Content uploaded by Kamau Wairuri
Author content
All content in this area was uploaded by Kamau Wairuri on May 21, 2020
Content may be subject to copyright.
DIGNITY Publication Series on
Torture and Organised Violence
Study
#
16
URBAN VIOLENCE IN NAKURU
COUNTY, KENYA
Kamau Wairuri
Ahlam Chemlali
Mutuma Ruteree
2 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 3
URBAN VIOLENCE IN NAKURU COUNTY, KENYAA
Study on urban violence prepared
by CHRIPS - The Centre for Human Rights and Policy Studies (CHRIPS)
DIGNITY Publication Series on Torture and Organised Violence No. 16
© 2018 DIGNITY - Danish Institute Against Torture
Midrift Human Rights Network
https://midrifthurinet.org/
CHRIPS - The Centre for Human Rights and Policy Studies
https://www.chrips.or.ke/
Danida
http://um.dk/da/danida/
DIGNITY – Danish Institute Against Torture
www.dignityinstitute.org
ISBN Print: 978-87-90878-87-0
ISBN Online: 978-87-90878-88-7
2 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 3
Contents
1. Executive Summary 9
2. Introduction 14
2.1 Causal Factors 16
2.2 Forms of Violence 18
2.3 Vulnerability and Victimization 22
2.4 Mitigation 23
Methodology 26
2.6 Overview 26
2.7 Study Sites 26
2.8 Quantitative (Household) Survey 26
2.9 Qualitative Data 29
2.10 Limitations and Mitigation Measures: 30
Key Findings and Analysis 32
2.12 Overview 32
2.17 Prevalence and Dynamics of Violence in Nakuru County 40
2.17.1 Sexual and Gender-Based Violence 40
2.17.2 Violence against children 41
2.17.3 Violent Crime 43
2.17.4 Police violence 45
2.17.5 Political and ethnic violence 46
4 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 5
3. Conclusion and recommendations 57
4. References 64
4 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 5
URBAN VIOLENCE IN NAKURU
COUNTY, KENYA
A STUDY ON ROOT CAUSES, RISK FACTORS AND PREVENTIVE STRATEGIES
Study on urban violence prepared by CHRIPS in collaboration with Midrift and DIGNITY
By Kamau Wairuri, Ahlam Chemlali, Mutuma Ruteree
6 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 7
Acknowledgments
This report was commissioned by DIGNITY - Danish Institute Against Torture together with Mid-
Rift Human Rights Network and undertaken by the Centre for Human Rights and Policy Studies
(CHRIPS).
The CHRIPS team would like to acknowledge the support of DIGNITY sta, in particular, Ahlam
Chemlali programme manager, Ane Kirstine Viller Hansen, health advisor and Anne Timm for
their leadership, guidance, support and hard work throughout all the phases of study.
We also warmly thank the sta of Mid-Rift Human Rights Network, Joseph Omondi, Executive
Director, Leonard Githae, Deputy Executive Director, and Walter Mwania, Programme Manager,
for the exceptional insights, advise and coordination support they provided during field research
for the study.
We are especially indebted to all the people in Nakuru County, who participated in the key
informant interviews, focus group discussions and the household survey for this study. This
study would not have been possible without their support.
6 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 7
8 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 9
1. Executive Summary
Rapid urbanisation has led to an increase in the prevalence of urban violence in many developing
countries. This is because of the mushrooming of densely populated informal settlements in
cities, which are characterised by deprivation and low quality of basic social services such as
healthcare and education. The situation is aggravated by the lack of jobs for most residents
of these settlements, who experience inequality, marginalisation and exclusion. Such an
environment facilitates the emergence and increase in violence in urban centres.
In Kenya, about 25.6 % of the population is urbanised, most of whom live in large cities such
as Nairobi, Mombasa, Kisumu, Eldoret and Nakuru, of which 56 % live in informal settlements.
Violence in these areas has emerged as a serious security and public health challenge which
the Kenyan security agencies have largely been unable to address. Violence has been shown to
have significant negative consequences including erosion of social cohesion, trauma, broken
families, injuries, deaths and loss of property.
The present study, commissioned by DIGNITY, is the first of its kind as it focusses on urban
violence in Nakuru County. It provides new knowledge that will inform the design of a three-
year multi-stakeholder urban violence program in selected areas of the county. The study was
conducted in five sites across Nakuru County: Bondeni and Kaptembwo, Nakuru Town; Karagita
and Kabati, Naivasha Town and Molo Town, where urban violence was previously understudied
and, therefore, not well understood. The methodology adopted involved primary research
using key informant interviews, focus group discussions and a household survey, with 43 % of
respondents from Nakuru Town Municipality, 38 % from Naivasha and 19 % from Molo Town.
Most of the respondents were women and young people, of which 69 % were below 34 years
and only 10 % above 55 years old. These were complemented by secondary material.
The findings revealed that violence is a major concern for residents of Nakuru County, most
of whom felt that violence was becoming more prevalent. The data collected highlights that
the most prevalent forms of violence in the County mirrored the patterns of violence in Kenya.
Those highlighted in Nakuru include sexual and gender-based violence (SGBV), violence against
children, violent crime, police violence and political and ethnic violence, which are distributed
among the various neighbourhoods. Notably, SGBV and violence against children were
highlighted as particularly significant challenges, by more than 70 % and 66 % of respondents,
respectively, who noted that these forms of violence mainly occur at the household level. The
study highlights that these forms of violence introduce unique challenges to intervention such
as when the victims do not want the perpetrator punished for fear of broader repercussions.
Violent crime is a significant problem, particularly in Bondeni Area in Nakuru, where 70 % of
respondents identified it as most prevalent.
Given the prevalence of SGBV and violence against children, it is unsurprising that most
respondents identified the bulk of victims as women, girls and children generally. Undoubtedly,
there were few noted cases where men were victimised by their wives as well, especially in
Naivasha. Many respondents also indicated the youth as victims of violence, especially with
respect to violent crime and police violence. Unsurprisingly, many respondents also identified
young people as the major perpetrators of violence, sexual assault and violent crime. This could
be related to the presence of gangs perpetrating violence in the County.
8 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 9
On the causes, of violence, about 80 % of respondents highlighted the underlying causes as
unemployment and idleness among the youth, poverty and 60 % attributed it to drugs, especially
alcohol abuse. This study however adopts a formulation developed by renowned scholars
Caroline Moser and Cathy McIlwane (2006) for understanding urban violence. They argue that
urban violence is a complex phenomenon that should be viewed through three lenses: structure,
identity and agency. As such, we argue that these underlying factors do not in themselves cause
violence but rather interact with other social and individual dynamics and triggers to precipitate
violence. This is also why the WHO and CDC, four-level social-ecological model (SEM) was
adopted, as part of the design of the household questionnaire and in the data analyses. The SEM
allows us to address the factors that put people at risk for or protect them from experiencing or
perpetrating violence (risk and protective factors) and the prevention strategies that can be used
at each level to address these factors.
The study notes that there are several interventions and coping mechanisms at the individual,
community and state levels. We noted individual behaviour changes to limit exposure to violence,
such as not walking after dark or bodaboda operators accompanying each other while taking
customers to neighbourhoods perceived to be more insecure. At the community level, we noted
eorts by civil society organisations, community-based organisations and religious institutions
to address dierent types of urban violence through peace initiatives, awareness creation, setting
up safe houses for victims of SGBV and the use of technology to alert authorities. At the State
level, the respondents noted the Nyumba Kumi initiative, a form of community policing that has
been applied across all areas under study, and other interventions by the County Government
such as setting up of SGBV unit at hospitals.
The study however noted gaps in these interventions in their approaches and/or coverage,
which presents an opportunity for additional programmatic interventions. We note, however,
that the success of any intervention will depend on the extent to which it relies on community
level structures that are seen as being trustworthy and eective in addressing these challenges
of crime and violence. Notably, any intervention would be well advised to include religious
leaders, community elders and local chiefs. About 70 % of respondents in this study found these
local level institutions as both trustworthy and eective in dealing with violence.
We conclude by proposing recommendations based on the data with a focus on prevention of
violence with limited attention paid to interventions after violence has occurred. Our proposed
interventions are largely based on how to prevent violence by improving the awareness and
understanding of the dynamics of violence and how they should be handled amongst the
community and hence empowering the community to deal with violence. The view is on
empowering the community to use existing mechanisms to address violence, with focus on
building on the community trust in the existing institutions such as community organisations,
police, chiefs, elders and religious leaders.
10 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 11
Recommendations:
SGBV (Sexual and Gender-Based violence)
1. Supporting the expansion of existing awareness creation and empowerment programs
beyond schools to the entire community, especially women and girls to understand
the dynamics of violence and how to respond in case of victimisation but also to help
men and boys understand what constitutes SGBV
2. Improving the incorporating of the most trusted actors in the communities in the
eorts of creating awareness and empowerment related to SGBV. In this regard,
programs to prevent SGBV cases should include doctors, religious leaders and the
police.
3. Support further dissemination of information on the legal procedures relating to sexual
violence including the preservation of evidence to facilitate arrest and successful
prosecution. The trusted institutions within the community such as religious leaders
and elders should be equipped with this information as they could be first points of
contact of victims or witnesses of violence.
Violence against children
1. Support programs carried out by trusted local institutions such as religious leaders
that raise awareness about the rights of children and the negative eects of violence
against children. Such programs should also include education on parenting skills and
other methods of disciplining children. Bringing doctors to speak to parent groups
about the long-term eects of violence against children could strengthen such
programs.
2. Develop school-based programs to empower children to be able to report on
cases of violence and abuse occurring to them or to their friends at home, in the
neighbourhood or in school including how to protect themselves from victimisation
3. Provide support to existing child support centres and homes for victims of this type
of violence or setting up new ones in partnership with the communities to deal with
the most serious cases.
10 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 11
Violent crime
1. Strengthening of existing community level interventions of dealing with violent crime
such as the Nyumba Kumi and Community Policing initiatives. This requires a more
fine-grained analysis of why they work better in some places and not others in order
to replicate best practices across the county.
2. Support the scaling of the Mulika Uhalifu program, a mobile phone program, which
enables individuals to report cases of violence by sending a simple message to a given
number that is linked to various Police station heads, or such similar initiatives across
the county to help in the reporting of violent crime.
3. Lobby the government to enhance the existing programs of dealing with proliferation
of small arms and light weapons in the county.
4. Work with the existing state mechanisms such as IPOA (Independent Policing
Oversight Authority) to enhance police accountability.
5. Support the National Police Service, and other mechanisms, at the local level to
enhance patrols, investigate reported cases, arrest and prosecute perpetrators.
6. Target out of school youths in peace and security programming.
7. Address the high rate of youth unemployment through provision of life skills, linkages
to access of government youth funds and the 30% tender provision by government.
8. Chiefs and police should crackdown on suppliers and brewers of illicit alcohol and
supply drugs.
9. Demobilize violent gangs and provide them with alternative life skills to deter re-
oending.
10. Promote citizen understanding and confidence of criminal justice system to improve
citizen willingness to stand-in as witnesses and pursue cases in the courts.
11. Encourage police ocers, directorate of public prosecutions and the courts to work
in tandem in expediting cases before them in time.
Police violence
1. Support the existing programs by local level NGOs to create awareness about the
rights of citizens when dealing with the police in order to increase demand for fair
treatment when engaging with the Police.
2. Work with state level institutions responsible for creating police accountability including
IPOA (Independent Policing Oversight Authority), NCAJ (National Commission on
Administrative Justice), KNHRC (Kenya National Human Rights Commission) to help
the residents better understand their roles and how to seek recourse in case of abuse
by police ocers.
12 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 13
3. Support local NGOs to fight the impunity of police ocers by enabling them to bring
cases against rogue police ocers so that they can face the consequences of their
actions.
4. Support community partnerships with the police where such issues can be raised and
addressed.
Political and ethnic violence
1. Supporting the existing mechanisms developed by local NGOs and Faith Based
Organisations to enhance inter-ethnic cohesion.
2. Lobby government institutions such as the NCIC (National Cohesion and Integration
Commission) and NPC (National Peace Committees) to have a broader and deeper
presence in Nakuru County.
3. Support and encourage the County Government of Nakuru to extend the Barazas
on understanding and tolerance beyond Nakuru Municipality to other areas of the
County.
4. Work with the County Government to develop inclusion policies that promote equality
of opportunity for all citizens in the County, regardless of ethnic background.
5. Support national level eorts, executed at the county level to address historical
injustice in the most inclusive manner. Most importantly, the resettlement of IDPs
displaced by politicised political and ethnic violence in previous elections.
6. Encourage community members to report cases of incitement by political leaders and
support the prosecution of such leaders in the community.
All the above-mentioned types of violence, can be addressed by focusing work and
interventions towards the following:
1. Participation and Norm change on how to prevent violence perpetration from
happening at all, and reducing risk factors for people becoming perpetrators, by
mobilizing participation and challenging norms.
2. Trust-building between police and community, as recommended by the World Bank.
3. Leadership development for intersectoral coalition building should be strengthened.
4. Forward looking we see a need for more studies and research into data on perpetrators
as opposed to victims and victim-based data.
12 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 13
2. Introduction
Over the last 60 years, developing countries have witnessed increasingly rapid urbanisation. By
2007, more than half of the world’s population (54 %) lived in urban areas compared to 30 %
world urban population in the 1950s (World Bank 2014, p.8). Trends indicate that nearly two-
thirds of the world’s population will live in urban areas by 2050 (World Bank 2014, p.8).
This rapid urbanisation has exerted enormous pressures on the eorts and resources of
developing countries, making it dicult for them to provide decent standards of living and
access to quality public services for most the urban residents. The inability of national and local
governments to build adequate infrastructure and housing facilities has led to the mushrooming
of unplanned, informal settlements in many cities and towns. These areas are characterised
by deprivation with low quality of basic social services such as healthcare and education,
which in any case, are inadequate for the large population found in these settlements. These
challenges expand to other areas such as security, water and sanitation and energy. The social
challenges of these neighbourhoods are aggravated by the unavailability of jobs for most of
the residents of these neighbourhoods despite having higher educational attainment compared
to rural populations. Most of the residents of these neighbourhoods are unemployed or make
meagre incomes in self-employment. In every sense, residents of these poor neighbourhoods
experience inequality, marginalisation and exclusion.
This situation is persisted around the country with about 56 % of the urbanised population
live in the informal settlements in large cities and towns including Nairobi, Mombasa, Kisumu,
Eldoret, and Nakuru.1 Like their counterparts elsewhere in the world, they live in overcrowded
conditions, lack of basic public services such as piped water or reliable electricity supply and
high unemployment levels and poverty rates.
Violence is an important development issue that requires attention.2 It is now widely recognised
that violence adversely aects a country’s macroeconomic growth and productivity.3 Analysts
have observed that violence undermines the development of nations, both in terms of economic
growth and investments and in relation to sustainable development and empowerment.4
Beyond these challenges, or perhaps because of them, violence in urban areas has emerged
as a serious security and public health challenge, which the Kenyan security agencies have
largely been unable to address. This situation disproportionately aects the residents of informal
settlements.5 Violence and violent crime have significant negative consequences including
erosion of social cohesion, trauma, broken families, injuries and deaths and loss of property.6 Even
1 World Bank, 2014.
2 Mcllwane, 1999.
3 Moser, C. & Mcllwane, C., 2006.
4 Mcllwane, 1999.
5 In fact, national polls show that violent crime/violence is one of the greatest social problems facing
citize ns.
6 Hillier 2007
14 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 15
though Kenyan security agencies have attempted to address these challenges, they have been
largely unsuccessful largely because their strategies and activities are not targeted at the overall
phenomenon of urban violence but rather on specific sub-categories of violence, particularly
political violence and gender-based violence. Thus, some aspects of violence, including social
and economic violence and institutional violence are neglected. Conceivably, approaching the
challenge holistically, that is seeking to understand the phenomenon of urban violence rather
than focusing on specific forms of violence is a starting point to address these issues.
The partnership on safety and security work between DIGNITY and the Nakuru-based Mid-Rift
Human Rights Network, on intersectoral urban violence prevention, is one of the few exceptions
to this approach. The two organisations have cooperated to implement activities aimed at
improving policing within Nakuru Town Municipality, especially in relation to prevention of
urban violence. With a view to build on this work more systematically, DIGNITY, the lead partner,
commissioned a baseline study on urban violence in Nakuru County in January 2017. This
study, the first of its kind to be undertaken in Nakuru, provides new knowledge and information
that will inform the design of a three-year, multi-stakeholder urban violence programme in
selected areas in the county. Towards this end, the study sought to illuminate the distribution
and prevalence of violence, the dynamics of violence in relation to social capital issues such as
socio-economics, gender, and social inclusion/exclusion, patterns of reporting and prevention
of crime and violence.
