Experiences in the Kenyan Criminal Justice System and Violent Extremism Dr Anneli Botha & Mahdi Abdile

FCA_KenyaAnalysis.indd 1 24.1.2020 12.06 Executive Summary

This study reports on research conducted in Kenya from 2017 to 2019. It explores the link between counterterrorism strategies and recruitment into violent extremism in Kenya as well as the role of the Kenyan criminal justice system in relation to violent extremism. Kenya has been experiencing violent campaigns links to violent extremism for at least two decades. Perpetrators of such acts usually come from counties and communities with Muslim majority residents. Compared to the rest of Kenya, these communities have unusually high unemployment, underdeveloped infrastructure, poor economic opportunities, long-standing local grievances, close-proximity to Somalia – all of which can serve as important internal and external radicalizing factors not only in Kenya but also in the wider region. Kenya’s contribution to foreign fighters to Somalia and to conflicts in countries such as Syria and Iraq have increased both in total and capital. Due to the violent and seditious nature of these terrorists, coupled with increased threats and a rise in the number of fatalities, Kenya has enacted an ambitious counter violent extremism and radicalization initiative aimed at preventing radicalization and rehabilitating those who have already been radicalized. While the emphasis continues to focus on individuals, there has also been recognition in that addressing violent extremism as part of prevention of terrorism has to include interventions aimed at addressing the environment conducive for violent ideologies and extremists who radicalise and recruit followers. It is evident from the Kenyan counter strategy initiatives that a holistic approach, that also addresses the development dimension of violent extremism, should be enacted. For counter initiatives to succeed, the government of Kenya should undertake two strategic initiatives. First, policies and strategies must be designed together with affected communities, including victims of violent extremism. The key to tackling violent extremism, and ultimately eliminating terrorism, lies in Kenya’s collective effort approach and less on a doing it alone approach. Second, evaluate what has worked and what has not through piloting and testing before rolling out. Critically, identify the gaps and limitations of current counter strategies.

FCA_KenyaAnalysis.indd 2 24.1.2020 12.06 Table of Contents

1. Introduction...... 8

2. Literature Review...... 9 2.1 Underlying Catalysts for Radicalisation and Violent Extremist Operations...... 9 2.2 State Response to Violent Extremism...... 11

3. Literature Review...... 13 3.1 Preliminary Stage...... 14 3.2 Field preparation...... 14 3.3 Fieldwork and Data collection...... 15 3.4 Focus Group Discussions...... 15 3.5 Analysis of data...... 17

4. Biographic background of respondents...... 19 4.1 Gender...... 19 4.2 Age...... 19 4.3 Marital status...... 19 4.4 Number of children...... 19

5. Education...... 25

6. Religion...... 27 6.1 Religious Identity...... 27

7. Government and National Identity...... 31 7.1 Ethnic Identity...... 33 7.2 Access to land...... 36 7.3 Create a sense of belonging in Kenya...... 39

8. Trust in Government...... 41 8.1 Participating in the political process...... 43 8.2 Access to national identity documents...... 44 8.3 Service delivery: Efficiency...... 47 8.4 Safety from foreign enemies...... 49 8.5 Safety from domestic enemies...... 50 8.5.1 Radicalisation and recruitment into illegal organisations...... 53 8.6 Corruption...... 63

9. Interaction with security forces...... 66 9.1 Police Legitimacy...... 70 9.1.1 Personal interactions...... 73 9.1.2 Procedural justice...... 76 9.1.3 Abuse of force and radicalisation...... 77 9.1.4 Conduct investigations and collect evidence...... 80 9.1.5 Religious and tribal representation of law enforcement...... 83 9.1.6 Profiling based on religion and ethnicity...... 85

FCA_KenyaAnalysis.indd 3 24.1.2020 12.06 9.2 Other security agencies...... 86

10. Recommendations...... 88 10.1 Government of Kenya...... 88 10.2 Kenya Police Service...... 89 10.3 Agencies involved in counter violent extremist activities...... 91

11. Conclusion...... 93

Figures

Figure 1: Relation to the person who went through the criminal justice system...... 17 Figure 2: Gender of respondents...... 20 Figure 3: Ages of respondents...... 20 Figure 4: Marital status of respondents...... 20 Figure 5: Number of children of respondents...... 21 Figure 6: Ages of children of criminal justice respondents...... 22 Figure 7: Number of children of individuals who experienced...... 23 the criminal justice system (detained) Figure 8: Education...... 25 Figure 9: Access to the same education opportunities...... 26 Figure 10: Reasons for not receiving the same education opportunities...... 26 Figure 11: Level of religious education received...... 28 Figure 12: Do you understand what you read...... 28 Figure 13: Reaction when confronted with conflicting preaching...... 28 Figure 14: Religious affiliation of respondents...... 30 Figure 15: Importance of religious identity...... 30 Figure 16: Identifying other religions as ‘them...... 30 Figure 17: Respondents’ association with being Kenyan (national identity)...... 32 Figure 18: Proud of being Kenyan...... 32 Figure 19: Identifying the Kenyan government as ‘them’...... 32 Figure 20: Ethnic/tribal representation of respondents...... 34 Figure 21: Importance of ethnic/tribal identity...... 35 Figure 22: Identifying other ethnic groups/tribes as ‘them’...... 35 Figure 23: Historic grievances based on ethnicity as a threat to domestic security...... 35 Figure 24: Unequal access to land and natural resources as a security challenge...... 36 Figure 25: Potential conflict over land...... 37 Figure 26: Need for land...... 37 Figure 27: Government’s success in providing access to land...... 37 Figure 28: Religious dialogue as a strategy to deal with security challenges...... 38 Figure 29: Government’s success in creating a sense of belonging...... 39 Figure 30: Initiatives to build national cohesion to deal with security challenges...... 40

