the culture of the community, as understood during the FGD, in
which children of the haves and the have-nots share food and there is no worry for OVC to
search for food by missing school.
Further analysis of extent of educational wastage by gender reveals that statistically there is
no considerable difference between Female and Male school age children (Table 5.5b).
On the other hand, analysis of educational wastage by age group depicts the proportion of
both OVC and non-OVC absent from the school for more than one term is getting higher
and higher as the age increases; and this variation is statistically significant (p<.001).
However, OVC are found to be more disadvantaged as compared to non-OVC at early age
(Table 5.5c)
Table-5.5a Extent of ‘Education Wastage’ among OVC and non-OVC: PC3 Baseline
Survey in Seven Selected Towns of Ethiopia –2005
Bahir
Addis Ababa Nazareth Shashemene Awassa Assaita Dire Dawa Dar Overall
n % n % n % n % n % n % n % n %
OVC
Valid N= 912 501 443 476 143 474 166 3115
Frequency 64 7.0 46 9.2 40 9.0 5912.4 10 7.0 44 9.3 20 12.0 283 9.1
non-OVC
Valid N= 1072 524 514 461 178 720 192 3661
Frequency 42 3.9 23 4.4 38 7.4 28 6.1 14 7.9 38 5.3 7 3.6 190 5.2
OVC vs non-
OVC Ratio 1.8 2.1 1.2 2.0 0.9 1.8 3.3 1.8
Table 5.6 gives an overview of the major reasons for not attending schools and absenteeism
from school for more than one term (educational wastage). Financial difficulty is the most
common reason among both OVC (50%) and Non-OVC (41%). Illness is a more common
barrier for OVC (11%) than for non-OVC (5.6%), perhaps due partly to their lower access
to modern health care services (see Section 5.1)
Table-5.5C Extent of Education Wastage among OVC and non-OVC Classified by age
group: PC3 Baseline Survey in Seven Selected Towns of Ethiopia –2005
OVC to
non-OVC
OVC non-OVC Total Ratio
Age 5-9
Valid N= 893 1221 2114
Cases Reported 49 5.5 28 2.3 77 3.6 2.4
Age 10-14
Valid N= 1307 1513 2820
Cases Reported 108 8.3 75 5.0 183 6.5 1.7
Age 15-17
Valid N= 780 760 1540
Cases Reported 116 14.9 76 10.0 192 12.5 1.5
Table- 5.6 Reasons for not Attending Schools Classified by OVC and Non-OVC: PC3
Baseline Survey in Seven Selected Towns of Ethiopia –2005
Bahir
Addis Ababa Nazareth ShashemeneAwassa Assaita Dire Dawa Dar Overall
n % N % n % n % n % n % n % n %
OVC (Valid N=) 106 66 120 73 23 68 37 493
Lack of time 8 7.5 3 4.5 6 5.0 1216.4 313.0 2 2.9 1 2.7 35 7.1
School Distance 1 0.9 3 4.5 1 1.5 5 1.0
Illness 11 10.4 812.1 14 11.7 811.0 1 4.3 8 11.8 3 8.1 53 10.8
Marriage 1 4.3 1 0.2
Lack of Educational
Facilities 3 2.8 3 2.5 3 4.1 2 8.7 2 2.9 13 2.6
Financial Difficulty 55 51.9 3857.6 75 62.5 3142.5 521.7 28 41.2 12 32.4244 49.5
Others 19 17.9 1015.2 9 7.5 1115.1 2 8.7 13 19.1 10 27.0 74 15.0
non-OVC
(Valid N=) 71 52 76 39 32 92 28 390
Lack of time 6 8.5 1 1.9 7 9.2 410.3 412.5 2 2.2 5 17.9 29 7.4
School Distance 6 8.5 1 1.3 1 2.6 1 1.1 1 3.6 10 2.6
Illness 1 1.4 611.5 3 3.9 410.3 2 6.3 5 5.4 1 3.6 22 5.6
Marriage 1 3.1 1 0.3
Lack of Educational
Facilities 1 1.4 1 2.6 9 9.8 11 2.8
Financial Difficulty 30 42.3 2548.1 49 64.5 1025.6 928.1 31 33.7 6 21.4 16041.0
Others 16 22.5 4 7.7 9 11.8 1025.6 412.5 16 17.4 2 7.1 6115.6
ii) Mood
vi) Functioning/capacity
The assessment tool consisted a total of 18 items on a scale of four points ranging from
strongly agree to strongly disagree. The items were constructed to measure the above-
mentioned psychological components as well as the overall psychosocial status of the study
population. The scores assigned to the items were as prescribed by the guide manual for
Monitoring and Evaluation of the National responses to OVC (2005). Thus, the total score
is an indicator of overall psychological health, while the scores on the subscales relate to
each of eight psychosocial dimensions. Based on the cut-off point prescribed by the manual,
a given score is determined to be adequate or inadequate. Accordingly, the maximum score
for a respondent on the composite scale is 72, and a score of 40 and above is considered as
adequate while any score below 40 is inadequate. For the subscales, the cut-off points vary
depending on the number of items allotted to measure each of the sub-components of
psychological health. Adequate score denotes positive emotional characteristics while
inadequate score indicates diminished well-being in the given category.