The study focused on identifying the:
• Root causes, mechanisms and dynamics, general risk factors and protective factors and
settings associated with urban violence.
• Vulnerable groups (at risk groups and victim categories)
• Factors that increase/decrease the probability of becoming a victim of violence.
• Factors that increase or decrease the probability of becoming a perpetrator of violence.
The study sites were Bondeni and Kaptembwo informal settlements in Nakuru Town Municipality;
Kabati and Karagita informal settlements in Naivasha Sub-County. These are also the sites where
the three-year programme will be implemented. Molo Town in Molo Sub-County was chosen
as a control site.
14 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 15
Understanding Urban Violence
Years of examination of the concept of violence, especially within the fields of criminology
and geography, have demonstrated how complex the phenomenon is. Even though many
theories have been advanced, none has yet to attain hegemonic status to provide researchers
with a frame to explain urban violence. Nonetheless, the theories that have been advanced give
us useful tools, which help us to understand some key aspects of this phenomenon. For the
purposes of this analysis, we will consider extant literature on four main themes: (1) underlying
causal factors, (2) the dierent forms of violence, (3) vulnerability and victimisation and (4)
interventions for dealing with violence.
2.1 Causal Factors
It is evident that urban violence is highly concentrated in poor neighbourhoods. In urban
contexts, a particularly important debate concerns the extent to which crime and violence are
causally rooted in inequality and exclusion. Much of the research on crime and violence in
developing contexts has demonstrated that the relationship between poverty and inequalities
with urban crime/violence is complex.7 This is not to disregard poverty as a contributor to high
levels of violence but rather to point out that there is no clear evidence of a causal relationship
between the two variables thought to contribute to urban crime/violence. Stewart for instance,
argues that violence, poverty and inequality are linked in a vicious cycle: inequality spawns
violence, which in turn worsens poverty and increases inequality even more.8 While this may
be so, researchers and academics still debate on the extent to which the factors are linked
with urban violence.9 Nonetheless, inequality is a major contributor to the high levels of crime
and violence in poor urban neighbourhoods. Winton argues that “...in situations of widespread
and severe inequality, the urban poor are undervalued and marginalized, and their daily living
conditions heighten the potential for the emergence of conflict, crime or violence.” 10
Inequality takes both economic (poverty) and socio-political dimensions11 with the latter being
a question of social and political exclusion which is often but not always, based on identity.
This nexus between exclusion, inequality and identity is useful in explaining high levels of inter-
communal violence in poor-urban neighbourhoods. This is the argument made by renowned
scholars of urban violence Caroline Moser and Cathy McIlwane noting that power and
powerlessness are fundamental to understanding the causal factors that underpin violence and
challenging the stereotypical view of poverty as a determinant of violence as too simplistic.12
It is important to distinguish between structural causes and trigger risk factors. While underlying
structural causes are generally related to unequal power relations, trigger risk factors, in contrast,
relate to situational circumstances that can exacerbate the likelihood of violence occurring.
This approach therefore helps to incorporate the wider political and socioeconomic power
structures within which individual realities are manifest into the analysis of urban violence while
7 Lemanski, 2012.
8 Stewart 2008
9 Moser 2004; Fainzylber et al. 2002; Neumayer 2005.
10 Winton 2004
11 Muggah 2012: 45
12 Moser, C & McIlwane, C.,2006.
16 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 17
also allowing for the recognition that victimisation is also influenced by elements of individual
identity formation as gender, age, ethnicity and race.
Moser and McIlwane (2006) developed one of the most potent frameworks for understanding
this phenomenon. Their work, builds on community perceptions of urban violence to locate the
situation-specific nature of people’s experience of violence within a broader structural context.
They convincingly argue that no single factor explains why some individuals behave violently
towards others or why some communities are more violent than others.
Noting that the causal factors underlying violence are multi-dimensional and interrelated, they
map the factors underlying violence, fear and insecurity on to three interrelated concepts of
structure, identity and agency.
Violence may be embedded in social institutions and material structures through the severe
inequalities in the distribution of economic, political and social resources that are closely linked
to poverty and inequality. Slums or poor urban neighbourhoods are spaces where multiple forms
of deprivation accumulate.13 Dark lanes, isolated bus stops or public latrines are often unsafe
spaces where rape, robbery and violent crime lurk. Others have noted how the life demands on
the poor, requiring them to commute long distances, and to work early in the morning or late at
night, exacerbate these spatial dangers. Another dimension regards socially constructed levels
of tolerance to violence, and perceptions of what are acceptable and unacceptable levels or
types of violent manifestations. Understanding how a society responds, or fails to respond, to
dierent types of violence is a vital component of understanding violence and crime.
Regarding identity, Moser and McIlwane argue that how people experience violence also
depends on their identity position primarily with respect to gender, age, ethnicity and race.
They call for a more nuanced approach to this question of identity noting that individuals have
a plurality of identities. They also highlight the need to factor in individual agency in any attempt
to understand crime and violence noting that individuals as social actors react to situations and
formulate objectives in dierent ways. They note that stereotypes, which treat groups of people
as “objects” denies them individual agency and their roles as actors. They specifically note the
common stereotypes that link poverty with high crime rates suggesting that all poor people are
violent. They conclude that the combination of structure, identity and agency that can assist in
understanding the underlying factors causing violence.
13 Auyero, et al 2013
16 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 17
2.2 Forms of Violence
Moser and McIlwane (2006) propose a conceptual framework that makes a four-fold distinction
between political, institutional, economic and social violence – with each category identified
in terms of the motivation for the physical act that consciously or unconsciously is used to
gain or maintain power. However, they themselves note that such categorization is too static
to represent a dynamic and holistic phenomenon, the four-fold typology identified above is
conceived as an interrelated continuum with close linkages between dierent types of violence.
Notably though, public records on violence and crime are not represented in this format.
Violence is prevalent in Kenya. Nearly 90 % of all adults in Kenya have experienced physical
violence since they turned 15 years old.14 The most prevalent form of violence, as shown by
Table 1 below, are Gender-Based Violence, Sexual Violence, Violence against Children, Murder,
Robbery, mob violence and extra-judicial executions by the police.
Sexual and Gender-Based Violence is quite common with women and girls, mainly, being
victims of assault, rape, threats and acts of intimidation.15 It is estimated that 39 % of married
women and nine % of men between the ages of 15 and 49 have experienced spousal violence or
sexual violence with many of them suering physical injuries.16 About 20 % and 12 % of women
and men, respectively, reported experiencing violence in the year before the 2015 survey was
undertaken. Further, it is estimated that 14 % of women and 6 % of men in the entire Kenyan
population, have been victims of sexual violence at least once in their lifetime. In 2016, 22,732
cases of sexual violence were reported to the police.17 Another dimension of this type of violence
is familicide in which parents, mainly men kill their spouses, children and then themselves.18
Table 1: Statistics on Violence Prevention in Kenya
Incidents
Female Male
Gender-Based Violence - 39 9
Sexual Violence 22, 732214 6
Violence against Children - 32 18
Murder 2,6483- -
Robbery 2,8654- -
Mob Violence 1,5005- -
Extra judicial executions 2626- -
14 KDHS , 2014, p. 291.
15 KDHS, 2014; C. McEvoy, 2012.
16 KDHS, 2015, p. 291.
17 KNBS , 2016, p.52.
18 In some cases, women kill their children and then themselves.
18 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 19
Violence against children is another predominant form of violence in Kenya even though data
on the victimisation of children is outdated. The most recent survey on violence against children
was undertaken in 2010.19 It showed that 32% of girls and 18% of boys in Kenya experienced
physical violence during childhood with parents being the main perpetrators of violence against
children.20 However, teachers and fellow students within the school environment have been
noted to also be key perpetrators of this violence. For instance, while corporal punishment in
schools was banned in 2001, it persists in some areas with school bullying being a significant
problem that has led to deaths and the maiming of children in some Kenyan schools.21
Major crimes such as murders, robberies and kidnappings constitute the other key category of
violence in Kenya. In 2015, the murder rate in Kenya that year stood at 5.63 murders per 100,000
people.22 Media reports show that robberies in Kenya mainly take the forms of muggings,
home invasions, car hijackings, and armed attacks on businesses. Robbery incidences, in which
robbers attack their victims with blunt objects, knives and guns, have ended up in deaths and
maiming of many more.
Table 2: Reported Cases of Murder and Robbery in Kenya (Source KNBS
Statistical Abstract 2016)
Violence 2010 2011 2012 2013 2014 2015
Murder 2,239 2,641 2,761 2,878 2,649 2,648
Robbery 2,843 3,262 3,262 3,551 3,011 2,865
In 2015, the rate of robbery was 6.09 robberies per 100,000 people.23 A 2010 UNODC survey
showed that at least 3.7% and 5.1% of Kenyans reported being robbed or assaulted/threatened
in 2010.24 This discrepancy between the murder rates based on ocial records, and this survey
data, seems to suggest robbery incidents are generally not reported to the police in Kenya.25
Kidnapping, especially of children, and car hijackings are also quite prominent., reported that
there were 588 cases of carjacking in 2012 which dropped to 470 in 2014 with kidnappings
19 U NICEF, 2010.
20 UNICEF, 2010, p. 2.
21 O. Okoth, 2014; J. Mwangi, 2013; J. Syanda, 2007; D. Ndetei, 2007.
22 This ratio is based on 2015 estimate national population, which was 47 million people (KIPPRA 2016).
It is calculated by dividing the national population by 100,000, and dividing the result by the number
of murders reported nationally. There are no systematic studies of the problem of murder in Kenya.
There are no published studies on the extent of this form of violence and its drivers.
23 NPS , 2016.
24 UNODC & KIPPRA, 2010, p.4.
25 Analysts suggest that in general victims of robberies and other crimes in Kenya are reluctant to make
reports to the police because they are perceived to be inept in taking the required action.
18 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 19
increasing from 250 in 2013 to 262 in 2014. Many of these crimes are gun related; in fact, firearms
were used in 12,877 violent crimes in Kenya during the period 2010-2014.26 It is estimated that
there are approximately 650,000 illegal firearms in circulation in Kenya (Regional Arms Centre
on Small Arms, 2015)
Organised Gangs play an integral role in urban violence in Kenya committing violent crimes,
including armed robbery, muggings and kidnapping, tracking of persons, extortion, firearms
smuggling and the drug trade.27 The National Crime Research Centre (NCRC) estimates that
there were at least 46 gangs in Kenya as of 2013 comprising mainly of young men in their 20s,
even though there is also evidence that children have been recruited into some of these gangs.28
Gangs are known to use violence against members of the public and with other gangs.29
Police abuses and violence have also been identified as a major challenge in Kenya by various
observers including government watchdog institutions, local and international human rights groups.30
The police have been known to misuse their power and harass citizens. The Data Department of
the Nation Media Group reported that they had documented 262 extrajudicial killings by the police
during in the past two years, about one incident every two to three days IMLU estimated that 308
people were summarily executed by police in various parts of the country between January 2015
and March 2017. 31 IMLU data shows that more than a third of Kenyans reported that they had been
victims of police brutality and harassment on at least one occasion since 2011.32 The poll found that
39.4% of violence against civilians by the police took place in police cells, while 41.2% took place
during arrests. Disappearances of suspected terrorists is another contemporary concern for Kenyan
human rights groups.33 The Kenya National Commission on Human Rights (KNCHR) estimated that
about 81 people had been forcibly disappeared, and may have been killed.34 It should be borne in
mind that the police have themselves been victims of violence in the line of duty, particularly when
they respond to armed robberies and terrorist attacks. In 2015, for instance, some 28 ocers were
killed in the line of duty, while 102 were injured and/or maimed.35
Politically-Instigated Political and ethnic violence is also an important type of violence, which occurs
mainly during election seasons but also related to heightened political activity. It is estimated that
political violence has led to about 3,500 deaths between 1997 and September 2013.36 Ocial state
inquiries into the 1992 and 1997 electoral violence found that political actors, particularly senior state
ocials opposed to democratic change, instigated violence against certain ethnic communities that
were deemed to support the opposition political parties.37 The motivation of fomenting violence was
26 The Regional Centre on Small Arms, 2015, p.9.
27 NCRC, 2013, p. 25.
28 NCRC, 2013.
29 D. Anderson, 2002, pp. 531-555..
30 IMLU, 2014; IPOA, 2013 see also reports by KHRC; Amnest y International, Human Rights Watch etc.
31 IMLU 2015; IMLU, 2016; NMG, 2017; See http://www.reuters.com/article/us-kenya-police-idUSKC-
N12206E
32 IMLU, 2016.
33 Amnest y International, 2015; Human Rights Watch, 2015.
34 KNCHR, 2015, pp. 6.
35 NPS, 2015, pp.8.
36 ACLED, 2017, pp.1.
37 See the various reports on electoral related violence such as the Kiliku Report (Republic of Kenya,
1992; Akiwumi Report ( Republic of Kenya, 1999)
20 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 21
to frighten the Kikuyu community, perceived opposition voters, living in the Rift Valley and Costal
regions, and prevent them from voting during the December 1991 and December 1997 elections.38
But the worst violence to rock the country was in 2007/2008 following disputed results of the
presidential election.39 That violence resulted in the death of more than 1,333 people, nearly 600,000
people displaced, and economic losses valued at more than KES 300 billion, equivalent to 3 Billion
USD.40 Violence between supporters of both the opposition and the government sides manifested
in communal riots, killings, massive internal displacements, rape and lynching around the country.41
Gangs of young men took advantage of the situation, and eectively paralyzed the country’s road
network by barricading roads, harassing, robbing and assaulting motorists and travellers. The ocial
inquiry into the 2008 post-election violence found that politicians incited their followers to engage
in violence, and supported the formation/activities of gangs and vigilantes, which carried out violent
acts.42 Some aided the formation of gangs, and others planned and funded them to mete out
violence against their opponents and ethnic groups not supportive of their political objectives.43
Resource based conflicts, related to access to land, pasture, and water are also common in
Kenya, even though they typically occur in rural and remote places.44 Commonly in the northern
and coastal areas of Kenya, the conflicts typically pit dierent pastoral communities against
each other, or pastoralists against farming populations. Such violent conflicts heighten during
periods of drought, when water and pasture resources are strained. Most of the perpetrators of
this kind of violence tend to be young men, armed with crude arms and increasingly firearms.45
It is claimed that political actors exploit these conflicts to marginalize their opponents.46 Large
numbers of livestock have been stolen during these attacks.
Kenya has also had a challenge of Terrorism since the 1990s. In the 1990s and early 2000s, Al
Qaeda elements infiltrated the country, carrying out several attacks, including the bombing of the
US Embassy in Nairobi in 1998. The problem of terrorism grew with the emergence of Al-Shabaab
terror group in neighbouring Somalia, which started to carry out attacks in the northern and coastal
parts of Kenya in 2011. In 2016, the National Counter-Terrorism Centre reported that approximately
900 people, including security personnel, had been killed in terrorist attacks in Kenya since the year
200047. The period 2013-2015 saw a major shift in the scale, frequency and style of the attacks. Six-
seven (67) people were killed in the Westgate Mall attack in Nairobi, while 280 people died in four
major attacks in 2014-2015.48 Before 2013, terrorist attacks took the form of suicide bombings, but
terrorist strategy has changed to favour shooting attacks on people.
38 KHRC, 1998; NCCK, 1997.
39 E. Okere, 2012; A. Nderitu, 2012; S. Sebastien et al, 2012; Republic of Kenya, 2009.
40 Republic of Kenya, 2009.
41 Ibid.
42 Ibid.
43 Kamungi, 2009, pp.353; Mohammed, 2015, pp.49.
44 CH RIPS , 2016.
45 Ibid.
46 Oucho, 2002.
47 Ombati, Cyrus (2016), 900 Kenyans killed in terror attacks in past 16 years, The East African Standard.
Link: https://www.standardmedia.co.ke/article/2000194525/900-kenyans-killed-in-terror-attacks-in-
past-16-years
48 ACLED, 2013; KNCHR 2015, pp. 3.
20 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 21
2.3 Vulnerability and Victimization
Emanating from the above analysis, there are several factors that can be used to explain the
vulnerability of certain groups of people to violence.
Most of the studies into the prevalence of violence in Kenya associate the phenomenon with
high levels of poverty, unemployment, disillusionment and weak social cohesions.49 As such,
there is a multiplicity of literature showing the violence in Kenya is particularly prevalent in poor
or low-income neighbourhoods in urban areas. As noted above however, the approach to this
question requires more nuance.
