‘We don't report because that's a family affair’:
A grounded look at community-based child
protection in three rural and urban areas in
Kenya
Inter-Agency Learning Initiative on
Community-Based Child Protection Mechanisms
and Child Protection Systems
April 23, 2014
The Inter-Agency Learning Initiative
 Widespread NGO and government use of community Child




Protection Committees or Child Welfare Committees
Questions about effectiveness, sustainability, relation to existing
community groups and wider child protection systems
Global review—weak evidence base, harmful practices, importance
of community ownership and linking community mechanisms
with formal aspects of national child protection systems
Multi-stage learning in Sierra Leone and Kenya
Bottom-up approach to systems strengthening—testing the
effectiveness of community driven interventions for strengthening
linkages between community based child protection mechanisms
and formal mechanisms
2
Global Reference Group
•
CPWG
•
•
Human Sciences Research Council
IICRD
•
Oak Foundation
•
•
•
•
•
•
•
•
Plan International
REPSSI
Save the Children
TPO Uganda
UNICEF
USAID (DCOF, PEPFAR)
War Child Holland
World Vision
3
Kenyan Partners
 UN/NGO: UNICEF-Kenya, Action Aid International,
AMREF, ANPPCAN, APHIA PLUS, CARE, CEFA, ChildFund,
Child Line, CLAN, CRADLE, Girl Child Network, IRC,
KAACR, Plan International, Save the Children, World Vision,
UNICEF/Kenya
 Government: Department of Children’s Services, Area
Advisory Councils (and LACs), District Children’s Officers,
Chiefs, Assistant Chiefs, and elders
 National Research Team: Ken Ondoro (Team Leader) with
Jemaiyo Chabeda-Barthe , multiple Kenyan researchers
 Youth, children, women, community people
4
DESIGN &
METHODOLOGY
5
Stages of the Research
 Qualitative/Ethnographic
 Definition of Well-Being Outcomes for Children & Construction,
Validation and Field Testing of Survey Instrument
 Action Research (with Quasi-Experimental Design)
- Baseline Measures (T1) of Children’s Risk and Well-Being
Outcomes—two matched sites
- Community selects, develops, and implements for two years a
linking intervention to strengthen child protection
- Endline (T2) Measures
 Interagency, multi-level consultation, participation, & learning
throughout
6
Research Areas and Sites
 Mombasa—two urban slums (informal settlement areas)
- Bangladesh (5,000 people)
- Tudor Moroto (3,000 people)
 Kilifi—two rural villages
- Bamba (900 people)
- Marafa (900 people)
 Kisii –two rural villages
- Kisii (4,000 people)
- Nyamira (4,400 people)
7
Ethnographic Phase:
Key Questions
 Who is a child and how do children develop?
 What are the main harms to children, aside from poverty




and health issues? How do they vary by gender, age, SES?
What endogenous and exogenous protection mechanisms
and practices are used?
What preventive factors support children?
When and how do people use the more formal child
protection system? What barriers are there to the use of the
formal aspects of the child protection system?
How do the more formal and more nonformal aspects of
the child protection system align?
8
Research Participants—Contrasts by Age,
Gender, SES, Positioning in Regard to Child
Protection









Women: Age 25 years and above
Young women: Age 18-25 years
Teenage girls: Age 13-17 years
Young girls: Age 5-12 years
Men: Age 30 years and above
Young men (typically not married): Age 18-30 years
Teenage boys: Age 13-17 years
Young boys: Age 5-12 years
Lower and higher SES (70% vs. 30%): based on criteria of
housing material, house size, location, source of income,
type of foods eaten
 Workers within the more formal child protection system
9
Mapping of Pathways of Response
 When problem X occurs, what happens—who responds




or who does child go to for help?
What are the steps involved?
Who makes the decisions and who is consulted?
What are the decisions taken or outcomes?
How do different stakeholders (parents, family,
community, child) view the outcome?