Table-6.1 Score Distributions for the OVC and non- OVC on the composite Scale of
psychological/emotional well-being
Addis Dire Bahir
Study Group Ababa Nazareth Shashemene Awassa Assaita Dawa Dar Overall
OVC Group
OVC Valid N= 154 65 73 61 50 102 48 541
% of OVC Youth
with adequate score 98.1 98.5 97.3 100 100.0 100.0 100.0 100
Mean score of
Adequate scorers 55.2
54.8 58.3 54.2 54.6 54.3 57.7 52.8 4
NON OVC Group
Valid N 146 43 52 56 50 114 40 495
% of Non-OVC
Youth with adequate
score 98.63 100.00 98.08 100 100.00 100.00 100.00 100
Mean score of
Adequate scorers
56.6 59.9 54.4 56.8 56.0 57.7 56.5 56.8
Ratio of OVC to Non-
OVC 0.99 0.98 0.99 0.95 1.00 1.00 1.00 0.98
As can be observed from Table-6.1, the overwhelming majority both in the OVC and non-
OVC groups have obtained adequate score in many towns covered by the survey, with
negligible differences between the two groups.
Key informants and FGD participants further explored psychsocial well-being and tended to
describe OVC as often being shy and timid, looking unhappy, grieving, sometimes crying,
being very reserved in many things, and appearing aloof.
In addition to the analysis done for the data obtained from the composite scale of
psychological health as indicated above, the data obtained from the sub-scales of
psychosocial components was also treated in a similar procedure. The results have been
shown as follows.
As explained some where in this report, the maximum score for each of the sub-scales of
psychosocial components covered by the study ranges from 4- 12 points. The points vary
depending on the number of items allotted for each of the dimensions. Similarly, the cut-off
points to determine adequate and inadequate scores are adjusted accordingly.)))
Table-6.3 shows the proportion or respondents who scored “adequate” in each of the
EIGHT? nine psychosocial components among both the OVC and non-OVC groups. With
one exception, the majority of respondents in both groups had adequate in all components.
But fewer than half of the respondents were adequate in the Functioning / Capacity
component, with a ratio of 0.91 OVC to non-OVC This component is intended to measure
problem-solving skills and personal competence of the respondent in facing the challenges
of life. . Less than 75% of respondent scored “adequate” in social connection and capacity.
OVC vs. non-OVC disparities are greatest in the stress and worry component and the
functioning/capability component. Otherwise there were few differences.
The three psychosocial components with the greatest proportion of inadequate scores ,
(Stress and worry, Social Connection and Functioning capabilities) were further analyzed as
shown the following Tables.
Survey towns
Addis Dire Bahir
Study Groups Ababa Nazareth Shashemene Awassa Assaita Dawa Dar Overall
.
OVC CHILD .
% With adequate
Score 84.2 82.3 83.7 84.9 89.1 92.6 80.0 85.2
There is local disparity in terms of the proportion and ratio of the respondents with adequate
score across the survey towns. Inadequate scores were most common in Nahir Dar and
Nazareth for both OVC and non-OVC. The trend across towns was that OVC were less
likely than non-OVC to have an adequate score. Statistical test was computed on the mean
scores of the two groups and the result has shown that significant differences exist between
the mean score. of the OVC and non-OVC respondents (P<. 05).
Table-6.5 Proportion and Mean Score Distribution for Functioning and Capability
Sub-scale Across the Study towns
Table-6.5 summarizes the output of Functioning and Capacity measures for the OVC and
non-OVC groups of respondents in the survey towns. As mentioned earlier, the majority of
respondents in both groups (about 58% and 54% for OVC and non-OVC respectively) did
not demonstrate adequate scores. There is substantial across towns, and in most cases,
unexpectedly, more OVC than non-OVC had adequate scores. Many theories in
psychology content that a certain among of stress is essential to motivate an individual
towards optimum utilization of his/her potential, which in turn determines one’s functioning
and competence.. It may be that OVC have actually gained functioning and capability
through enduring stress and worry and/or from programs emerging for OVC and increasing
psychosocial support to them. Interventions targeting OVC are now underway in many
survey towns, as noted during qualitative data collection, and may develop life skills.