The second is gender. As highlighted above, women are a greater risk of experiencing violence
in their homes than men.50 Similarly, girls are more likely to have experienced physical violence
than boys.51 Even so, elderly, widowed women in places like Kilifi and Kisii counties, are at higher
risk of being lynched by being falsely accused of being sorcerers so that they may be dispossessed
of the land and property. However, political violence disproportionately aects men. Men are by
far more involved in perpetrating or countering this form of violence. Statistics of various ocial
inquiries into past episodes of politically-instigated political and ethnic violence in Kenya show
that men were the overwhelming majority of people killed.52 Additionally, men are much more
likely to be the victims of mob violence than women in Kenya.
The third risk factor is age. Nearly 70% of Kenya’s population are youth, defined in government
policy as persons between the ages of 18 and 3453. It can therefore be inferred that people of
in this age bracket are more likely to experience violence in contrast to those falling below or
above the bracket. Intuitively, children (defined as persons below 18 years in Kenyan law) are
more likely to experience violence, the associated psychological trauma, and neglect, because
of their vulnerability. However, there have been no violence victimization surveys that sampled
children, and it is therefore not possible to determine the extent of the risk they face vis-à-vis
adults.
The fourth risk factor is class and ethnic identity. From the review of available literature,
it does not seem that socio-economic status and ethnicity are individual risk factors with
respect to domestic violence, sexual violence, violence against children or murder, robbery
and kidnapping in Kenya. They are, however, clear individual risk factors in political violence in
the country. Poor people in urban areas are more at risk of being victims of political violence
than their auent, middle-class counterparts. Minority ethnic communities in regions with
ethnically hegemonic groups are also at greater risk of being victims of political violence.
49 Schuberth, 2014; LeBas, 2013; KNCHR, 2014; National Crime Research Centre, 2012; UNICEF, 2011.
50 KDHS, 2015 .
51 UNICEF 2010, p.2; Ruto 2009, p. 181.
52 Republic of Kenya, 2009; Republic of Kenya, 1998.
53 Constitution of Kenya, 2010.
22 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 23
2.4 Mitigation
Kenyans assess the criminal justice system (police, courts and prison) as generally ineective,54
and distant, complex and expensive.55 Nonetheless, studies in poor urban neighborhoods
have shown that Kenyans prefer the police to the other institutions in addressing their security
concerns. In his study of policing in Githurai, Wairuri noted that this preference is informed by
the proximity and accessibility of the police as well as their ability to dispense justice in a manner
that meets community expectations.56 This notwithstanding, Kenyans still complain about police
corruption.57 Some complain that reporting thieves to the police therefore loses meaning, as it
becomes just another opportunity for the police to collect bribes. The important take away from
here is that people often report cases to the police, despite the challenges noted above, but not
with the intention of the cases ending up in court. Rather, they usually expect the police to resolve
the cases.
Since the government is unable to guarantee security to every person, individuals and communities
develop and employ their own mechanisms for generating security and justice outcomes .58 These
alternative mechanisms take dierent forms including improving security of homes (dogs, metal
grills), community vigilantism either in the form of mob justice or through organized groups such
Mungiki, Kamjesh and Taliban .59 Many communities have developed their own mechanisms for
dealing with violence. Lynching or Mob Violence, popularly known as ‘mob justice’ in Kenya appear
to be a response to the perceived ineectiveness in the criminal justice system60. Mob violence
typically targets criminals, especially petty thieves, and persists in both urban and rural areas. Elderly
people in rural areas in parts of coastal (Kilifi) and Western Kenya (Kisii) have been victimized on
spurious claims that they are sorcerers.61 These accusations are instigated by criminals seeking to
steal land and property from lonely, elderly people. Local media regularly report on incidents of
mob violence. But there has been no consistent documentation of cases and trends by the police,
civil society or research institutions. Robert McKee has attempted to address this gap by reviewing
media reports on mob violence.62 McKee found that some 1,500 people were reportedly lynched
during a 17-year period, from 1996 to 2013.
Another key dimension is the employment of private security agencies. As Garland noted,
once security ceases to be guaranteed to all citizens by a sovereign state, it tends to become a
commodity, which like any other is distributed by market forces rather than according to need.63
54 Wairuri, K (2015), Popular Suppor t for Police Violence in Nairobi, Unpublished MSc Thesis: University
of Oxford
55 See Owen & Cooper-Knock, 2014
56 Wairuri, 2015; See also CHRIPS recent conflict assessment of 7 counties where people complained
about the police but still saw the police as important. Ayiera, E (2017) studying violence in poor urban
neighborhood of Nairobi reached the same conclusion.
57 Akech, 2005; Ruteere & Pommerolle, 2003; Anderson, 2002.
58 Hornberger, 2013; Owen & Cooper-Knock, 2014; Baker, 2006; See also, CHRIPS,2017.
59 Baker, 2006; Tankebe, 2009; Anderson, 2002; Katumanga, 2005; Ruteere, 2008; Rasmussen, 2013;
Kagwanja, 2003.
60 Baker, B. Multi-Choice Policing in Africa, Journal of Modern African Studies, 47:01, 154-155, 2009.
61 Joseph, 2009;Miguel, 2005.
62 Mckee, 2001: 1
63 Garland, 1996:463.
22 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 23
2.5
This section highlights the conceptual choices we have made to help us understand the
phenomenon of urban violence in Nakuru better. This include the definition of violence, the
categorisation of violence, and analysis of the vulnerability.
We recognise that the definition of violence is highly contested. While we acknowledge, the
distinctions made by Moser & McIlwane for instance, between political, institutional, economic
and social types of violence, we have adopted a parsimonious definition of violence.64 For this
study, we adopt the instrumental definition of violence advanced as the use of physical force,
which causes injury to others to impose one’s wishes.65 The only addition we make to the definition
is to include the threat of violence and verbal abuse which consider significant manifestation
of non-physical violence because the result is still the imposition of the perpetrator’s wishes
against a victim. This is very much in line with the violence definition as defined in the WHO
World report on violence and health (WRVH), namely: “the intentional use of physical force or
power, threatened or actual, against oneself, another person, or against a group or community,
that either results in or has a high likelihood of resulting in injury, death, psychological harm,
maldevelopment, or deprivation.” This is not to disregard the evident forms of structural violence
embedded in the Kenya society, especially in the deprived neighbourhoods we are studying, but
rather to make a pragmatic choice to understand the types and dynamics of violence that can
be addressed through programmatic interventions.
Consequently, rather than seek to understand violence as political, institutional, social or
economic as suggested by Moser and McIlwane, we have decided to categorise violence as
Sexual and Gender Based Violence (SGBV), Violence against Children, Criminal Violence, Police
Violence and Political and ethnic violence. The category of criminal violence includes major
crimes such as murders, robberies and kidnappings constitute the other key category of violence
in Kenya.66 This decision is based on several considerations. First, the data available does not
lend itself to a neat mapping onto the schema proposed by Moser and McIlwane but rather is
received in this format that we are proposing. In fact, Moser and McIlwane observe that they are
not neat, clearly distinguishable categories. Secondly, our categorisation will help us develop a
more nuanced view of the dynamics of violence and therefore inform the recommendations
made for interventions to address specific forms of violence.67 We agree with the proposal
by Moser and McIlwane to understand violence through the framework of structure, identity
and agency. In our attempt to understand perpetrators and victims, we are careful to employ
this framework.68 This is particularly helpful in avoiding the common analytical traps that
sometimes deny key actors agency by, for instance, suggesting that all poor people are violent.
This framework also helps us to remain mindful not to allocate the same amount of agency to
all actors despite evidence of dierential power relations within the communities that we are
studying. Finally, we also adopt the recommendation by Moser and McIlwane to understand
the dynamics of violence - and response mechanisms - as occurring at three dierent levels:
individual, community and state. Similar to the four-level social-ecological model (SEM) which
seeks so enlighten why no single factor can explain why some people or groups are at higher
64 Moser & McIlwane, 2006.
65 Keane, J. (1996). Reflections on Violence. London: Verso.
66 Moser & Mcllwane, 2006.
67 Ibid.
68 Ibid.
24 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 25
risk of interpersonal violence, while others are more protected from it. This framework views
violence as the outcome of interaction among many factors at four levels - the individual, the
relationship, the community, and the societal. The SEM allows us to address the factors that put
people at risk for or protect them from experiencing or perpetrating violence (risk and protective
factors) and the prevention strategies that can be used at each level to address these factors.
Our study therefore makes an attempt wherever possible to distinguish the dynamics of the
various types of violence at each of these levels of analysis.
24 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 25
Methodology
2.6 Overview
The study adopted a mixed-methods approach consisting of both quantitative survey and
qualitative research. The study employed some qualitative approaches - literature review and
collection of routine statistics (including NGOs reports and maps) – to refine the key questions
for the study.
2.7 Study Sites
The study covered Nakuru Town Municipality (Bondeni and Kaptembwo informal settlements),
Naivasha Sub-County (Kabati and Karigita informal settlements) and Molo Sub-County (Molo
Town). These sites were partly selected based on the fact that they are informal settlements. It is
estimated that most of the county’s 1,867,461 people live in informal settlements in Nakuru Town
Municipality concentrated in Bondeni-Kivumbini, Free Area, Lake View, Mwariki, Kaptembwo
and Kwaronda.69
2.8 Quantitative (Household) Survey
A key component of the study was a quantitative survey that reached individuals in 1,780
households, covering Nakuru Municipality, Naivasha Town and Molo Town, through face-to-
face interviews. The survey was conducted based on a structured questionnaire (annexed) that
was revised three times and then translated into English and Kiswahili. The household survey
questionnaire was designed to gauge respondents’ knowledge, perceptions and experiences of
violence, violence distribution and prevalence, violence typology, risk factors and violence trends
and dynamics. The survey was administered through a Mobile Data Collection (MDC) platform.
Questionnaires (annexed) were administered to randomly selected household respondents in
randomly selected areas in the five study sites. Once data was gathered from each respondent, it
was immediately transmitted to a secure server for later retrieval and analysis.
Field Interviewers and Training: A team of 35 interviewers were recruited from Nakuru Town
from CHRIPS database of experienced field researchers to conduct the survey. A 1:1 ratio for
gender balance was ensured in the recruitment. Given the total sample size used was 1,780
households, each interviewer conducted approximately 50 interviews over a six-to-seven-day
period. Therefore, the interviews allocated to each interviewer amounted to about 2.8% of the
total sample, failing within the recommended maximum of 5%. To ensure uniformity and for
quality control purposes, the research team, in collaboration with DIGNITY, trained enumerators
on: (a) Description of survey; (b) Roles of team leaders and interviewers; (c) Sampling intervals
to be used; (d) Handling of questions from respondents; (e) the MDC platform; and (f) the Kish
Grid. Each enumerator spent a whole day carrying out seven pilot surveys to gain familiarity with
the questionnaire. During the ensuing debrief, their comments on the wording and flow of the
questions were particularly useful, and informed the fourth and final revision of the questionnaire.
69 Lowe, Lucky (2001), Nakuru - a study in urbanization, http://livelihoodtechnology.org/home.
asp?id=csNakuru .
26 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 27
Household Identification: Like many household surveys, a household is primarily-(a) unit of people
who eat from the same food pot and (b) are answerable to one household head (c) could also be
living under the same roof though not necessarily. This information was verified by the interviewers
asking the questions about the number of people who live under the same roof and eat together.
Once at the sampling point (sub-location), the interviewers reported to the gatekeepers, in this
case the assistant chief/chief to whom they introduced the survey. In some cases, the chief then
personally or through an assigned village elder help them identify the boundaries of the sub-
location. This measure was taken to ensure that the survey was conducted within the randomly
selected sub-location. The interviewers then identify a landmark e.g. school, church, mosque,
police post, chief’s camp, shop etc. closest to the selected households. For this purpose of
household based surveys, a landmark is defined as a permanent feature at a specific location that
will be available for many years to come.
Once they reached the relevant sub-location, the interviewers situated themselves at the selected
landmark and using the “date score”, decided on the first household to be interviewed (sampled).
This was done by summing up the date in case the date has two digits. For instance, on the 21st
February 2017, all the interviews started at the 3rd household from the landmark, (i.e. 21st; 2+1=3),
while on the 22nd, the interviewers summed up 2+2=4, then started the interviews at the 4th
household from the landmark, and so on. Where the interview at the first sampled household
was successful, four households were then skipped, and the next interview was undertaken at
the 5th household. After every successful interview, this skip pattern was repeated to ensure that
each sampling point was well covered until the required numbers of interviews were achieved. In
the event of an unsuccessful interview, the interviewer knock on the immediate household until a
successful interview was achieved. As the interviewer walked from the landmark and the sampled
household, they kept left as they look for the next household. (This is called the “left hand rule”.)
The respondent selection process was done after a successful household identification process.
The respondents were selected by use of the Kish Grid (annexed) in which case the interviewer
listed all the occupants of the household aged 18 years and above. The interviewers then used
the last digit of the serial number on the questionnaire to select the respondent to be interviewed.
In instances where the selected respondent was not available at the time of call, three call backs
were made in attempt to find the respondent. Where the respondent was completely unavailable,
the interviewer made a substitution by moving to the subsequent household. Debriefs with team
leaders, the research team learned that the data enumerators encountered few challenges in the
administration of the survey. Respondents were generally cooperative, although some appeared
to be uncomfortable with some of the questions that focused on issues such as marital rape.
Chiefs and their assistants were also cooperative in providing the required support to interviewers.
Sampling frame: The overall sampling frame to be used was derived from the 2009 population
and household census data.70 The research team settled on using a multi-stage stratified sampling
strategy. The study targeted the adult population, (18 years and above), male and female living
in the selected urban areas. Refugees and non-members of the household were not interviewed.
To ascertain this, the interviewer listed all those present in the household, and by asking specific
questions, clarifying whether they would be considered household members or not, Kenyans or
non-Kenyans. As such visitors and refugees were not be considered members of the household
thus not listed in the Kish Grid.
70 KNBS 2009 National Population Census.
26 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 27
Sample size determination and distribution: In determining the sample size of this study based
on the sampling requirements, the following formula was applied.
n = Nz^2pq/(E^2(N-1)+ z^2pq)
Where
N=adult population size of Nakuru Town (308,791)
z=confidence level (1.96)
E=margin of error (3.45%)
p=the proportion of a population with a particular characteristic (0.5)
q=the proportion of a population not having this characteristic (0.5)
On this basis, a total sample size of 1,780 respondents was drawn from the 2009 National
Population Census data71 that had estimated the combined population of the five study sites be
187,887 people. This sample size allowed for a low margin of error of +/-2.3%, ensuring a high
degree of accuracy and credibility of the data gathered. It should be noted that the research
team unsuccessfully attempted to get more recent population data from the Kenya National
Bureau of Statistics (KNBS), the Nakuru County Government, and NGAO administration chiefs.
All the sources indicated that they did not have any new estimates of the change in population in
the study sites since 2009, and therefore relied on that year’s census. It is unlikely any new data
will be available until the planned 2019 census is undertaken.
The sample of 1,780 respondents was distributed across the five study sites as follows:
Table 3: Distribution of Sample in the Five Study Sites
Town Fields of study Households
Percentage
in the Total
Nakuru Town
Bondeni 20,574 6,148 11% 195
Kaptembwo 70,352 23,200 37% 666
Naivasha Town Karagita 39,209 12,953 21% 371
Kabati 25,437 9,194 14% 241
Molo Town Molo Town 32,315 8,664 17% 306
Total 187,887 60,159 100% 1,780
The survey achieved broad demographic representation by reaching respondents across age and
gender. Fifty-six percent of the total participants were female while 44% were male. Twenty-seven
percent of all the respondents were between the ages of 18 and 24, 41% between 25 and 34 years
old, 19% between 35 and 44 years; and 13% were 45 years and above. More than 60% had lived in their
areas of residence between 1 and 10 years. Regarding their sources of income, 34% of the respondents
claimed to be doing business, 27% were manual labourers and 20% were unemployed.
71 KNBS 2009 National Population Census.
28 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 29
2.9 Qualitative Data
Review of the relevant literature, helped us to develop an overview of the pertinent issues
with respect to urban violence in poor, urban neighbourhoods in Kenya. The literature
review, which continued throughout the study period, also helped in the development of the
research instruments, which included questions for key informant interviews (KIIs) and the
Focus group discussions (FGDs). The findings of the preliminary literature review (analysed
below) demonstrated that the key questions for this study were appropriate and had not been
adequately explored in previous studies. The review also enabled the identification of key
institutions working on issues related to violence in Nakuru, from which key informants were
recruited for the study.