10
Research Team
 17 Kenyan researchers (9 males, 8 females)
 Team of 5-6 national researchers per area
 Understanding of local languages (Kiswahili & Sheng in
Mombasa; Mijikenda in Kilifi; Ekegusii in Kisii-Nyamira)
 Two weeks of participatory training
 Team Leader (Ken Ondoro)—checking, backstopping,
mentoring, analyzing
 Two international researchers (Drs. Kathleen Kostelny &
Mike Wessells)
11
Research Tools
(1) Participant observations (267) & living in villages
(2) Group discussions (recorded, 90 min.) of risks and pathways of
response (104 groups; 1,045 participants)
- identification of ‘most serious harms to children’
- mapping of typical pathways of response
(3) Group discussions (recorded, 60-90 min.) of preventive factors (57
groups; 500 participants)
(4) In-depth interviews (recorded): tailoring to people’s interests,
probing
(5) Timelines: childhood, roles, markers of transition
(6) Risk mappings (56 groups; 540 participants): 5-8 & 9-12 yrs.
(7) Body mapping (68 groups, 668 participants): learning from younger
children (5-8 and 9-12 years)
(8) Key informant interviews (recorded): Government (e.g., Children’s
Officers, AAC & LAC members, police, teachers), Chiefs, Assistant
Chiefs, elders, religious leaders, health workers, NGO workers(117
participants)
12
Ethics of Research
 KEMRI review and approval
 Save the Children’s Child Safeguarding Policy
 Respect for people’s dignity
 Use of general questions that avoided opening
personal wounds
 Informed consent and assent—no coercion, systematic
explanation of potential risks and benefits
 Confidentiality
 Expectation management
13
Data Collection & Work Plan
 Timeframe: July, 2012 to August, 2013
 Four weeks per site
 Flow of work: first week devoted to participant
observations , group discussions, & body mappings; in
the second and following weeks, individual methods
such as in-depth interviews became more prominent
 Digital recordings with verbatim English records
(identifiers removed), with checking for accuracy,
validity, etc.
 Follow-up by Team Leader
14
Analysis
 Grounded analysis by Kostelny, Ondoro, & Wessells, with






follow-up when needed
Emergence of natural categories
Triangulation by data source and method
Analysis of pathways of response/community-based child
protection mechanisms
Identification of typical narratives for various categories of
risk and response
Discussion and resolution of differences between researchers
Method of contrasts to track different views of subgroups
such as young women, young men, adult women & men
15
Limitations
 Short time frame
 Incomplete skill development
 Relatively short timeframe for analysis
 No measurement of incidence rates
 Not a national sample
16
Q&A
Key Finding 1
Children faced multiple
severe and interacting risks
30
24.1
Percentage of Participants
25
20
15.6
13.6
15
10.3
10
8.4
7.4
5.5
5
0
3.2
2.8
2.4
1.8
1.4
1.2
1.1
1.2
Local Views of Harms to Children
‘Most serious’ harms
Additional harms
 Out of school children
 Poor parenting
 Sexual abuse and
 Hard and heavy work
exploitation
 Early pregnancy
 Alcohol and drugs
 Lack of food/poverty
 Negative influences
 Orphans
 Child beating
 Witchcraft
 Child rights
20
Key Finding 2
Widespread poverty was a
driver of many of the harms
to children
 Key role in other harms:
 Out of school (not pay school fees)
 Sexual abuse
 Heavy labor (parents required children work)
 Alcohol abuse (poor people start chang’aa businesses;
bring children into it)
Out of School - Lack of School Fees
 Although the government announced that
education was free for all, there are still
small expenses that force parents to
choose between educating the child and
buying food for the family. (Key informant
interview, Tudor Moroto)
Sexual abuse: Prostitution
 Some of the parents can’t buy these girls things like
'Always' and some of them do not even have enough
food to eat. So when it becomes too much for these
children, they drop out of school and get into
prostitution in order to survive.... (Young woman, indepth interview, Marafa)
Early Pregnancy
 It is because fending for themselves. You get parents
not providing for their children, and girls set themselves
and start looking for men and after getting a baby is
when they realize the difficulty and pain of getting a
child without anything to give or care for them.