Table-6.6 Proportion and Mean Score Distribution for Social Connection and
Capacity.
% With adequate
Score 70.1 79.1 74.2 83.8 71.7 80.6 55.6 73.9
Mean 6.1 6.2 5.7 5.7 5.9 6.1 5.9 41.4
Valid N 204 115 93 74 53 108 63 710
NON OVC CHILD
% with adequate
Score 77.2 74.4 63.8 77.3 71.2 78.0 65.3 74.2
Mean 6.2 6.6 5.7 6.0 5.9 6.1 5.9 42.4
Valid N 193 78 69 75 59 127 49 650
Ratio of OVC to
Non-OVC 0.91 1.06 1.16 1.08 1.01 1.03 0.85 1.00
Table-6.6, shows a an equivalent proportion of OVC and non-OVC with adequate scores
(74%), but with a lot of regional variation. In some towns, more OVC have adequate
scores, in other towns more non-OVC have adequate scores.
Similar to the findings observed under Functioning and Capability, the OVC group of
respondents appeared to be more advantaged as compared to their counterpart non-OVC in
the social connection capacity. This finding looks contrary to what is usually anticipated
from causal observation and intuitive thinking that presumes orphaned and vulnerable
children are disadvantaged in terms of social connection and capacity. In order to clear this
doubt further analysis was done using the procedure of Inter Item Correlation. It was found
out that there is a strong positive correction (r =0.73, P< 0.01) between
Functioning/capability and Social Connection/Capacity sub-scales, implying that the OVC
respondents who score adequate in the measures of functioning likely score adequate in the
measure of social connection, too. The average time lapsed in orphan hood as reported
somewhere in this study is 6 years and some children are lucky enough to get adequate
social support and alternative care which enable them wear off the traumatic experiences of
parental loss.
It may come as a surprise that OVC and NON-OVC would be some similar with respect to
social connection and capacity. But it support qualitative findings that in some cases at
least, OVC are in fact very well supported. This fact was noted in several of the case studies
held with orphan children. For instance, a child, 12, who was interviewed in Kirkos Sub-
City of Addis Ababa, described his situation as follows;
I was 8 when my mother died. I do not know my father. I was told he also died earlier. My
mother was sick for some times before her death. I had never expected her death. Even, the
moment she died, I did not know. When I come back home from playground I found many
people in the house. They told me that my mother was asleep. I did not know that they were
deceiving me. I learnt lately what death meant. After my mother died, our neighbor took me
to their home. Ever since I have been living with them as a member of the family. I am
happy that I am doing my schooling and d not feel bad for losing my parents.
As explained earlier, the respondents in the OVC group may have access to services
targeting that build their social skills and competence. It appeared that the social support
that children get access to and provided with is indispensable to foster favorable attitude and
good self esteem and this in turn has a lot to with building social connection and capacity of
orphan children.
The majority of OVC and non-OVC have adequate child-care giver connection scores.
Specifically, the proportion of OVC and non-OVC respondents with adequate scoreS is
88.20% and 91.21 % respectively. Local variation is observed across the study towns for
both the OVC and the non-OVC groups. However, more non-OVC than OVC have scored
adequate in all towns, except Awasa and Bahir Dar. The over all ratio of OVC to non-OVC
stands at 0.97 with in the range of 0.91 in Shashemene to 1.06 in Bahir Dar. Statistical test
was done to explore the differences in the mean score of the OVC and non-OVC group of
respondents in the measure of child-caregiver connection. The result revealed that there is a
significant difference between the mean scores of the two groups (p<0.05).
Overall, relative to non-OVC, OVC may be moderately disadvantaged in having stable and
positive emotional connection with adult caregiver. This finding has several practical
implications for actors and policy makers in the area of children and youth. Many studies
confirmed that emotional ties between children and adult caregivers determine the
children’s social behavior in later adult life. Children who have enjoyed positive emotional
bonding with their parents/guardians tend to develop high self-esteem, trust other people
and exhibit pro-social behavior. In contrast lack of such nurturing from an adult leads to
hostile, anger, unenthusiastic and anti-social behavior. Therefore, children and adolescents
who lack caring adults need to be supported so as to enable them develop a positive
relationship with the adult caregiver to lay down the basis for healthier development.