The study also included key informant interviews and focus group discussions. Key informants
were selected across dierent categories: men, women, youth, civil society organisations,
government institutions and individuals. We conducted a total of 55 KIIs in Nakuru, Naivasha
and Molo. The interviews were semi-structured in nature with the main questions based on
key themes emerging from the literature review and informed by the objectives of the study.
The interviewees were selected through purposive sampling techniques from the key groups
including civil society organizations, community-based organizations, religious organizations,
community opinion leaders, local politicians, ocials of the Nakuru County Government and
the National Government Administration Oce, health workers, senior National Police Service
ocers and private sector actors. Key informant interviews took between 1 and 2 hours. In some
cases, informants provided supporting documentary and statistical data/evidence, which has
been used in this report. The full list of the interviewees is annexed.
Additional qualitative data was gathered through 13 focus group discussions (FGDs) held in
all the study sites. They were particularly useful in gaining local residents perspectives on key
questions of urban violence, identifying how they feel about the causes and consequences of
urban violence, as well as gaining ideas on potential solutions to these problems. The FGDs
brought together between six (6) and eight (8) participants from the various groups including
local opinion leaders, women, youth and men/elders. The discussions were moderated by
a member of the research team, and focused on the key themes and questions identified in
the initial literature review. As with the key informant interviews, responses from FGDs where
incorporated into Google Forms-based questionnaires.
Additionally, the study included a quantitative household survey, which relied on a structured
questionnaire designed to gauge respondents’ knowledge, perceptions and experiences of
violence, violence distribution and prevalence, violence typology, risk factors and violence
trends and dynamics.
28 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 29
Research Ethics Considerations
The study was guided by the principle of informed consent: All participants in the study, including
survey respondents, key informants and FGD participants, were clearly informed about the risks and
benefits of their participation in the study. The study was guided by the principle of informed consent.
Each participant was informed that their participation in the study would be kept confidential, and
that information provided would not be directly attributed to them. They were also informed that
they had a right to decline to participate, and their right to withdraw from the study at any time.
Quality Assurance and Review Mechanisms:
Several complementary approaches were applied to ensure data quality during the qualitative and
quantitative phases of the baseline study. Refresher training and induction of enumerators for the
quantitative survey was conducted on February 7, 2017 in Nakuru Town. All enumerators signed
on to CHRIPS’ research principles and ethics code document. As part of quality control, 5 % of all
interviews for the survey were accompanied by team leaders in the field. The measure was taken to
ensure that interviewers followed the instructions and procedures in the sample specifications in the
questionnaires and conducted the interviews per the specified standards.
Regarding quality control for the collection and analysis of qualitative data, all data gathered through
qualitative research—key informant interviews and FGDs—was corroborated by cross-referencing it
with dierent sources.
2.10 Limitations and Mitigation Measures:
Violence is complex and sensitive issue in Kenya, as it is anywhere else. Therefore, there is always a
likelihood of getting inaccurate or unreliable information from key informants, FGD participants and
quantitative survey respondents. While this limitation is not one that can be completely overcome,
the study attempted to do so by using a good sample size (for both the key informant interviews and
the quantitative survey), as well as multiple data collection methods to increase the odds of getting
accurate information. To enhance the accuracy of information, all data gathered through- key
informant interviews, FGDs and quantitative survey- has been triangulated to generate the report’s
findings and analysis.
Being aware of the security sensitivities surrounding the subject of violence generally, the research
team accommodated the concerns of key informants about confidentiality to ensure that their
participation in the study does not expose them to any security threats.
Another key challenge faced in the field was lack of access to certain key informants, in particularly
Members of County Assembly and County Government ocials. This challenge was partly addressed
by finding alternative respondents who are knowledgeable on the issues in question.
The field research phase was conducted in three phases. The first phase, which took place between
6th and 10th February 2017, involved the pre-testing of all the data collection tools in Nakuru Town
through mock interviews and FGDs. Pre-testing enabled the research team to identify gaps and
problems in the tools, and provided the basis for refining them accordingly. The second Phase took
place between 20th February and 1st March 2017 and involved the collection of data including
key informant interviews, FGDs and the quantitative survey in the five study sites. The third phase,
30 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 31
which took place between 10th and 12th April 2017 included additional key informant interviews
with senior leaders of the National Police Service in all the five study sites, and two additional FGDs
in Nakuru Town (Bondeni and Kaptembwo).
2.11
In total 1,780 household interviews were performed. 19% of the population was from Molo Town, 15%
from Naivasha Kabati, 23% from Naivasha Karagati, 12% from Nakuru Bondeni Village and 31% from
Nakuru Kaptembo. 55% of the population consisted of women and most of the study participants
were younger. 27% were 18-24 years and 42% were 25-34 years, 18% were 35-44 years old, 10%
were 45-54 years and the oldest 55 years and above only consisted of 4%. Almost a quarter of the
participants had 5 or more members in their household (22.8%). 38.5% had a household consisting
of 3-4 members and almost the same number of participants had a household consistent of only
1-2 people (38.7%). These detailed table showing the characteristics of the participants is included
in Appendix 1.
The number of participants with no formal education was very small, standing at 1.6%, 8.9%
had started a primary education and more had completed primary education (21.5%). The most
participants had finished a secondary education corresponding to 40%, whereas some had begun,
but not finished (13.9%). 15% had a higher education than secondary.
Over half of the population were married (56.5%) and 35% were single, probably due to the high
number of young people in the study population. 3.8% were divorced/separated and 3.6% were
widowed.
Most people were involved with business (35%), some involved with manually skilled and unskilled
work and more than 1/3 were unemployed. Other practices like clerical, domestic service,
professional and others consisted of the rest from 3-5% in each category. Most participants had
an income Kshs 5,000-10,000 per month (37%). 16% had an income between Kshs 1-5,000, 26%
between Kshs 10-20,000 and 8% had a monthly income of Kshs 20,000-40,000 per month. 1%
of the population had an income higher than Kshs 40,000 and 11% refused to answer. A report
published by the World Bank in 2016, details the decline of those living in global poverty, which is
reclassified as living on $1.90 or less a day (corresponding to less than 196 Kshs per day), to a forecast
9.6%of the world’s population in 2015 - a projected 200 million fewer people living in extreme
poverty than in 2012.72 This means that at least 16% of the study population live in poverty along with
a proportion of the Kenyans that in the next category (5,000-10,000) earn less than 6,000 Kshs per
month. Furthermore, there is a chance that the reason for not responding to the question is due
to a low household income. 200 participants corresponding to 11% of the study population chose
not to answer this question. The sensitivity of this category has been identified in other studies as
well.73 In addition to the income estimates, it was seen that almost the whole study population used
communal toilet facilities corresponding to 94.4%.
Almost all participants were Christian (95%). Half of the respondents belonged to Kikuyu (47%). 13-
72 http://www.worldbank.org/en/publication/global-monitoring-report.
73 The high non-response rate of the income question in sur veys has been the subject of many other
studies. See for instance Riphahn, R.T., & Serfling, O., (2002), Item Non-Response on Income and
Wealth Questions, IZA Discussion Paper No. 573; Yan, T., Curtin, R., Jans, M. (2010), Trends in Income
Nonresponse Over Two Decades, Journal of Ocial Statistics, Vol. 26, No. 1, pp. 145-164.
30 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 31
14% of the study population were either Kisii, Luo or Luhya and 7% were Kalenjin. 2% were Kamba
and 1% were Maasai or Meru. Also, 1% refused to answer. The rest belonged to ethnic groups:
Turkana, Swahili, Somali, Samburu, Mijikenda, Meru, Maasai, Borana and Embu and were placed in
the “Others” category.
In the below analysis, age was categorized into four categories 18-24, 25-34, 35-44 and 45+. The
45+ group had less than the other categories, which was why it made sense to combine 44-55 and
55+ to into one category.
Key Findings and Analysis
2.12 Overview
The following section presents an analysis of the research data collected from participants in
the study sites. It highlights the types of violence experienced in the areas, their distribution and
prevalence and dierent response mechanisms at the individual, community and State levels.
2.13
Nakuru County is in Kenya’s south Rift Valley area, and is the former headquarters of the Rift
Valley Province. It shares borders with several counties: Kericho and Bomet (west), Baringo
and Laikipia counties (north), Kajiado and Kiambu (south), Nyandura (east), and Narok (south
west). The county has elleven administrative sub-counties: Naivasha, Gilgil, Nakuru, Rongai,
Nakuru Town West, Nakuru Town East, Subukia, Njoro, Molo, Kuresoi South and Kuresoi North.
The Nakuru County Government estimates that the county had a population of 1,867,461 in
year 2014.74 Of the total population, 937,131 were male, while there were 930,330 females. A
large proportion of the county’s population live in informal settlements in Nakuru and Naivasha
Municipalities. This population is concentrated in Bondeni-Kivumbini, Free Area, Lake View,
Mwariki, Kaptembwo, Kwaronda, Karagita-Mirera, Kihoto and Kabati in the two municipalities
of the municipality.
Nakuru County’s economy is mainly based on small-scale agriculture, dominated by dairy farming
and grain production. Retail, tourism and light manufacturing also contribute significantly to
the local economy. Most people are self-employed or work in the so-called informal sector,
and poverty levels remain high at 45 %.75 The Nakuru County Government recognizes these
challenges and has developed the County Integrated Development Plan (CIDP)2014, and the
County Medium Term Expenditure Framework (MTEF) 2014-2016 to partially address them.76
One of the county’s key policy priorities is the reduction of high levels of recurrent expenditure
(51% of the county’s budget pays wages of government ocials) so as to free up funds for
development. Another key priority is the improvement of access to better quality healthcare
74 CI DP, 2 014.
75 KIPPRA, 2016: 177.
76 The Nakuru County Integrated Development Plan is aligned to priorities of various national and in-
ternational development frameworks, including Kenya Vision 2030 and the Millennium Development
Goals (MDGs).
32 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 33
services and social protection for more people, with a focus on controlling communicable
diseases and reducing child mortality.
Bondeni and Kaptembwo (Nakuru Town Municipality), Karagita and Kabati (Naivasha Sub-
County) and Molo Town are classified as informal settlements, and share many characteristics.
While all the five areas are considered cosmopolitan, the dominant ethnic groups are the Kikuyu
and Kalenjin. However, there are sizable populations of Luhya, Luo and Kisii people. The youth
are the majority in these areas, mirroring the national situation. Data from the household survey
for this study shows that these areas have a large transient population, with most people having
lived in them for between 1 and 10 years only. Most residents earn their living work in the low-
income informal sector.
All the study sites areas suer from high levels of poverty and unemployment, particularly
among young people. The limited land available in these areas cannot properly sustain their
large populations. The result is overcrowding compounded by poor housing conditions, lack of
adequate physical infrastructure, in particular drainage systems, water and sanitation systems,
access roads, adequate street lighting, and reliable electricity supply. Most residents live in
shacks and one-room homes with their families, and few own land in any of the five sites.
Insecurity persists in the five sites owing to a combination of weak police presence, frustration
among young, unemployed people, and low levels of social community cohesion.
Table 4: Key Indicators in Five Study Sites
Study Site Bondeni Karagita Kabati Molo Town
Location Nakuru Town
Municipality Naivasha Sub-County Molo Sub-
County
Size (sq km) - 5.1 - - -
Population
(2009 census data) 20,574 70,352 39,209 25,437 32,315
No. Households (2009
census) 6,148 23,2000 12,953 9,194 8,664
% aged 18-34/youth
(survey data) 72% 71% 67% 64% 65%
% of population with
secondary school
qualifications
37% 38% 43% 35% 34%
% of population with tertiary/
higher education-level
qualifications
19% 21% 15% 14% 11%
Urban violence in the five study sites, Bondeni and Kaptembwo (Nakuru Town); Karagita and Kabati
(Naivasha Town) and Molo Town—is generally understudied. Of the 26 studies reviewed for this study,
only one has a specific focus on one of these sites—Kaptembwo in Nakuru Town. Judy Ngina (2016)
uses data collected from the informal settlement to examine the relationship between gender violence
32 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 33
and its relationship to inaccessibility of water. She concludes that a significant proportion of women
in Kaptembwo face the risk of sexual and gender-based violence when they go out of their homes
to fetch water for their families. However, the study does not have a focus on SGBV in other settings
e.g. in homes. A second refereed study by Karanja Njoroge et al about street children in Nakuru Town
asserts that they are at risk of violence, but is thin on data touching on the exposure of these children
to violence or their role in committing violence.77 The other studies (which happen to be masters level
theses) broadly examine violence in Nakuru Municipality (Town), but make no mention of Bondeni or
Kaptembwo.78 Overall, it does not appear that there are any studies on violence in Bondeni (Nakuru),
Karigita and Kabati in Naivasha Town, or Molo Town. While several studies have been conducted on
the problem of politically-instigated political and ethnic violence in Molo Sub-County, they focus
mainly on the violence that took place in the rural parts of the county, and not Molo Town.
Studies show that in Nakuru County, the typical settings for violence in Nakuru County are
informal settlements in urban areas, including Nakuru Town Municipality.79Living standards
and the quality of life in Nakuru’s informal settlements are low, mainly because of poor spatial
planning of urban areas, particularly of low-income areas and informal settlements. As a result,
health, hygiene, safety, the quality of the environment and security have been adversely aected.
The first Nakuru County Government recognized the gaps in the availability of spatial data, and
its impact on planning, and started a digital topographic mapping exercise in 2015 to gather
the data required for spatial planning that will be implemented through to 2024.80 The spatial
plan commits to the building of new physical infrastructure. It also prioritizes the need for clear
planning of residential, business, industrial, and recreational areas. While the planning processes
has been completed, but the spatial plan is yet to be publicly released.
The patterns of violence identified above in the analysis on violence in Kenya can also be discerned at the local
level in Nakuru County. For instance, the victimisation of women through Gender-Based Violence prevails in
Nakuru.81 Similarly, several studies have shown that children in Nakuru, especially street children are particularly
vulnerable to violence meted out by municipal inspectorate ocers, the police and members of the public.82
Many street children in Nakuru Town Municipality come from poor and/or dysfunctional families, and have
run away from home to have more independence.83 Further, Nakuru is known to host several gangs including
Mungiki, Backyard, Gengejuu, Lumumba, Makaveli, Mtaro Base, Wajanja, Wazelendo and Westside.84
Patterns of the nature of violence in Nakuru County are also like the national trends. For instance,
it is evident that firearms are used in in crime in Nakuru County with evidence that residents of
middle-class suburbs, such as Kiamunyi, Milimani, Nakka and Section 58, have been the victims
of car hijackings, robberies, assaults and rapes in which firearms were used (R. Abduba 2010, p. 3).
Similarly, Nakuru County has also borne the risk of police excesses and extra-judicial executions.
For instance, 14 of the 308-summary execution by the police between January 2015 and March
77 Njoroge et al., 2013.
78 See J. Nyaga, 2014; R. Abduba, 2010; W. Maina, 2010.
79 See E. Majanga 2015; C. Tanui 2015; F. Vosevwa 2014; UNICEF & Save the Children 2012; P. Muiru
2012, p.135; G. Lubaale 2012, p. 38; E. Okere 2012, p.22; A. Mugalavai 2012; R. Abduba 2010, p.42; M.
Ndung’u 2009; B. Lang et al. 2008; M. Ndung’u 2009, p.77; B. Lang et al. 2008, p.8.
80 Nakuru CIDP, 2014.
81 See E. Onsumu et al., 2015; WHO, 2014; W. Maina et al., 2013.
82 Karanja et al., 2013; UNICEF, 2012, pp. viii.
83 Karanja et. Al., 2013, pp. 219.
84 Mkutu et. al., 2015.
34 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 35
2017 occurred in Nakuru County.85 Nakuru County has also borne a disproportionate level of the
politically-instigated political and ethnic violence since the return to multi-party democracy in the
early 1990. Nakuru County, then Nakuru District, was one of the places within the Rift Valley where
the Kikuyu community was attacked for being supporters of the “opposition”.86 In Kenya’s 2008 post-
election violence, Nakuru County was once again a centre of conflict partly due to its cosmopolitan
nature particularly in Nakuru, Naivasha and Molo towns. The ocial inquiry into the violence found
that some 263 deaths occurred in what is now Nakuru County during the crisis.87 Additionally, Nakuru
County, especially Naivasha and Molo, have experienced internecine violence over access to and use
of land in the past two decades.88 In Molo settlement schemes, corruption in the allocation of title
deeds has been a key source of land-related violence.
2.14
In table 5, when investigating the answer to perpetrators, the data file was split according to who
had experienced violence (EV1). Only data on persons who had experiences violence was inserted.