(Woman, in-depth interview, Nyamira)
Heavy Work
 Children from poor family background are doing very
hard work in well off families for pay. These children are
very young and are supposed to be in school but
because of poverty and they need to feed, they get
themselves doing these odd jobs to get money and buy
food. (Young women, group discussion, Kisii)
Key Finding 3
Gender related issues of sexual abuse
and exploitation, teenage pregnancy,
and early marriage were prominent
 Consensual sex with age mates
 Rape of young girls associated with brewing dens, disco funerals,
working as househelp
 Sexual abuse by uncles, teachers
 Incest
 Transactional sex common
- motorbike drivers
- older men cheated girls
 There are also other people who give children money
and every now and then, they keep on buying for them
“viazi” for ten shillings. Later on, they call these
children inside their houses and start touching them on
their private parts. (Adult man, in-depth interview,
Bangladesh)
 When the mother leaves and you are left alone with the
father, then father turns to you and rapes you. (Teenage
girls, group discussion, Tudor Moroto)
 'Watoto kuingia kwa biashara ya mwili' [children doing
the business of the body] is the main harm. Some
children, especially those in secondary schools, when
they don’t have money, they sell their bodies so that
they can get money. (Key informant interview,
Nyamira)
Sexual abuse by those with money
and power
 That’s why we don’t want tuition because teachers tell us to
stay in school up to eight o’clock in the night and then they
come at that time and start touching you on the thighs.
And some do bad things to girls at that time (Teenage girls,
group discussion, Bangladesh)
 When your mother dies and then you are taken to stay with
your uncle, the uncle will now be the person who buys you
everything…. Then after some time, the uncle comes to you
and tells you that you have to sleep with him…(Teen girls,
Tudor Moroto)
Sexual Abuse of Boys
 There is also the issue of young boys “kuwekwa na wamama
wazee.” (Being misused by the old women, including
sexually). They (young boys) brew chang’aa for the women
and they also perform “husband duties.”.(Key informant
interview, Bangladesh)
Early Pregnancies
 We have had so many cases where girls engage in sex
with men and later they get pregnant and drop out of
school…You know these are very young girls who don’t
even know how to cook and we don’t expect them to
become somebody’s wife… Old men who ought to be
fathers of these girls are the ones moving around with
the girls. They cheat them with money and after they
are pregnant they run away. (Young woman, in-depth
interview, Kisii)
Abortions
 There are so many fetus which are thrown in the ocean.
You see from my house, this is where the garbage starts
to be dumped so I always see so many fetus. (Woman,
in-depth interview, Tudor Moroto)
 Mostly we receive calls to go and pick children thrown
in the garbage and most of the time it is a mature fetus.
(Key informant interview, Tudor Moto)
Key Finding 4
The views of harms to
children varied according to
gender, SES and age
90
79.2
Percentage of Participants
80
66.7
70
60
51.5
50
40
28.6
25.5
30
23.4
20
10
0
Women
Young Women
Teenage Girls
Men
Young Men
Teenage Boys
Figure 2.The percentage of participants in urban slums in Mombasa by age and gender sub-group that rated sexual abuse and exploitation
as one of the top three harms to children (n=334).
Gender
- Teenage girls and women ranked sexual abuse as top
harm more often than did teenage boys and men
- Teenage boys and young men more often ranked
alcohol and drug use as harm than did teenage girls
and young women
SES
- Low SES were more likely than high SES families to
rank out of school children as harm
- High SES families were more likely than low SES
families to rank early pregnancy as harm
- High SES families were more likely to rank alcohol
and drug use as harm
Age
Young children were concerned about:
 heavy work
 parents fighting
 being called abusive names
 being beaten and caned
Key Finding 5:
HIV and AIDS and related
problems were significant
sources of vulnerability for
children
HIV and AIDS cont.
- HIV and AIDs were widespread in all three areas
- boys and girls were sexually active at early age
- children born to HIV positive mothers often became
HIV positive
- HIV spread by transactional sex
- some people spread HIV deliberately so they would
not die alone
Maltreatment of orphans affected
by HIV and AIDs
 The villagers look at this [HIV positive child] as if he is
not a person. The disease brings some hatred. The
people form opinions towards the children. The
orphans -- especially those who are known to have
been born with the disease -- are hated and are told
'your parents died of HIV'. (Woman, group discussion,
Kisii)
Q&A
Preventive Factors and Response
Pathways
Preventive Factors
 Valuable preventive factors at family and community
levels were operative yet had limited capacity,
effectiveness, and reach.