Table-6.8 Perception of House Heads on the Frequency of children's crying during the
normal day
Survey Towns
Addis Ababa Nazareth Shashemene Awassa Assaita Dire Dawa Bahir Dar Over all
An attempt was made to tape the responses of the household heads about how often the
children under their care cry in daily life. More than half of the respondents (about 55%) in
the non-OVC households witnessed that children in their respective homes never cry in
normal days while this figure stands at 47.3% in OVC households. Therefore, it appeared
evident that children in the OVC households more likely to cry than their counterpart
children in the non-OVC households
Similar proportion of household heads in the OVC and non-OVC categories (72.4% and
73.8% respectively) reported that children under their respective care never face difficulty
in making friends.
On rare
Occasion 65 10.0 36 10.4 50 17.9 56 20.3 10 8.3 17 5.2 16 13.0 250 12.4
Never 445 68.7 219 63.5 157 56.3 150 54.3 71 58.7 239 73.5 81 65.9 1362 67.8
Don't know 3 0.5 11 3.2 3 1.1 7 2.5 3 2.5 2 0.6 1 0.8 30 1.4
Total 2007 99.7
Mostly 28 4.9 23 8.7 24 11.3 13 6.3 19 15.3 15 3.8 6 5.3 128 6.8
Sometimes 36 6.3 21 8.0 24 11.3 16 7.8 12 9.7 35 9.0 8 7.1 160 8.5
On rare
Occasion 48 8.5 25 9.5 41 19.3 32 15.6 6 4.8 29 7.4 13 11.5 194 10.3
NON
OVC Never 440 77.6 169 64.3 121 57.1 130 63.4 85 68.5 286 73.3 83 73.5 1314 69.8
Househol Don't know 8 1.4 12 4.6 2 0.9 2 1.0 2 1.6 14 3.6 0 0 40 2.1
d No response 7 1.2 13 4.9 0 0 12 5.9 0 0 11 2.8 3 2.7 46 2.4
Total 1882 99.9
As illustrated in the last column of Table-6.10, the majority of the respondents, 69.8% and
67.8%, respectively for the OVC and non-OVC households reported that children are never
alone in their respective home. On the other hand, 11.8% of OVC and 10.3 % of non-OVC
house holds said that children tend to be alone some times while the proportion of the
respondents who said “on rare occasion” is 12.4% and 10.3% for OVC and non-OVC
Households in that order. Thus it can be inferred that slightly more of the respondents in the
OVC Household perceived that children often tend to be alone in their respective houses
than it was the case for the respondents in the non-OVC households.
In sum, the data obtained from the household survey with regard to the perceptions of the
Household heads on the psychosocial well being of children revealed few differences
between the OVC and non-OVC households.
As shown in table 7.1, 687 (35.4%) out of 1942 children under 5 represented in the
household survey in the survey were had birth certificates, 1238 (63.7%) did not, and 15
(0.8%) did not know.
Higher proportions of children under 5? with birth certificates were seen in Addis Ababa,
Dire Dawa and Bahir Dar with proportions ranging from 44 – 71%, while lower proportions
were seen in Shashemene, Assaita and Nazareth in the range of 10 - 20 %. OVC were less
likely (33%) than non-OVC (37%) to have a birth certificate, implying an OVC to non-
OVC ratio of 0.89.
The ratio between OVC to non-OVC birth registration status in each of the study towns
individually does not show any consistent pattern. Some agencies (MEASURE Evaluation
Project) suggest that it is not necessary to disaggregate the birth registration information by
OVC status, since birth registration is critical for all children. The registration of OVC
usually should take place long before a child is likely to have been orphaned. registration of
orphans could probably be higher in some settings as children are being registered in
response to having become an orphan.
The finding in general suggests that the issue of birth registration and certification is a
problem to both OVC and non-OVC as implied by the low overall average proportion of
children with birth certificates, i.e. 35.4%; while the ratio OVC to non-OVC doesn’t show
any clear pattern as to suggest any consistent disparity between OVC and non-OVC groups
as regards their access to birth certificates.
Table 7.1 Proportion of OVC and non-OVC who have Birth Certificates : PC3 Baseline
Survey in Seven Selected Towns of Ethiopia -2005.