As table 5 below illustrates, the prevalence of violence within the last six months is 2% higher
in Naivasha(20.1) than in Molo(18.6) and 2% higher in Nakuru(22.5) than in Naivasha. The
perpetrators are more often organized groups and gangs (29.6 %), thiefs (24.3 %) and youths
(23.2 %). However, current spouses (10.9 %) and police (8.4 %) also make up a relevant part
of the perpetrators. Former spouses, other family members, local administration and others
accorded for less than 5% of the violence experienced by respondents .
Nakuru n (%) Naivasha n (%) Molo n (%) Total n (%)
Sample 596 (33.5) 595 (33.4) 589 (33.1) 1780
Experience of violence
Yes 174 (22.5) 135 (20.1) 62 (18.6) 371 (20.9)
No 599 (77.5) 537 (69.5) 272 (81.4) 1408 (79.1)
Perpetrators
Current spouse 24 (14.1) 7 (5.5) 8 (13.1) 39 (10.9)
Former spouse 8 (4.6) 5 (3.7) 2 (3.2) 15 (4.0)
Other family member 1 (0.6) 1 (0.7) 2 (3.2) 4 (1.1)
Neighbor 29 (16.7) 12 (8.9) 16 (25.8) 57 (15.4)
Organized groups/gangs 46 (26.4) 47 (34.8) 17 (27.4) 110 (29.6)
Thief 42 (24.1) 27 (20.0) 21 (33.9) 90 (24.3)
Police 13 (7.5) 14 (10.4) 4 (6.5) 31 (8.4)
Local administration 2 (1.1) 3 (2.2) 3 (4.8) 8 (2.2)
Youths 43 (24.7) 35 (25.9) 8 (12.9) 86 (23.2)
Others 2 (1.1) 2 (1.5) 1 (1.6) 5 (1.3)
Table 5: Selected Outcomes stratified by Town.
85 IMLU 2017; IMLU 2016; IMLU 2015.
86 KHRC, 1998; NCCK, 1997.
87 Republic of Kenya, 2009: 308.
88 Lang & Sakdapolrak, 2014; Boone, 2011; Kenya Land Alliance, 2009; Republic of Kenya, 1998; KHRC,
1998.
34 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 35
2.15
violence
The tables below show the Odds Ratio (OR) for Nakuru, Naivasha and Molo town, and the
confidence interval at 95 % as well as p-values. This allows for analysing what factors increase
or decreases the probability of becoming a victim of violence
In Nakuru, women may have a higher risk of being exposed to violence, whereas in Naivasha men
have an increased risk of being exposed to violence than women. Both in Nakuru and Naivasha,
individuals in 25-34 and 34-44 age brackets, have a lower risk of being victims of violence than
the younger age group 18-24. However, none of these associations were statistically significant.
In both towns, the older group (44+) where at a greater risk than the youngest group (18-24),
however this changed after additional control for marital status, education and occupation. This
tendency also changed for the age group of 25-34, going from 4% lower risk to 3% higher risk
after controlling for covariates. In Molo, all the age groups tended to be more at risk of violence
than the youngest age group between 18-24 years. The covariates therefore accorded for some
of the association between age group and violence exposure. The descriptions are a depiction
of the risks if they had been statistically significant.
In Nakuru and Naivasha, it appears that unmarried/single individuals are at a higher risk of being
exposed to violence. In Naivasha and Molo the divorced where
Table 6: Probability of becoming a victim of violence
36 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 37
more exposed than married respondents, however this tendency was reversed in Nakuru. The
widowed were more exposed in Naivasha compared to the married. This tendency was also the
other way around in Molo, whereas in Nakuru there was only a slight raised risk for the widowed.
Having a higher education level corresponding to secondary or above was related to a 10%
decrease in risk in both Nakuru and Naivasha compared to respondents with an education level
under secondary. In Molo, there was not a distinguished dierence between the groups.
Being employed seemed to be associated with a higher risk (15-20 %) of violence exposure
compared to unemployed respondents in Nakuru and Molo. In Naivasha on the other hand,
employment was a protective factor of violence, as employed had a 17 % lower violence
exposure risk compared to their unemployed counterparts.
2.16
When asked about their perception of violence, most people revealed that they feel generally
safe in their homes and about 78 % of household members had not experienced violence during
the last six months. This is consistent with the response that the majority had never considered
moving to another neighbourhood due to violence or safety reasons. The gender comparison
further revealed that women generally felt safer than men. The detailed responses of participants
regarding perceptions on safety are shown in Fig. 1.
Figure 1: Perceptions on safety and insecurity
39%
41%
36%
35%
30%
34%
38%
36%
21%
22%
22%
13%
13%
17%
15%
22%
23%
18%
26%
24%
27%
9%
16%
16%
15%
29%
26%
16%
0%
0%
4%
10%
1%
2%
1%
House
Neighbourhood
At Work
Place of study
The Streets
Market place
Public transport
Percepons on safety and insecurity
Fairly safe Very safe Bit unsafe Unsafe Does not apply
Does not apply
Unsafe
Bit unsafeVery safeFairly safe
House
Neighbourhood
At Work
Place of study
The Streets
Market place
Public transport 38%
34%
30%
35%
36%
41%
39%
17% 27% 16%
13% 24% 26%
13% 26% 29%
22% 18% 15% 10%
22% 23% 16%
21% 22% 16%
36% 15% 9%
1%
2%
1%
4%
0%
0%
Perceptions on safety and insecurity
0,00,20,40,60,81,0
At least a third of respondents indicated that they felt safe at various places, including
public transport, markets, open streets, educational institutions, work places and their own
neighbourhoods and homes. People felt most safe in their own homes and most unsafe in
the open streets. This intuitively makes sense because it people are likely to feel safest in their
own homes, where they are most familiar with conditions. It also intuitively makes that people
would not feel very safe in the open streets in the night as it is reasonable to expect that criminal
activities are likely to be undertaken at night as darkness oers cover to criminals.
36 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 37
Victims of violence:
Most people thought that the bulk of victims are youth and women though there is a perception
that girls are disproportionately targeted; and men are more victimised in Nakuru Municipality’s
Kaptembwo neighbourhood; which could be attributed to the high number of gangs perpetrating
violence in this area, as was revealed through interviews and focus group discussions. The major
gang operating in Kaptembwo is Gaza, which engages in violent crime, robbery, extortion and
political violence. There are also turf wars between Gaza and other gangs including Confirm,
which mainly operates in Bondeni.89
Gangs and neighbours were identified as major perpetrators of violence by 23 % and 14 %
of respondents respectively. Among the demographic groups, young people are also major
perpetrators, according to 24 % of survey respondents.
Causes of violence
When people were asked to identify the most important causes of violence, they noted several
causes, the most important of which was unemployment (81 %), poverty (75 %) and alcohol
and drugs (65 %). The main causes of violence within families include alcohol and substance
abuse (73 %), marital discord (62 %), poor parenting practices (61 %) and low socio-economic
household status (59 %). These views are similar to the responses from the qualitative research
where respondents highlighted the importance of unemployment and idleness, especially
amongst the youth, as a key driver of violence in the area. For some of the respondents, especially
from Naivasha Sub-County also noted that even those who are working, especially in the flower
farms, do not earn enough money to meet their needs and may therefore resort to crime. Many
of those who participated in this study also noted the issue of alcohol and drug abuse as a trigger
for domestic violence. A trader from Naivasha noted this as a big concern indicating that he sells
more alcohol that foodstu in a day. Some respondents also noted infidelity between spouses,
which causes tensions in the family. Other respondents highlighted the issue of poverty as driver
of violence in these neighbourhoods. Some respondents for instance observed that the lack of
money within the families contribute to disagreement between couples in marriages.
Nonetheless, it is important to remember that many of these issues that are being highlighted
as major causes of violence explain only one part of the drivers of violence in poor urban
neighbourhoods. That is, the respondents focus largely on the structural factors that underlie
violence such as material deprivation and unemployment as seen in figure 3. A few of the
triggers such as drug and alcohol abuse are also noted but there is little regard to the agency of
the individuals perpetrating the violence and those being victimised. It is therefore important to
emphasize that this analysis should be viewed in context alongside other dynamics of violence
highlighted in the sections that follow.
89 Interview with senior member of Nakuru County Peace Committee; and senior ocer in Kaptembwo
Police Station.
38 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 39
Figure 2: Most prevalent types of violence
0% 10% 20% 30% 40% 50%
Sexual violence
Political violence
Ethnic violence
Gang related violence and vigilantism
Youth violence 45%
38%
34%
28%
14%
What type of violence is mostly manifested in your area?
As the figure above shows, violence perpetrated by youth and gangs is perceived to be
most prevalent in Nakuru County. This makes sense as the profile of most violent criminals
in Kenya, and indeed across the world, tends to be people drawn from the youth category.
Political and ethnic violence have aected Nakuru in each of the four general elections
held in the country since 1992, and it is therefore not surprising that a large proportion of
respondents, 34% and 28%, respectively, cite these types of violence as prevalent in the country.
Figure 3: Most important causes of violence
0 20 40 60 80 100
Others
Inequality between gender and social groups
Availability of arms
Poor quality of education
Ethnicity
Politics
Cultural norms that support violence
Lack of street lighting
Poor rule of law
High crime levels
Local illicit drug trade
Poverty
Unemployment 81%
Most importnat causes of violence in your opinion?
75%
65%
43%
36%
27%
27%
25%
25%
20%
19%
15%
2%
38 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 39
Unemployment and poverty were cited by survey respondents as the leading factors for violence
in Nakuru country. The link between joblessness and poverty has been clearly established in
many academic and policy studies in Kenya as well as globally. The drug trade is also cited as a
leading cause of violence. This is not surprising as it is well-established globally that drug dealers
enforce the rules of their trade and gain the initiative by resorting to, or threatening rivals, and
even their clientele, with violence.
2.17 Prevalence and Dynamics of Violence in Nakuru County
The study revealed that violence is a major concern for residents of Nakuru County. Most
respondents to the study felt that violence was becoming more prevalent. On the most
common types of violence in the county, 46 % of victims had suered physical assault, 32 % were
threatened and intimidated while 26 % were violently robbed or mugged. Sexual and gender-
based violence appears widespread in the County as the most common type of violence among
households, according to 76 % of respondents. Child abuse was also cited as prevalent by 19 %
of respondents, though it appeared more prevalent in Molo (at 29 %). As such the data revealed
five major types of violence namely: (1) SGBV, (2) violence against children, (3) violent crime, (4)
police violence and (5) political and ethnic violence. Nonetheless, the study also revealed that
the types of violence are distributed dierently among the various neighbourhoods in the study.
2.17.1 Sexual and Gender-Based Violence
Key informants and FGD participants reported that sexual assault is a leading to broken families, spread
sexually transmitted infections and HIV/AIDS, trauma and even death. This kind of violence was seen
to mainly target women and children, with 25-40-year-old men (husbands or other men) as the main
perpetrators. However, some cases of men being victimised were also noted. For instance, police
ocer from SGBV department in Kaptembwo told the story of a man who was assaulted by his wife:
“The husband was misusing money and eating
food when the children had not eaten…the woman
poured hot water on him and he was hospitalised
for five days. He pleaded with the police not to
arrest his wife because she was the breadwinner.”90
The respondents indicated that the perpetrators are often known to their victims. This type of
violence was found to be present in all areas of the county included in the study. Also common
is sexual violence against children, by men. Respondents in Nakuru Town expressed worry over
the increase in child rape cases.
Victimisation appears to be closely linked to space, home or street, and gender identity of the victim
even though men remain the main perpetrators in both cases. Our respondents noted that the main
forms of SGBV within the home included domestic violence and sexual assault. A few cases of men
sexually assaulting their own children were highlighted. In one case, a respondent highlighted how
90 Interview with police ocer in SGBV depar tment, Kaptembwo Police Station
40 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 41
a man violated his two-year-old daughter.91 Female house-helps were also noted to be victims
to male sexual assault within the homestead. There were also a few cases of young men raping
their own mothers, which were highlighted during the interviews. Children are the most vulnerable
outside the home. Several cases of children being sexually assaulted after being lured with gifts from
relatives and strangers were cited. Some of the respondents also noted cases of prostitutes being
assaulted by young men92. The use of abusive language and threats of sexual assault against women
especially among matatu operators, was also highlighted as another common form of SGBV. A
senior ocial in the Nakuru County Transport Department pointed out that:
“Sexual violence against women is a major issue.
Conductors touch them or talk to them in ways
that amount to sexual harassment.” 93
Female Genital Mutilation (FGM)94, was also highlighted as a common form of SGBV especially
in Naivasha. Some social groups, for instance the Wakorino religious formation, wasi dentified
as perpetrating this practice.
When asked to identify the major causes of SGBV, respondents identified the usual factors that
have been highlighted in previous studies including alcohol and drug abuse and infidelity in
marriages and general moral decay in society. Some respondents reduced the causes of the
violence to material conditions within the family. While some argued that violence results from
tensions caused by material lack or inequality caused by increased access to financial resources
for women through the growing micro-finance institutions. From the conceptual framework, it
is important for us to note that these perspectives however take away the agency of the victims
and perpetrators. Additionally, it appears that this latch on to other external discourses since it is
obviously not true that all cases of material lack lead to domestic violence. Another interesting
observation is that while people are often able to identify dierent types of SGBV, they often
only speak about domestic violence when it comes to identifying causes.
2.17.2 Violence against children
In addition to aforementioned sexual abuse and assaults towards children, respondents across
the study highlighted violence against children as prevalent. More than half (54 %) of the study
population identified it as a problem in their neighbourhood. This type of violence appeared to
be particularly serious in Molo Town, with 66% of the respondents identifying it as a problem,
though the prevalence is still high in other areas.
Violence against children was noted to have negative social eects including negative impacts
on the victim’s health that spill over to the community. The victimised children suer trauma and
other psychological problems which often leads to them dropping out of school. Violence also
contributes to juvenile delinquency with children who run away from home ending up in the streets.
91 FDG, Bondeni Women.
92 Interview with representative of NGO working on resource conflict in Nakuru Town.
93 Interview with ocial, Nakuru County Transport Department, Nakuru Town
94 There is an ongoing debate over the use of the term FGM, as opposed to the more culturally-correct
Female circumcision. The respondents, however, referred to the practice as FGM.
40 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 41
Respondents indicated that children across all ages are victimised at home especially by parents
(and step parents), siblings and relatives. As noted by the Chief of Naivasha Town location:
“The defilement of children (both boys and girls)
rising, particularly by closer members of the family.”95
It was also noted that children are sometimes vulnerable to sexual violence by neighbours
and other adults because their parents neglect them and are not keen on them.96 Others also
pointed out that some mothers expose their children to sexual violence when they are involved
in prostitution and bring their clients home.97 Respondents also pointed out that some parents
and guardians exploit their children for economic gain and. A key informant pointed out that:
“Children are exposed to sexuality at a young
age, sometimes pressured to enter into
prostitution by parents to help make ends meet.”
Outside the home, the perpetrators of violence against children are also known to the children. They
often include teachers at school and members of religious groups. A respondent revealed that:
“Murder of children is especially high in Karagita;
five children went missing and later found
dead. Church leaders and other grownups are
implicated in this. This is common in Karagita
and Kihoto.”98
Figure 4: Prevalence of child abuse
0% 20% 40% 60% 80% 100%
Don't knowRefuse to answerNot a problemProblem
Total
Molo
Naivasha
Nakuru
How prevalemt is the problem of child abuse in your area?
50%
44%
66%
54%
31% 18%
1%
4%
3%
2%
25% 27%
7%
17%
24%
27%
95 Inter view with senior national administration ocial, Naivasha Town Location.
96 Interview with nurse, Naivasha District Hospital.
97 Focus Group Discussion with women in Bondeni
98 Interview with Director of NGO working on Conflict management in Naivasha
42 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 43
A key dynamic of this type of violence to note is that, it can occur over long periods of time
unlike the other types of violence. Children can be victimised repeatedly before the abuse
becomes a matter of concern. There are two possible reasons for this. First, physical abuse
of children is sometimes seen as permissible by the society. It’s often seen as a part of the
‘disciplining’ process necessary for proper development of the children, which has to be
carried out by the responsible adults (parents, teachers or relatives). Secondly, children rarely
report such cases, which can partly be explained by their dependence on parents. This will
often lead them to fear further victimisation for reporting. As such, these issues come into the
fore only in extreme circumstances. Some of the respondents noted cases of children being
denied food for long periods of time.