Preventive Factors
 Family (grandmothers, parental guidance)
 Education, Discipline and Economic support
 Religious groups and leaders
 Preached good values and morality to the youth
 Preached about good parenting
 Counseled the youth
 Raised money for children’s fees
 Helped orphans
Preventive Factors
 Youth group: Alpha and Omega, Young mothers,
Amkeni Youth Group
 Educated the youth on safe sex
 Distributed condoms
 Educated youth on dangers of early pregnancy
 Women’s groups: ‘Merry Go Round’
 Was a good source of money to help keep children in
school
 Preventive factors were strained or overwhelmed by
the burden of accumulating risks
Response Pathways
 For 70-80% of harms to children, the response was
through non-formal family and community
mechanisms
 Pathways of response were particularly weak in regard
to sexual abuse within the family
 Family honor and avoidance of shame were often
prioritized over the well-being of girl survivors of
sexual abuse
 ‘Traditional’ mechanisms such as traditional courts or
councils of elders were rare to nonexistent
Non-formal Mechanisms
 Nonformal family and community mechanisms were
of central importance in responding to risks to
children yet were overstretched and variable in their
effectiveness
 Some nonformal family mechanisms caused harm to
children
 Abortion
 Early and forced marriage
 Family agreement in cases of rape
 Corporal punishment
Mombasa: Pathways of response for out of school children who had
been unable to pay school fees.
Child is out of school because school
fees had not been paid
Mother borrows or
obtains money from
relative, pays fees
Child tells mother or
mother notices
Teacher accepts
and child goes
back to school
Mother talks with
teacher and
promises to pay
Mother gets money
from Merry Go
Round
Teacher does
not accept
Mother goes to
Head Teacher,
pleads case
Head Teacher
allows child back
into school
Mother looks for
more work if no
father
Mother tells
the father
Head Teacher
does not allow
child back into
school
While out of school, child makes money and
decides not to return or falls in with bad peers
and does not return to school; or a girl out of
school becomes pregnant and does not return
Child returns to
school
Child returns to
school
Father looks
for casual
labor
Child out of school until
money is raised
Child returns to
school
Q&A
Linkage of CBCPMs
With the Formal System
 Importance of linkages: referrals, prevention,




authority, capacity, etc.
Diverse connectors: Chiefs, Assistant Chiefs, elders, police,
DCOs, VCOs, AACs, LACs
These connectors did in fact respond to harms to children,
encourage prevention, and make appropriate referrals.
Area Advisory Councils (AACs) and LACs emerged as
potentially useful in bridging formal and nonformal child
protection supports, yet most were in the planning stages
regarding child protection.
Overall, there was mixed evidence regarding the use of
linkages with the formal system – local social norms frequently
pulled in a contrary direction.
Alignment Between Formal and
Nonformal Elements of the Child
Protection System
 Alignment is important for a coordinated, consistent
approach.
 Overall, there was a moderate level of alignment.
- Views of harms to children (out of school, teenage
pregnancy, drugs and alcohol, sexual abuse, poor
parenting, etc.).
- Some consistency of response by formal and non-formal
actors to those harms.
 But, social norms frequently ran counter to the laws and
policies of the formal system.
Challenges of Non-Alignment between
Nonformal and Formal Elements:
Who is a Child?
 Chronological definition: Kenyan national law
 Local views: varied ages, emphasis on dependency,
limited responsibility, small stature, etc.
A child is one who cannot determine what is right or wrong.(Teenage boy,
group discussion, Nyamira)
A child is the one who is very small ... when you see their body, the way
they appear, you definitely know that this is a child. (Man, group
discussion, Marafa)
A child is a child until he is an adult when he/she is married and has his
own people/ kids. (Women, group discussion, Nyamira)
Challenges of Non-Alignment:
Harms to Children
 Kenyan law prohibits early marriage, female circumcision,
and sexual exploitation of children, yet each of these was a
customary practice in the research sites.
 In Kisii, workers in the formal system viewed local culture
as a harm to children, though other citizens did not.