Addis
Ababa Nazareth Shashemene Awassa Assaita Dire Dawa Bahir Dar Overall
n % n % n % n % n % n % n % n %
OVC
Valid N= 164 130 93 110 51 113 21 682
Frequency 82 50 25 19 11 12 36 33 7 14 50 44 15 71 226 33
NON OVC
Valid N= 231 185 208 145 115 326 50 1260
As depicted in table 7.1, 7.2 orphaned children are about twice as likely as non-orphans to
engage in income earning for their household.(8.8% vs. 4.6%). The ratio ranges from 1 in
Shashemene to 4.4 in Bahir Dar, and overall it stands at 1.9 or 90% more orphans as
compared to non orphans are engaged in income earning activities. Child labor exposes
children to accidents and denies their proper development in terms of participating in school
and play.
On the other hand, almost equal proportion of both orphan and non-orphan children are
assisting their households with chores. As per table 7.3 and the information we gathered
from a number of FGDS, males are more likely than females to work outside the home for
wages, while females are more likely than males to do chores at home.
Table 7.2 Engagement of children earning income for households and Household Chores
– OVC vs, Non-OVC: PC3 Baseline Survey in Seven Selected Towns of Ethiopia -
2005.
Table 7.3 Engagement of Children Age 10 and Over earning Income for the Household and
Assisting in Household Chores by Gender: PC3 Baseline Survey in Seven Selected Towns of
Ethiopia -2005
Bahir Dar
Dire Dawa
Assaita
Study areas
Total
Awassa
Non-OVC HHs with Succession plan
OVC HHs with Succession plan
Shashemene
Nazareth
Addis Ababa
Of those OVC households who reported having a succession plan in place to their children,
about one-half of them (48.6%), have reportedly shared the plan with potential guardians.
Of those non-OVC households, who reported to have a succession plan 46.3% of them have
affirmed that they have shared the plan with the potential guardian.
1. The total number of children included in this study was 9935. Of these 22% were
single or double orphans and 45% were OVC. The total number of households was
###. Of these, 32% accommodate one or more single or double orphans and 53% of
accommodate OVC. (Note that orphans are included among OVC and households
with orphans are included among households with OVC).
2. The age distribution of OVC Vs. non- OVC showed that the majority (57%) belongs
to age category of above 10, while non –OVC of this age group share only 47% of
their group.
3. About 48%of orphans are living with a surviving parent while 52% are living under
the custody of non-parent families. Female orphans are more likely than male
orphans to be living with non-relations or employed as helpers.
4. The vast majority of all children do not have birth registration. This is a critical area
of intervention for all children. OVC were just slightly less likely than non-OVC
(33% vs. 37%) to have had their birth registered.
5. The majority of all households are unable to provide some basic material needs
(defined as two shoes and two sets of clothing), with OVC households being less
likely yo be able to do so than non-OVC households (18.2% vs. 25.3%) Thus
families of OVC appear less capable of fulfilling basic material needs of their
children, but even a majority of non-orphan children are lacking these items.
6. OVC households were observed to be proportionally less food secure and more
often employing various forms of coping mechanisms such as decreasing the
amounts of food and skipping meals as compared to the non-OVC households. This
implies the need for intensive livelihood support to OVC households.
7. Adults in general and women in particular were subject to reducing the amount of
food or skipping their meals in times of household food shortage. This implies the
presence of further vulnerability of women in view of the prevailing high level of
maternal malnutrition.
8. Differences are small, but overall there is much need for succession planning.
Twenty percent of the OVC households have reported that they have a succession
plan for their children, while 23.7% of the non-OVC households reported to have
one. In both the OVC and non-OVC households three-fourth of the respondents have
reported that they do not have a succession plan. This implies that there is a need to
intensively and extensively promote the practice of succession planning.
10. The majority of the respondents failed to score adequate in three out of eight?
psychosocial areas, namely; Stress and Worry, Functioning and Capability and
Social Connection Capacity.
11. Fewer OVC than non-OVC had adequate scores in Stress and Worry (85% vs. 91%).
12. More than half of the respondents both in the OVC and non-OVC groups obtained
inadequate score in the measure of functioning and capability, showing that the
majority of the youth groups involved in the survey in general likely were in deficit
of life skills and competence.
13. Equivalent proportions of OVC than non-OVC respondents (74%) were found to
score adequate in functioning and social connection capacity sub-scales, but overall
more than a quarter of all youth scored inadequate in this area. It is possible that
ongoing OVC interventions and social support networks for OVC existing in the
community might be offsetting negative life events, have positive impacts in shaping
the personal functioning and social skills of OVC.
14. Remarkable disparities were observed in the proportion and ratio of the respondents
who scored adequate across the study town indicating that the upcoming
psychosocial interventions need to into account the socio-cultural variations of the
target areas and find out appropriate entry points.