When asked what they perceived the root causes of such violence to be, more than 70%
of the respondents attributed it to drug and alcohol abuse by parents and about 60% on
poor parenting. Others identified poverty which, they claim, resulted in stress and tensions
in the home, often leading parents to treat children more harshly and possibly violently than
under less stressful circumstances. As stated by a police ocer in Nakuru, “Most cases of
violence against children is economically related. Either because of the stress experienced
by the parents in eking a living unleashing this stress on their children or children left on their
own as parents look for resources leaving them vulnerable to violence in the neighborhood.
There are not too many reported cases of economically endowed persons meting violence
on their children.”99
2.17.3 Violent Crime
For analytical ease, we have categorised several types of violence including theft, robbery,
mugging, kidnappings, extortion and car-jackings under the tag of ‘violent crime’. Aviolent
crime is when someone physically hurts, with or without a weapon, or threatens to hurt
someone.100 These violent crimes occur on a regular, even daily basis across the country, a
pattern replicated in Nakuru County.
Violent crime aects people from all walks of life whether they are motorists, tourists or
businessmen. For instance, respondents to this study noted that gangs and cartels extort
matatu operators regularly. Furthermore, several key informants noted that violent crime has
negative eects on economic growth and development of an area since insecurity discourages
investment.
The perpetrators of violent crime are as varied as their victims. Nonetheless, most of the
respondents to the study noted that gangs are the main perpetrators of violent crime in
Nakuru County. This was especially highlighted in Bondeni area, Kivumbini and Kaptembwo.
Respondents noted that the dierent areas are controlled by dierent gangs, which often have
some members roaming the streets to identify potential targets.101 Some of the respondents
noted that these gangs often operate in collusion with street children, rogue police ocers
and matatu operators who provide them with information on the movement of their targets.102
99 Interview with Police Ocer in GBV Desk, Nakuru Town.
100 https://ww w.victimsupport.org.uk/crime-info/types-crime/violent-crime.
101 Interview with chairman of a Matatu Sacco in Nakuru Town.
102 FGD with CBOs and CSOs in Bondeni.
42 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 43
The Chair of a matatu SACCO103 in Nakuru Town, for instance revealed that:
“Since there is no clear framework or criteria for
employment in this sector, many criminals join in and
corrupt and make the sector criminal and violent;
many crew member work in league with other
criminals, robbers, drug dealers, rapists, Police.”
Furthermore, human rights groups have reported that:
“Some police have been assisting the criminals
with weapons and they share the loot.”104
When the respondents were asked to identify the main causes of this type of violence, they
tended to fall back on the same standard arguments. About 80 % identified poverty and
unemployment (93 % in Bondeni area in Nakuru), while 65 % attributed it to alcohol and drug
abuse. However, explaining the causes of this type of violence away as caused by socio-
economic factors that are beyond the control of the perpetrators amounts to denying them
agency which reduces our ability to understand the complexities of this type of violence. This
is precisely what Moser and McIlwane warn us against. In fact, it is easy to point out that not
all young unemployed and poor young men turn to crime.
Notably however, the respondents noted two potent explanations for the prevalence of crime
in Nakuru County. The first was the availability of small arms, which embolden criminals
during robberies, muggings and car jackings. The arms are said to come from neighbouring
Baringo County,105 where they are commonly used by cattle rustlers. The second, emphasized
at FGDs in Nakuru Town, is corruption in the Police force, where rogue ocers are accused
of colluding with criminals – specially with gangs such as Confirm in Bondeni and others
in Kivumbini and Kaptembwo areas - to provide information about potential victims.106
Figure 5: How prevalent is violent crime in your area?
0% 10% 20% 30% 40% 50% 60% 70% 80%
Total
Molo Town
Naivasha Kabati
Naivasha Karagita
Nakuru Bondeni Village
Nakuru Kaptembwo
0 20 40 60 80 100
Others
Inequality between gender and social groups
Availability of arms
Poor quality of education
Ethnicity
Politics
Cultural norms that support violence
Lack of street lighting
Poor rule of law
High crime levels
Local illicit drug trade
Poverty
Unemployment
Most importnat causes of violence in your opinion?
How prevalent is violent crime in your area?
27%
27%
55%
70%
36%
39%
40%
45%
103 Savings and Credit Organization.
104 Interviews, Nakuru Human Rights Defenders Network.
105 Interview with regional coordinator of an FBO working in Nakuru Count y
106 Focus Group Discussion, Kaptembwo Men.
44 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 45
2.17.4 Police violence
Evidently, Nakuru County is no exception to the prevalence of police violence that has been
noted to be prevalent across the country. Police in Kenya are documented to misuse their
power to harass and threaten citizens, which has resulted in severe distrust of the Police by
the public. In some cases, the police go to the extreme of summary executions, extra-judicial
killings and enforced disappearances. Respondents to our study noted that the police often
intimidate and extort money from matatu operators, businessmen and hawkers. This is often
done through threats to charge the victims with unsubstantiated oences. Those who fail to
give in to the demands are usually locked up arbitrarily or even physically harmed. The head of
a matatu SACCO in Nakuru explained that:
“The police are also involved in extorting money
from and intimidating matatu crews; those
who fail to give bribes are locked up arbitrarily.
At night you have regular (not trac) police
arresting conductors who are carrying the day’s
money, most times they may be roughed up, not
necessarily beaten, but will be locked up” 107,
As highlighted above, some of the respondents from Kaptembwo and Bondeni in Nakuru
municipality, noted that the police collude with gang members in identifying victims, attacking
and covering up for the gangs whenever the cases were reported. The respondents also
highlighted that police violence is also witnessed during public protests and demonstrations.
Although these are usually peaceful, the police are known to stop them with disproportionate
force, using tear gas, beating and locking up protesters.
40 % of respondents felt that the Police use unnecessary violence/force when responding to
incidents of violence. However, this perception varied according to the type of violence that the
Police were responding to. While more than 70 % were against Police using violence in cases
of murder, break-ins, rape, theft and mugging, the change was significant in the case of armed
robbery as more than 40 % felt that the use of violence by the Police was justified. It is likely
that more people justify police violence in cases of armed robbery rather than murder, because
most people have more of a direct experience with violent robbery, than they have with murder.
Respondents identified the culture of corruption and impunity as the main reasons for the
prevalence of police violence. They noted that this environment only serves to encourage these
law enforcement ocers to engage in abuse. They cited a sense of entitlement by the police, as
the latter often felt they had a right to extort civilians, even turning violent in the process. This
was especially the case in Nakuru Town, where 60 % of the respondents expressed mistrust of
the Police. Others attributed Police violence and executions to frustrations among the ocers
by the Judiciary giving bail to repeat oenders who, often seek revenge against the Police.108
107 Interview with Chair of Matatu SACCO, Nakuru; and FGDs by CSOs and CBOs in Nakuru Municipality.
108 Interview with Executive Director of an NGO working on issues of Internally Displaced Persons in
Nakuru.
44 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 45
Figure 6: Use of violence by Police in response to incidents of violence
0% 20% 40% 60% 80%
YesNo
Murder
Armed robbery
Break-ins
Rape
Assaults
Theft and mugging
Unsafe
27%
24% 26%
29%
18%
23%
22%
36%
Are the police justified to use violence in these cases?
71%
29%
79%
21%
70%
30%
77%
23%
58%
42%
81%
29%
2.17.5 Political and ethnic violence
Nakuru County is one of the hotspots of inter-political and ethnic violence in Kenya. The county
has witnessed several bouts of inter-political and ethnic violence since the early 1990s.109 The
wounds from this violence are yet to heal, respondents in this study expressed concern that the
factors leading to the political violence that rocked the country in 2007/8 are still present, in an
environment clouded in poor inter-ethnic relations and lack of tolerance.
34 % of respondents in the present study identified inter-political and ethnic violence as a major
type of violence in the County. This was even higher in specific areas: 45 % in Karagita (Naivasha
Municipality) and 44 % in Kaptembwo area (Nakuru Municipality).
There are material dimensions to the inter-political and ethnic violence that has been witnessed
in the county. Respondents identified conflicts over natural resources, particularly land – in Molo
Town - and fish landing beaches – in Naivasha. In the case of land in Molo Town, for instance,
respondents noted claims of double registration of land titles especially in the Kibagundi area.
They traced this tension back to the 1992 political violence, which led to the displacement of
some members of the Kikuyu community and the occupation of their land by the Kalenjin to
date. Some respondents highlighted ongoing favouritism of certain communities by the County
government. Men in an FGD in Bondeni pointed out that:
109 For further reading on electoral violence see, for instance, Human Rights Watch, Divide and Rule:
State-Sponsored Ethnic Violence in Kenya (New York: Human Rights Watch, 1993); Mwangi Kagwanja,
Killing the Vote: State Sponsored Violence and Flawed Elections in Kenya (Nairobi, Kenya Human
Rights Commission, 1998); Human Rights Watch, Ballots to Bullets: Organized Violence and Kenya’s
Crisis of Governance, (New York: Human Rights Watch, 2008; Kenya National Commission on Hu-
man Rights, On the Brink of the Precipice: A Human Rights Account of Kenya’s Post-2007 Election
Violence, Nairobi: The Commission, 2008.
46 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 47
“Ethnic tension and violence are linked to political violence such that there is a feeling among
certain communities that certain political actors favour certain ethnic communities in Bondeni
when it comes to distribution of resources.”
Others also noted that there are often ethnic dimensions when it comes to providing job
opportunities and public projects, adding that it fuelled ethnic tension.110 There were some
respondents who highlighted the marginalisation of certain communities by the County
government in respect of jobs and other opportunities. A respondent in Naivasha identified that:
“Political and ethnic violence is caused by discrimination against certain ethnic groups. Some
tribes are not given jobs in flower farm companies.”111
Figure 7: Prevalence of political and ethnic violence
0% 10% 20% 30% 40% 50%
Nakuru Bondeni
Naivasha Kabati
Molo Town
Nakuru Kaptembwo
Naivasha Karagita
40% 60% 80% 1,0
Are the police justified to use violence in these cases?
How prevalent is political and ethnic violence in yout area?
71%
45%
44%
37%
11%
4%
70%
58%
42%
The other material dimension of this conflict is when violent crimes are mapped on to ethnic
cleavages. For instance, our respondents identified theft of livestock, and break-ins into shops
as often resulting in inter-ethnic conflicts in Molo. Since the two main communities in Molo
Town, Kalenjin and Kikuyu, are generally divided along distinct lines of economic activities with
the latter tend to be businessmen and the former animal and crop farmers, theft cases often
take an ethnic dimension.
Some of the respondents identified politicians as the main perpetrators of acts of inter-political
and ethnic violence as they often mobilize support on ethnic lines and fund gangs to execute
such acts. Others noted that the elders of the various ethnic communities also serve to instil
ethnic animosity in young people. This serves to shape perception of the other, occasioning
a situation where behaviour is mapped into ethnic stereotypes. As such, young people seek
110 Interview with a Community Mobiliser fand DPC member, in Molo Town.
111 Inter view with Director of an NGO working in Nakuru Town .
46 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 47
vengeance for the loss of loved ones or property against a community escalating inter-ethnic
tensions and violence. People expressed concern on the risk of violence erupting with regards
to the forthcoming elections, citing possible major cases as incitement by politicians (64%),
politicization of ethnicity (69%), verbal assault towards other ethnic groups (54%) and rumours
about vote rigging (56%).
2.18
This section highlights the responses taken at individual, community and State levels to counter
each of the types violence that were found to be prevalent in Nakuru County. The section adopts
the three-level mitigation framework discussed earlier in the paper, examining the actions taken
by individuals, the community and the State in preventing and countering violence.
2.19
Any intervention aimed at reducing violence will depend on a good understanding of the
organisation of the community. This involves noting the types and number of groups that people
are members of as groups are often the first source of intervention in case of violence. The survey
revealed that the groups that attracted the highest membership were religious groups (34% of
respondents), men, women and youth groups (34% of respondents), community associations
(10% of respondents) and sports groups (7% of respondents). None of the respondents indicated
being involved in the Nyumba Kumi initiative. It also revealed interesting insights on age dierence
in group membership. Most of those who indicated that they were members of the religious
groups and community associations were 45 years and older. Young people represented the
highest category among those who did not belong to any group, though they also made up a
bulk of the membership in sports groups. Nonetheless, even their participation in these sports
groups is limited with only 12% of 18-24 and 7% of 25-34 indicating that they were members of
these groups.
The study also sought to identify where the residents of Nakuru had sought assistance over
the previous twelve months. The study revealed that Men, Women and Youth groups as well as
religious groups were the most important source of assistance and information for the residents
of Nakuru County. One-in-four of the respondents had received assistance from their men,
women or youth groups, with women (28 %) benefiting more than men (21 %). Most of the
beneficiaries were those aged 45-54 (36 %). An almost similar number (23%) of the respondents
had received support from religious organisations, with women (26%) still benefiting more than
men (20 %). The older groups over 55 (37 %) and 45-54 (31 %) benefitted more from the religious
groups than the younger ones 18-24 (21 %) and 25-34 and 35-44 both standing at 22 %).
About a third of the respondents (34 %), indicated that they had received support from their
family. Friends (30 %) and neighbours (30%) were also noted as significant sources of assistance
for Nakuru residents. The importance of family, friends and neighbours as sources of support
run across both genders and across the county.
48 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 49
2.20
Many respondents noted that neighbours (35 %) and family (19 %) came to their rescue whenever
they experienced violence, while only 9 % acknowledged having received Police assistance.
This highlights the fact that people rely more on their neighbours and families when they are
victimised. The variance between neighbours and family response may be due to intervention
of neighbours during cases of domestic violence.
The study also sought to identify what mechanisms the people trust to address the challenge of
violence. The most trusted – as shown in the figure 9 – are medical doctors (71 %),112 religious
leaders (70 %), elders (69 %) and NGOs (62 %). The Police were identified as the least trusted (51 %).
Figure 9: Level of trust in violence prevention actors/mechanisms
Not Trustworthy
Trustworthy
How much do you trust them?
Police
Community Policing
Court Judges
Medical Doctors
Local Church/Priests
Nyumba Kumi
Chiefs
Elders
NGOs
40% 60% 80% 1,0
62%
69%
64%
57%
70%
71%
58%
60%
51% 49%
40%
42%
29%
30%
43%
36%
31%
38%
42%
As noted earlier, Kenyans appear to have a lot more trust in the police than in other institutions within
the criminal justice system.113 Notably, many of them report incidences to the police, expecting
them to resolve the issues rather than taking the matters to court. As such, the police clearly form an
important part of the acceptable response mechanisms to violence. Local Chiefs were also noted as
key actors in response to violence as shown in the figure below. Further research would be helpful to
establish exactly what sort of crime and violent incidences that residents of these neighbourhoods
report to the police.114
112 This is probably mostly relating to Sexual and Gender Based Violence as well as Violence against
children
113 See Wairuri, 2015; ibid.
114 Extant literature suggests that people will report major incidences such a infant rape, violence rober
48 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 49
Figure 10: Reporting violence incidents
0% 10% 20% 30% 40% 50%
Local NGOs
GBV units at health facilities
Local churches/religious leaders
Nyumba Kumi
Local elders and Headman
Local council
Chiefs
Police
Where did you report the incidents of violence?
80% 1,0
50%
15%
6%
5%
2%
2%
2%
1%
Unsurprisingly, religious leaders who are highly trusted by the members of the community were
also identified as the most eective in dealing with violence.115 Local leaders and chiefs who
were also highly trusted by members of the community were also seen as quite eective. This
indicates that the local communities have a high preference for local, grounded institutions to
address their challenges with crime and violence. Interestingly, the community is split almost
halfway in their opinion of the police both in terms of trustworthiness and eectiveness. Proper
appreciation of the nuances of how residents interact with the police would help to improve
understanding of these views deeper. Interestingly, while 62 % of the respondents found NGOs
trustworthy a slightly lower number (55 %) found them eective in dealing with issues of crime
and violence.
Figure 11: Eectiveness of violence intervention actors/mechanisms
How eective are they handling violence
49%
40%
42%
29%
30%
43%
36%
31%
0% 20% 40% 60% 80% 100%
Not eectiveEective
Community leaders
NGOs
County Government
Religious leaders
Police
Local Elders
Chiefs 65% 35%
33%
44%
28%
52%
45%
34%
67%
56%
72%
48%
55%
66%
115 Though it is not clear for what type of violence.
50 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 51
On the reduction of violence, respondents proposed various response mechanisms. 75 %
proposed community policing as the main mechanism at community level. At State level, 62
% proposed government-driven interventions such as alcohol and substance control, Police
control and street lighting. Within the family, people felt that education (62 %), parental support
(61 %) and proper living conditions (60 %) are important for preventing violence.