Kisii culture is a harm to children. When the father passes away and
leaves the wife and children behind, most of the times you find that
children are sent away from home by their father’s relatives. This really
affects children because some of them go to the streets and become
street children and completely stop going to school. Fathers also defile
their children and this has become a problem around here. (Key
informant interview, Kisii)
Inconsistency within the Formal
System—Child Rights
 In Kilifi, some workers in the formal system
identified child rights ('haki za watoto' ) as a harm
to children.
But I will say again that this 'haki za watoto' have
contributed immensely to this issue of early pregnancies.
Nowadays these children don’t fear getting pregnant, they
know very well that they will give birth and go back to
school and there’s nothing that parents can do to them.
(Key informant, Bamba)
Challenges of Non-Alignment:
Child Protection Practices
 Kenyan law forbids corporal punishment of children, yet
parents saw child beating as essential for teaching children
obedience and good behavior.
 Actors within the formal system (Chiefs, teachers, elders)
frequently used corporal punishment on children.
 Approximately one quarter of the participants were not
willing to report a statutory crime such as the rape of a
child to the authorities.
 People reportedly took direct collective action themselves
against strangers who were caught raping a child.
Underuse of the Formal System:
Four Main Obstacles
(1) Bribery and corruption: Chiefs or Assistant Chiefs ask
for ‘something small’; perpetrators of statutory
crimes reportedly paid off police to drop the case
against them, or they paid money to the victim’s
family, who then refused to press charges or testify.
(2) Preference for family and community solutions:
Teenage pregnancy: family first, and preference for
reaching a ‘compromise’ between families; for
children who were out of school due to inability to
pay school fees, the main response was often through
religious groups or the village ‘merry go round.’
Underuse of the Formal System:
Four Main Obstacles, cont.
(3) Taboo against reporting family abuses and problems
outside the family—honor, stigma, lost support.
 Because that is a family affair and the uncle can be called and asked why he
has done that. Then he can be told to stop it…. That’s a family issue. (Adult
women, Bangladesh)
 'Titogweitera' [We cannot kill for ourselves]. What we do is that we call a
meeting and see how to solve the issue before it is known because if it is
known, it will spoil the name of the family. The uncle will be asked to pay
some fine, maybe an animal. (Young man, group discussion, Nyamira)
 The rapist might be the bread winner in the family. If he gets arrested,
people will suffer in the family. I might end up dropping out of school
because there will be no one to pay my fees. In fact, that is family thing that
people outside the family should not even know. It should be kept a secret.
(Young men, Bangladesh)
Underuse of the Formal System:
Four Main Obstacles, cont.
(4) Perception that reporting violations to authorities
leads to inaction.
- Citizens and workers in the formal system reported
that the courts were backlogged.
- Chiefs described how upward referrals of cases
sometimes led to inaction or the case being handed
back down to them.
Q&A
Action Priorities






Enable a coordinated, holistic approach
Integrate economic support and child protection
Address gender issues
Place families and communities at the center
Strengthen prevention
Improve the use, functionality, and accountability of
the formal child protection system
 Support internally driven social change
 Stakeholders in Kenya develop specific, contextualized
recommendations
Community Child Protection Exchange
www.childprotectionforum.org
Kostelny, K., Wessells, M., Chabeda-Barthe, J., & Ondoro, K. (2013). Learning about
children in urban slums: A rapid ethnographic study in two urban slums in Mombasa
of community-based child protection mechanisms and their linkage with the Kenyan
national child protection system. London: Interagency Learning Initiative on
Community-Based Child Protection Mechanisms and Child Protection Systems.
Kostelny, K., Wessells, M., & Ondoro, K. (2014a).Community-based child protection
mechanisms in Kilifi, Kenya: A rapid ethnographic study in two rural areas. London:
Interagency Learning Initiative on Community-Based Child Protection Mechanisms
and Child Protection Systems.
Kostelny, K., Wessells, M., & Ondoro, K. (2014b).Community-based child protection
mechanisms in Kisii and Nyamira: A rapid ethnographic study in two rural areas of
Kenya. London: Interagency Learning Initiative on Community-Based Child
Protection Mechanisms and Child Protection Systems.
Wessells, M., Kostelny, K., & Ondoro, K. (forthcoming). A grounded view of
community-based child protection mechanisms and their linkages with the wider child
protection system in three rural and urban areas in Kenya: Summary and Integrated
Analysis.