Figure 12: Mechanisms for violence reduction
10% 20% 30% 40% 50%
How do you think violence could be reduced in your community?
0% 10% 20% 30% 40% 50% 60% 70% 80%
Alcohol and substance control
Community policing
Government driven interventions
Information materials/campaigns
Police control
Street lighting 59%
54%
46%
62%
75%
76%
2.20.1 Effectiveness of Current responses to SGBV
Most of our respondents, confirmed expectations from the literature that victims do not always
report SGBV cases. They noted that the victims, mainly women and children fail to report these cases
for fear of being kicked out of their homes or victimised/stigmatised by neighbours and relatives.
Some individuals in Naivasha also failed to report incidents of SGBV due to influence from religious
leaders (pastors of certain churches), who asked their congregations to forgive perpetrators.116
Our study found several mechanisms, which the communities have adopted to deal with the
issue of Sexual and Gender Based Violence. Most notably, there are eorts by several Community
Based Organisations (CBOs) and Civil Society Organisations (CSOs), dedicated to tackling SGBV,
particularly in Naivasha. Some of their eorts include creating awareness in schools, setting
up safe houses and oering psychosocial support to victims of SGBV. Young Men Champions
in Ending GBV (YMCEGBV) in Naivasha has a program, which creates awareness through
mentorship programs in schools and dialogue with girls on areas of SGBV. However, residents
expressed concern that there are not enough mechanisms to protect abused women and
children. According to a police ocer in Kaptembwo:
“Women and children [are the most common victims
of violence in the community]. They are abused
and chased from home but there are no shelters for
abused women. There are no places to go.”
116 Interview with Chairperson of a youth group in Naivasha.
50 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 51
The respondents noted several other school programs aimed at sensitizing children about
defilement and other forms of abuse and encouraging them to report as soon as they witness
or experience this. They noted that these initiatives are eective as children are more aware.
Some even noted that the children are now more likely to report cases of violence to the
Police, teachers and other adults. They conduct forums, dubbed Kahawa Forums to which
they invite senior police ocers to discuss ways of mitigating issues of security. In partnership
with the Federation of Women Lawayers of Kenya (FIDA), they have also started programs to
help victims of violence seek proper legal procedures and follow up of cases in the court.
There is also Amani House, a safe house for GBV victims which has been set up by the Agatha
Imani House organisation. Additionally, the Naivasha Gender Based Violence Network, which
comprises ordinary citizens, hospital workers and gender-desk ocers from the Police, oers
psychosocial support and other response services to victims of SGBV.
Respondents showed support for the use of mob ‘justice’ as a response to incidents of
sexual violence. In the case of rape, for instance, 66 % felt it was justified to resort to mob
‘justice’. Most of the respondents expressed their dissatisfaction with the traditional response
mechanisms by the state ranging from intervention by Chiefs at the local level, especially
for domestic violence cases, and arrest and prosecution of perpetrators by the Police. They
accused the Chiefs and the police of ineectiveness and corruption. They highlighted cases
where perpetrators of SGBV were released after they allegedly bribed authorities.
The respondents also discussed the new approaches developed by the government in
attempt to address insecurity, including the Nyumba Kumi initiative, a form of community
policing. The respondents noted that the Nyumba Kumi had been applied across all areas
under study, but with dierent levels of success. Whereas respondents in Molo Town had
a positive outlook of the initiative, those in Naivasha and Nakuru mostly claimed it was
ineective. This was particularly notable in Bondeni and Kaptembwo areas where more
than 50 % of respondents felt that the Nyumba Kumi committees were untrustworthy. In
the latter, some respondents accused certain members of Nyumba Kumi committees of
colluding to protect perpetrators of violence that they knew and were related to. However,
in some parts of Molo Town, the initiative seems eective as these committees have been
instrumental in stopping cases of domestic violence.
Finally, the County government has also developed its own mechanisms to help tackle the
challenges related to SGBV. Most notably, they have established and equipped a full wing for
handling SGBV cases at the Provincial hospital to provide treatment and counselling services
for SGBV victims. The County government has also established sexual violence committees
at sub county level bringing together various government ministries and ocials to explore
further ways to tackle SGBV. Another notable eort is the establishment of a gender desk in
the Police service, which, in addition to investigating and following up cases to the end, also
attends community awareness forums.
52 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 53
2.20.2 Effectiveness of responses to Violence against children
At the Individual level, the study found that most children often do not report cases of violence.
Many of these cases of abuse are only discovered after a long period by neighbours or relatives,
or when they are highly dramatic in nature, such a when a child gets burnt. This failure to report
is underpinned by fear of being harmed further by the perpetrators of the violence upon whom
the children often depend on - mostly parents, guardians, neighbours or teachers.
Respondents highlighted that the community level interventions are mainly carried out by
NGOs, but were nonetheless noted to have much less interest than the SGBV, as highlighted
above. They noted that the Mid-Rift Human Rights network is working to raise awareness
across the county and provide training on community policing.
At the State level, the respondents noted that the police responses take the traditional
approach of rescuing victims of such violence and taking them to Children’s Homes and
prosecution of perpetrators. The challenge, however, is that most cases are not reported,
given that they occur within the privacy of a home or school. The other challenge, which
is particularly, more pertinent is that the children may not want to be separated from their
parents. As such the legal approach of charging the parents and taking the children away
may not be fully acceptable. Perhaps, this is one of the reasons why many such cases are
not reported. Expressing the inadequacies of mechanisms to protect abused children, a
police ocer said:
“Children who are abused at home are
sometimes rescued by police and taken to
children’s homes; some are taken to borstal
schools even though they are not delinquent,
the fear is that they will run from normal
homes”117
2.20.3 Effectiveness of Current responses to Criminal violence
The survey revealed that 78 % of the population in Nakuru County do not report incidents of
crime witnessed or experienced and of the 22 % who do, only half report to the Police and
a few to the Chiefs.
The respondents to the study noted that individuals have come up with their own coping
mechanisms to deal with the risk of victimisation. In response to the common theft of motor
bikes from Boda Boda operators, the latter avoid taking clients after dark and to well-known
crime spots. As is the case elsewhere in the country, some businessmen and citizens across
the selected study sites often hire private security services (from private firms or the Police)118
at their business premises and homes.
117 Interview with police ocer on GBV Desk, Kaptembwo
118 Interview with Businessmen in Naivasha Town.
52 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 53
At the community level, several mechanisms for responding to insecurity have been
developed. For instance, due to the common theft of motor bikes, the Boda Boda operators
have come up with a system in which they escort one another whenever working late or
maintain contact with colleagues - as a tracking system - while taking clients to presumable
violence-prone destinations. Also, a local non-state actor in Nakuru Town developed Mulika
Uhalifu, a mobile phone program, which enables individuals to report cases of violence by
sending a simple message to a given number that is linked to various Police station heads.
The respondents acknowledged that the program has been eective in increasing reporting
and Police interventions119. Another mechanism has been the growth of Matatu SACCOS,
which have helped to improve the financial stability of their members. The respondents
noted that this has created meaningful opportunities, which have kept the youth away from
crime. As one respondent attested:
“the matatu SACCOs … brought in a
better business framework and ethic.” The
respondents noted that the SACCOs make
their own investments which oer a return that
benefit members through dividends and low
interest loans.”
Community members also resorted to ‘mob justice’ in dealing with cases of violence, though
this was often a spontaneous response. Interestingly, whereas 80 % of respondents in the survey
were against mob justice in general, more than 75 % claimed it was justified in responding to
cases of murder or armed robbery and over 65 % justifying the act in cases theft and mugging.
At the state level, the respondents commented on eorts of the police including new
initiatives such as Community Policing and Nyumba Kumi at the national level. Respondents
also observed that the Police force increased its patrols in the central business district and
residential neighbourhoods police arrests and prosecution of perpetrators has been on
the rise and in Nakuru Town. There is also more frequent rotation of ocers to limit their
opportunities for engaging in corruption and colluding with criminals and gangs.
Respondents from Naivasha, noted that the Community Policing program, in which 64 %
of respondents expressed trust, has been useful in improving the timeliness of the response
of police ocers to public reports of violence. They noted that the Community policing
committee meets regularly, sometimes with Police ocers, and present evaluation reports on
what issues were reported and how they were dealt with. The residents of Molo Town, expressed
satisfaction with the Nyumba Kumi initiative. They observe that they conduct neighbourhood
patrols and surveillance mounted which have resulted in successful investigations and arrest
of cattle thieves.
They also noted eorts by the County government, including street lighting project in towns
that have made public spaces safer at night. Respondents in Nakuru Town, however, noted that
the initiative is being undermined by gangs who have been vandalizing some of the lights120.
119 Interview with senior county administration ocial, Nakuru.
120 Noted in FGD for Men in Bondeni area.
54 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 55
Figure 13: Use of mob ‘justice’ in response to violence
60% 80% 1,0
Is it justified to use mob justice in these cases?
71%
70%
58%
0% 10% 20% 30% 40% 50% 60% 70% 80%
Yes
No
Murder
Armed robbery
Break-ins
Assaults
Theft and mugging
26%
68%
60%
33%
33%
62%
20%
75%
20%
78%
As the figure above shows, a large percentage of people believe that it is justified for the
public to engage in mob violence against criminals as a way of addressing crime. This mirrors
the broader national pattern of support for mob violence, which is partly the result of the
public’s frustration with ineective responses to crime by the police. Moreover, individuals
involved in mob violence are rarely arrested and punished, and this engenders impunity and
the perception that it is justified for the public to take the law into their own hands.
2.20.4 Effectiveness of Current responses to Police violence
Key informants cited Nakuru as one of the counties trying very hard to improve police-citizen
relations and interactions.121 Most respondents, across the County, indicated that citizens
are afraid of reporting cases of Police violence for fear of becoming targets themselves,
especially by what they termed as ‘rogue ocers’. This fear is undergirded by the perception
of corruption and impunity by the Police. Respondents noted that the police are known to
have links with perpetrators of violence.
At the community level, it is clear that NGOs have been active in campaigning against Police
violence for a long time within the county and are trusted by more than 60 % of respondents.
The Mid-Rift Human Rights Network for instance, has helped people know about police
work and their rights vis-à-vis the police. Several key informants, particularly those working
in civil society organizations, noted that these kinds of trainings have resulted in reduction
of police violence and brutality in the dealings with citizens.
Respondents noted the state level interventions, including the frequent rotation of ocers to
limit their opportunities for engaging in corruption and colluding with criminals and gangs.
121 Interview with Regional Coordinator for FBO in Nakuru
54 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 55
2.20.5 Effectiveness of Current responses to Political and ethnic violence
At the individual level, the respondents expressed the importance for individuals to be more
tolerant towards members of other ethnic groups or those with dierent political opinions.
However, given the complexity of political violence, they did not provide concrete actions
they took to counter this form of violence.
At the Community level, however, several eorts by NGOs and Faith Based Organisations
were noted. Organisations such as Mercy Corps, USAID and Peace Caravan, for instance,
have played a major role in reducing instances of political and ethnic violence in Molo by
collaborating with community leaders to foster peace and harmony. Interviewees and focus
group participants noted that these eorts have been eective. They also noted the eorts
of the NCCK in terms of oering counselling and spiritual support to couples aected by
political violence and also getting\ scholarships for victims who are not able to acquire
education. Respondents also highlighted Peace Circles, organised by the Catholic Church in
Bondeni to promote peace. The Peace Circles bring together small groups of about twenty
people to discuss matters of peace and social cohesion. This program initially started to
address post-election violence and inter-ethnic tensions in 2007/8 but has continued.
Respondents highlighted civic education to the public through Barazas to promote
understanding and tolerance carried out by Ward administrators in Nakuru Municipality.
Some also acknowledged the existence of Peace Committees but complained that these
lacked sucient financial support from the National and County Governments. Peace
Committees are reportedly more eective in rural areas than urban; and they are said to
work closely with CSOs and churches in identifying and planning interventions. These
include peace related events, dialogues and mediation between rival communities, business
groups and families involved in violence122.
122 Interview with members and leaders of Peace Committees for Nakuru East and West Sub-counties
and County-wide.
56 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 57
3. Conclusion and recommendations
This study has demonstrated that urban violence is an issue of concern in Nakuru County,
particularly in the poor urban neighbourhoods. This study has shed more light on this phenomenon
particularly on the prevalent types of violence and their causes, the perpetrators and victims of the
various types of violence and the eectiveness of the current response mechanisms.
Evidently, therefore there are gaps in the current response mechanisms. We have noted these gaps in
the study including impunity of perpetrators, a lack of trust in the mechanisms by the victims of violence,
inadequate understanding of what constitutes violence and crime amongst the populace (for instance with
respect to violence against children), and the failure of victims and witnesses to report cases of violence to
the authorities (Police, Chiefs) due to fear of further victimisation. Notably, however, the study also reveals
important dynamics with respect to the structure of the communities we have studied, victimisation and
response mechanisms that are useful in developing recommendations for programmatic intervention.
This section puts forward recommendations that will help fill the identified gaps and strengthen the
existing response mechanisms, for the dierent types of violence discussed. The recommendations
will facilitate better programming for tackling urban violence in Nakuru County. The focus of
the recommendations here is primarily on prevention of violence with limited attention paid to
interventions after violence has occurred. The proposed interventions are largely based on how to
prevent violence by improving the awareness and understanding of the dynamics of violence and how
they should be handled amongst the community and hence empowering the community to deal with
violence. The view here is on empowering the community to use the existing mechanisms to address
violence. There is also ned to focus on building on the community trust in the existing institutions such
as community organisations, police, chiefs, elders and religious leaders. The interventions will also
include comments on reporting of violence within the formal criminal justice system.
56 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 57
i Sexual and Gender-Based Violence (SGBV)
Future programs seeking to address the challenge of SGBV need to take proper cognisance of
the community dynamics of this type of violence including the causal and/or trigger factors and
the eectiveness of the current response mechanisms. The study noted that the most prevalent
forms of this violence include domestic violence, rape, threats and acts of intimidation that mainly
targets women and girls. Our respondents identified alcohol and drug abuse, infidelity, material
conditions within the family. From the conceptual framework, it is important for us to note that
these perspectives however take away the agency of the victims and perpetrators. For instance, it
should be clear that while alcohol contributes to domestic violence, not all people who use alcohol
are violent. As such, the focus should be in understanding the underlying factors.
The other key finding with respect to this kind of violence is that victims of SGBV do not always
report cases for fear of being further victimised or stigmatised by neighbours and relatives. The
study found several mechanisms adopted by the communities to deal with SGBV including creating
awareness in schools, setting up safe houses, oering psychosocial support to victims of SGBV
and helping victims to seek legal redress. Nonetheless, respondents expressed their dissatisfaction
with the traditional state response mechanisms including the police and chiefs accusing them of
corruption. As a result, many of them expressed support for mob ‘justice’ as a way of dealing with
sexual violence. The County Government of Nakuru has also developed its own mechanisms to deal
with SGBV including establishing sexual violence committees at sub county level bringing together
various government ministries and ocials to explore further ways to tackle SGBV.
Recommendations for programmatic interventions:
i Supporting the expansion of existing awareness creation and empowerment
programs beyond schools to the entire community, especially women and
girls to understand the dynamics of violence and how to respond in case of
victimisation but also to help men and boys understand what constitutes SGBV
ii Improving the incorporating of the most trusted actors in the communities
in the eorts of creating awareness and empowerment related to SGBV. In
this regard, programs to prevent SGBV cases should include doctors, religious
leaders and the police.
iii Support further dissemination of information on the legal procedures relating
to sexual violence including the preservation of evidence to facilitate arrest and
successful prosecution. The trusted institutions within the community such as
religious leaders and elders should be equipped with this knowledge as they
could be first points of contact of victims or witnesses of violence.
58 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 59
ii Violence against children
Violence against Children is another predominant form of violence in Nakuru with 54% of
our respondents identifying it as a problem. The study found that parents and teachers
are the main perpetrators of violence against children though fellow students were also
found to be perpetrators of this violence. The violence takes the form of beating, corporal
punishment and bullying. It is important to bear in mind that most children are at the risk
of repeated victimisation as this is often seen as an acceptable way to discipline children.
Another important finding is that children rarely report such cases, partly because they may
also see it as normal and also for fear of further victimisation. As such, these issues come
into the fore only in extreme circumstances. The respondents noted that drug and alcohol
abuse by parents and poor parenting were to blame. It was evident that this is a neglected
area even though some NGOs such as the Mid-Rift Human Rights network are working to
raise awareness across the county on this type of violence. The study also noted that the
traditional state response of rescuing victims from the parents and prosecuting the parents
may not be eective as the children may not want to be separated from their parents.
Recommendations for programmatic interventions:
i Support programs carried out by trusted local institutions such as religious
leaders that raise awareness about the rights of children and the negative eects
of violence against children. Such programs should also include education on
parenting skills and other methods of disciplining children. Bringing doctors to
speak to parent groups about the long-term eects of violence against children
could strengthen such programs.
ii Develop school-based programs to empower children to be able to report on
cases of violence and abuse occurring to them or to their friends at home, in
the neighbourhood or in school including how to protect themselves from
victimisation
iii Support local level mechanisms of resolving minor cases of violence against
children rather than focussing on arrest and prosecution of perpetrators.
Interventions by religious leaders, elders and the local chief could be more eective
in reducing this kind of violence while keeping families intact, as much as possible.
The incarceration of a parent or guardian may have further negative eects on
families.
iv Provide support to existing child support centres and homes for victims of this type
of violence or setting up new ones in partnership with the communities to deal
with the most extreme cases.
58 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 59
iii Violent crime
The study found violent crime to be a major component of violence in Nakuru County. Our respondents
indicated that violent crimes occur on a regular, even daily basis across the county aecting people
from all walks of life. The respondents further noted that gangs are the main perpetrators of violent
crime in Nakuru County, especially in Bondeni area, Kivumbini and Kaptembwo. Respondents noted
poverty and unemployment, and alcohol and drug abuse as the main drivers of this type of violence.
However, explaining the causes of this type of violence away as caused by socio-economic factors
that are beyond the control of the perpetrators amounts to denying them agency which reduces
our ability to understand the complexities of this type of violence. But the most potent explanations
given include availability of small arms and the corruption of police ocers which embolden
criminals during robberies, muggings and car jackings. Due to this perception of police corruption,
78 % of the population in Nakuru County do not report incidents of crime witnessed or experience.
The respondents revealed that they have come up with their own coping mechanisms to deal with
the risk of victimisation including hiring private security or escorting each other to prevent attacks.
Another mechanism that was highlighted is the Mulika Uhalifu, a mobile phone program, which
enables individuals to report cases of violence by sending a simple message to a given number
that is linked to various Police station heads. Respondents also noted that the growth of Matatu
Saccos which have provided livelihoods for most of the youth have contributed to a reduction in
crime. Another local level mechanism they noted is ‘mob justice’. State driven mechanisms such
as Nyumba Kumi and Community Policing were evaluated dierently in various areas. Community
policing was most popular in Naivasha while Nyumba Kumi was evaluated positively in Molo Town.
Most people expressed support for the street lighting project by the County government.
Recommendations for programmatic interventions:
i Strengthening of existing community level interventions of dealing with violent
crime such as the Nyumba Kumi and Community Policing initiatives. This requires a
more fine-grained analysis of why they work better in some places and not others
in order to replicate best practices across the county.
ii Support the scaling of the Mulika Uhalifu program or such similar initiatives across
the county to help in the reporting of violent crime
iii Lobby the government to enhance the existing programs of dealing with
proliferation of small arms and light weapons in the county
iv Work with the existing state mechanisms such as IPOA to enhance police
accountability
v Support the National Police Service, and other mechanisms, at the local level to
enhance patrols, investigate reported cases, arrest and prosecute perpetrators.
60 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 61
iv Police violence
Similarly, to the national level picture, police excesses and violence were found to be prevalent
in Nakuru County.123 Our respondents noted that the police often intimidate and extort money
from matatu operators124, businessmen and hawkers. Some of the respondents, especially
from Kaptembwo and Bondeni in Nakuru municipality, noted that the police collude with
criminals. 40 % of the respondents to the household survey indicated that the police use
unnecessary violence/force when responding to incidents of violence.
The study found that local NGOs have been active in campaigning against police violence for
a long time within the county. These institutions have a high level of trust in the community.
It was noted that some of these NGOs, such as the Mid-Rift Human Rights Network, has
helped people know about police work and their rights vis-à-vis the police. Respondents
noted that these kinds of trainings have resulted in reduction of police violence and brutality
in the dealings with citizens.
Respondents noted the state level interventions, including the frequent rotation of ocers to
limit their opportunities for engaging in corruption and colluding with criminals and gangs.
Notably, the respondents did not speak about the Independent Policing Oversight Authority
(IPOA).
Recommendations:
i Support the existing programs by local level NGOs to create awareness about
the rights of citizens when dealing with the police to increase demand for fair
treatment when engaging with the Police.
ii Work with state level institutions responsible for creating police accountability
including IPOA, NCAJ, KNHRC to help the residents better understand their roles
and how to seek recourse in case of abuse by police ocers
iii Support local NGOs to fight the impunity of police ocers by enabling them to
bring cases against rogue police ocers so that they can face the consequences of
their actions.
iv Support community partnerships with the police where such issues can be raised
and addressed.
123 See reports by IMLU, 2014; IPOA, 2013; see also reports by KHRC; Amnesty International, Human
Rights Watch etc.
124 Interview with Chair of Matatu SACCO, Nakuru; and FGDs by CSOs and CBOs in Nakuru Municipality.
60 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 61
v Political and ethnic violence
Nakuru is on the counties that have borne the blunt of politically-instigated political and ethnic
violence. This type of violence, most common during electoral periods, is the result of incitement
of ethnic animosity by political leaders. 34 % of the respondents to our survey identified this type
of violence as a major concern. This was even higher in specific areas: 45 % in Karagita (Naivasha
Municipality) and 44 % in Kaptembwo area (Nakuru Municipality. Politicians who were identified as
the main perpetrators of acts of inter-political and ethnic violence were said to mobilise support
on ethnic lines and fund gangs to execute such acts. The elders of the various ethnic communities
were also noted to instil ethnic animosity in young people. The respondents expressed concern
on the risk of violence erupting with regards to the forthcoming elections, citing possible major
cases as incitement by politicians (64%), politicization of ethnicity (69%), verbal assault towards
other ethnic groups (54%) and rumours about vote rigging (56%).
The respondents noted that several NGOs, such as Mercy Corps, USAID, Peace Caravan and Faith
Based Organisations such as the NCCK have played a major role in reducing instances of political
and ethnic violence in Molo by collaborating with community leaders to foster peace and harmony.
They noted Peace Circles, organised by the Catholic Church in Bondeni to promote peace, where
small groups of about twenty people to discuss matters of peace and social cohesion as eective.
Though they didn’t speak much of state level interventions. They noted the use of civic education
through Barazas to promote understanding and tolerance carried out by Ward administrators.
Some also acknowledged the existence of Peace Committees but complained that these lacked
sucient support from the Government. Peace Committees are reportedly more eective in rural
areas than urban; and they are said to work closely with CSOs and churches in identifying and
planning interventions. These include peace related events, dialogues and mediation between
rival communities, business groups and families involved in violence125.
Recommendations for dealing with this type of violence:
i Supporting the existing mechanisms developed by local NGOs and Faith Based
Organisations to enhance inter-ethnic cohesion
ii Lobby government institutions such as the NCIC, National Peace Committees to
have a broader and deeper presence in Nakuru County
iii Support and encourage the County Government of Nakuru to extend the Barazas on
understanding and tolerance beyond Nakuru Municipality to other areas of the County
iv Work with the County Government to develop inclusion policies that promote
equality of opportunity for all citizens in the County, regardless of ethnic
background.
v Support national level eorts, executed at the county level to address historical
injustice in the most inclusive manner. Most importantly, the resettlement of IDPs
displaced by politicised political and ethnic violence in previous elections.
vi Encourage community members to report cases of incitement by political leaders
and support the prosecution of such leaders in the community.
125 Interview with members and leaders of Peace Committees for Nakuru East and West Sub- counties
and County-wide.
62 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 63
vi Intersectoral violence prevention
All the above-mentioned types of violence, can be addressed by focusing work and interventions
towards the following:
Participation and Norm change on how to prevent violence perpetration from happening and
reducing risk factors for people becoming perpetrators, by participation and change of norms.
Trust-building between police and community, as recommended by the World Bank.
Leadership development for intersectoral coalition building should be strengthened.
Forward looking we see a need for more studies and research into data on perpetrators as
opposed to victim-based.
62 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 63
4. References
ACLED. (2013). Country Report: Kenya. Armed Conflict Location & Event Dataset, December, 1.
Akech, M. (2005). “Public Law Values and the Politics of Criminal (In) Justice: Creating a Democratic
Framework for Policing in Kenya”, Oxford University Commonwealth Law Journal, 225-256.
Alice. 2009. “Policing in Kenya: A Selective Service.” Chap. 11 in Policing Developing
Democracies, by M S Hinton and TNewburn.
Amnesty International. (2015). The State of the World’s Human Rights: Report 2014/15.
Anderson, D. (2014). Vigilantes, Violence and the Politics of Public Order in Kenya, African Aairs,
Auyero, J., de Lara, A. B., & Berti, M. FViolence and the State at the Urban Margins.
Anderson, D M. 2002. “Vigilantes, violence and the politics of public order in Kenya’.” African
Aairs 101 (405): 531-555.
Baker, B. 2009. Multi-Choice Policing in Africa, Journal of Modern African Studies, 47:01, 154-155.
Baker, B. (2006). ‘‘Beyond the State Police in Urban Uganda and Sierra Leone.’’ Africa
Spectrum, 41(1), 55-76.
Baker, B. (2004). Protection from crime: What is on oer for Africans.” Journal of Contemporary
African Studies, 22(2): 165-188.Journal of Contemporary Ethnography, 43(1), 94 - 116.
Fajnzylber, P., Lederman, D. & Loayza, N. (2002) ‘Inequality and Violent Crime’, Journal of Law
and Economics, XLV (April), pp. 1-40.
Hillier, D. (2007) ‘Africa’s Missing Billions: International arms flows and the cost of conflict’,
Oxfam briefing paper, http://policy-practice.oxfam.org.uk/publications/africas-missing-
billions-international-arms- flows-and-the-cost-of-conflict-123908
Hornberger, J. (2013). “From General to Commissioner to General: On the Popular State of
Policing in South Africa.” Law and Social Inquiry, 38(1), 598–614.
Human Rights Watch. (2015). World Report: Events of 2014.
IMLU (2017). Extra- Judicial Executions Full Report. Independent Medico-Legal Unit, Nairobi.
IMLU (2016). Extra- Judicial Executions Full Report. Independent Medico-Legal Unit, Nairobi.
IMLU (2015). Extra- Judicial Executions Full Report. Independent Medico-Legal Unit, Nairobi.
IPOA. (2013). Baseline Survey on Policing Standards and Gaps in Kenya. Independent Policing
Oversight Authority, Nairobi, PP. 13-20.
64 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 65
Owen, O., & Cooper Knock, S. (2015). “Between vigilantism and bureaucracy: Improving our
understanding of police work in Nigeria and South Africa.” Theoretical Criminology, 19(3), pp.
355-375.
Keane, J. (1996). Reflections on Violence. London: Verso.
Kagwanja P. (2003). “Facing Mount Kenya or facing Mecca? The Mungiki, political and ethnic violence
and the politics of the Moi succession in Kenya, 1987–2002. “ African Aairs, 102(406), 25–49.
Katumanga, M. (2005). “A city under Siege: banditry & modes of accumulation in Nairobi, 1991-
2004.” Review of African Political Economy, 32(106): 505-520.
KDH. (2014). Demographic and Health Survey. Kenya Demographic and Health Survey. 291-322.
KNBS. (2016). Statistical Abstract 1.0.0. Kenya National Bureau of Statistics, Nairobi.
KNCHR. (2015).The Error of Fighting Terror with Terror. Kenya National Commission of Human
Rights, pp. 3.
Lowe, L. (2001), Nakuru - a study in urbanization, Available at http://livelihoodtechnology.org/
home.asp?id=csNakuru .
Lemanski, C. (2012). “Everyday Human (in)Security: Rescaling the Southern City.” Security
Dialogue, 43(1), 61-78.
McEvoy, C. (2012). Battering, Rape, and Lethal Violence – A Baseline of Information on Physical
Threats against Women in Nairobi. Small Arms Survey, Graduate Institute of International and
Development Studies (GIAL), Working Paper, No. 13.
McKee, R. (2013). Lynchings in modern Kenya and inequitable access to basic resources: A major
human rights scandal and one contributing cause. Graduate Institute of Applied Linguisticts &
College of International Studies (GIAL), Special Electronic Publications, pp 1.
Mcllwane, C. (1999). Geography and Development: Violence and Crime as development issues,
Progress in Human Geography, 23(3), pp: 453-463.
Moser, C. (2004) ‘Urban violence and insecurity: an introductory roadmap’, Environment and
Urbanization Vol 16, No 2, pp. 3–16.
Moser, C. & Mcllwane, C. (2006). Latin American Urban Violence as Development Concern:
Towards a Framework for Violence Reduction. World Development, 34(1), pp.89-112.
Muggah R. (2012) Urban Resilience in Situations of Chronic Violence: Case Study of Nairobi,
Kenya, Center for International Studies.
Mwangi, J. 2013. “Administration and Supervision in early Childhood Education.” Centre for
Open and Distance Learning, Vol 1.
64 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 65
NCRC. 2013. Organized Crime Profiling. National Research Centre. 23 -25.
NPS. (2015). Crime Situation Report. National Police Service (Nairobi: Government Printers), p 8.
Okoth et al. (2014). “Services provided to households enrolled on orphans and vulnerable
children intervention programmes in Kisumu County, Kenya. “ Global Journal of Human Social
Science, Arts & Humanities Psychology, 14(9), pp. 1-9.
Oucho, J. (2002). Undercurrents of Ethnic Conflicts in Kenya. Leiden: Brill, 100-146.
Owen, Olly, and Sarah Jane Cooper-Knock. 2014. “Between Vigilantism and bureaucracy:
Improving our understanding of police work in Nigeria and South Africa.” Theoretical
Criminology, 1-21.
Rasmussen, J.(2013).The chameleon and the mugumo tree: the politics of the Kenyan Mungiki
movement. Roskilde Universitet.
RCSA (2015). A Report of Analysis on Armed Crimes in East Afria Community Countries: Burundi,
Kenya, Rwanda, Tanzania and Uganda. Regional Centre on Small Arms, pp 9.
Republic of Kenya. (2008). Report of the Commission of Inquiry into Post Election Violence.
Nairobi: Government Printer.
Republic of Kenya. (1999). Report of the Judicial Commission Appointed to Inquire into Tribal
Clashes in Kenya. Nairobi: Government Printer.
Republic of Kenya. (1992). Report of the National Assembly’s Select Committee to Investigate
Ethnic Clashes in Western and other Parts of Kenya. Nairobi: Government Printers.
Ruteere, M., & Pommerolle, M. (2003). Democratizing Security or Democratizing Repression?
The Ambiguities of Community Policing in Kenya. African Aairs, 102, 587-604.
Ruteere, M. (2008). Dilemmas of Crime, Human Rights and the Politics of Mungiki Violence in
Kenya. Kenya Human Rights Institute.
Stewart, F. (2008) Horizontal Inequalities and Conflict. Understanding Group Conflict in
Multiethnic Societies, Basingstroke: Palgrave Macmillan.
Syanda, J., & Ndetei, D.et al. 2007. ‘‘Bullying in Public Secondary Schools in Nairobi, Kenya.‘‘
Journal of child and adolescent mental health, 19(1), 45-55.
Tankebe, J. (2009). “Self-Help, Policing, and Procedural Justice: Ghanaian Vigilantism and the
Rule of Law.” Law & Society Review, 43 (2): 245-270.Vol. 101, No. 405 (Oct., 2002), pp. 531-555.
UNICEF. (2010). Violence Against Children in Kenya: Findings from a 2010 National Survey.
United Nations Children’s Fund.
66 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 67
UNODC & KIPPRA. (2010). Victimization Survey in Kenya. United Nations Oce on Drugs and Crime, 4.
Wairuri, S K. (2015). ‘Thieves should not live amongst people: Popular support for police violence
in Nairobi.’ Unpublished dissertation. University of Oxford.
Winton, A. (2004) ‘Urban violence: a guide to the literature’, Environment and Urbanization, 16(2).
World Bank. 2011. Violence in the City: Understanding and Supporting Community Responses
to Urban Violence
66 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 67
68 | INTERSECTORAL URBAN VIOLENCE PREVENTION URBAN VIOLENCE IN NAKURU COUNTY, KENYA | 68
Midrift Human Rights Network
www.midrifthurinet.org
Danida
www.um.dk/da/danida
CHRIPS - The Centre for Human Rights and Policy Studies
www.chrips.or.ke
DIGNITY – Danish Institute Against Torture, Denmark;
www.dignityinstitute.org