4 5

FCA_KenyaAnalysis.indd 4 24.1.2020 12.06 Figure 31: National service of the youth...... 40 Figure 32: Community involvement to address security challenges...... 40 Figure 33: Trust in the presidency...... 42 Figure 34: Trust in national government...... 42 Figure 35: Trust in local government...... 42 Figure 36: Trust in politicians...... 42 Figure 37: Participation in the political process...... 43 Figure 38: Reasons for not participating in the political process...... 44 Figure 39: Stopped and asked for ID...... 45 Figure 40: Government’s success in providing identity documents...... 46 Figure 41: Government’s success in providing passports...... 46 Figure 42: Need to receive education...... 47 Figure 43: Ability of government to provide education...... 47 Figure 44: Need for healthcare...... 48 Figure 45: Government’s success in providing healthcare...... 48 Figure 46: Need to be protected against a foreign enemy...... 49 Figure 47: Success of the Kenyan government to protect citizens against foreign enemies...... 49 Figure 48: Terrorism as a main threat to domestic security ...... 50 Figure 49: Threat of al-Shabaab attacks in Kenya...... 50 Figure 50: Threat of Islamic State attacks in Kenya...... 51 Figure 51: Crime as a threat to domestic security...... 51 Figure 52: Serious nature of criminal activities: Criminal justice respondents...... 51 Figure 53: Serious nature of criminal activities: Family and friends respondents...... 52 Figure 54: Serious nature of criminal activities: Respondents history of being detained...... 52 Figure 55: Security threat overestimated...... 53 Figure 56: Seriousness of Kenyan nationals being recruited into illegal organisations...... 53 Figure 57: Most vulnerable to be recruited into an illegal organisation...... 55 Figure 58: Aware of initiatives to prevent people from joining illegal organisations...... 56 Figure 59: Duration being aware of counter measures...... 56 Figure 60: Leading agency responsible for counter initiatives...... 57 Figure 61: Type of initiatives...... 57 Figure 62: Assess the success of counter-initiatives (70-100% success rate)...... 58 Figure 63: Reasons why counter initiatives may not be successful...... 59 Figure 64: Aware of the Kenyan Government’s Strategy against violent extremism...... 59 Figure 65: Informed of Kenya’s counter strategy...... 69 Figure 66: Aware of the content of government’s strategy...... 60 Figure 67: Perception of corruption per level of government, department or institution...... 63 Figure 68: Actual corruption - family and friends sample...... 65 Figure 69: Actual corruption - detained sample...... 65 Figure 70: Number of times family members and friends interacted with the police...... 67 Figure 71: Number of times detained respondents interact with the police...... 67 Figure 72: Offences for being arrested...... 68 Figure 73: Arrest followed by another arrest...... 68 Figure 74: Relationship to the person arrested first...... 69 Figure 75: Age arrested for the first time...... 69 Figure 76: Who will be trusted with this information...... 71

4 5

FCA_KenyaAnalysis.indd 5 24.1.2020 12.06 Figure 77: Prevent family members and friends to come forward with information...... 72 Figure 78: Prevent respondents previously detained from coming forward with information...... 72 Figure 79: Were you treated fairly by security forces...... 73 Figure 80: Call the police when in danger...... 74 Figure 81: Personal experience of individuals previously detained...... 75 Figure 82: Personal experience of family members and friends...... 75 Figure 83: Presence of a catalyst to initiate joining of an illegal organisation...... 78 Figure 84: Enhancing the abilities of the police to be able to address security challenges...... 81 Figure 85: Ability of security officials to work with vulnerable communities...... 82 Figure 86: Did security forces in the area reflect the main religious affiliation ...... 84 found in the community Figure 87: Majority of police officers represent ethnic groups in the area...... 84 Figure 88: Singled out based on religious affiliation...... 85 Figure 89: Singled out based on ethnic/tribal affiliation...... 85

Tables

Table 1: Possible developmental effects on children of parents arrested and incarcerated...... 24 Table 2: Intergenerational behaviours, crime and incarceration...... 24 Table 3: Ethnic composition...... 33 Table 4: Trust in government and its institutions...... 41 Table 5: Reasons for joining illegal organisations based on priority...... 54 Table 6: Radicalisation facilitated through...... 55 Table 7: Priority of counter strategies...... 61 Table 8: Who should take the lead implementing the strategy...... 61 Table 9: Priority of proposed changes to enhance the effectiveness of counter measures...... 62 Table 10: Level of confidence in training...... 82 Table 11: Negative interactions with criminal justice actors...... 87

6 7

FCA_KenyaAnalysis.indd 6 24.1.2020 12.06 Commissioned by

Acknowledgement

This research on Experiences in the Kenyan Criminal Justice System and Rates of Violent Extremism was carried out in partnership between Finn Church Aid (FCA), The Network for Religious and Traditional Peacemakers, Muslims for Human Rights and Kenyan Muslim Youth Alliance and funded by the U.S. Department of State’s Bureau of Counterterrorism. On behalf of the project consortium, we would like to thank all the project partners and sup- porters, who have enabled the successful completion of this research project. We would also like to express our gratitude towards the Government of Kenya, especially the Ministry of Interior and the Office of the President, for their support throughout the process.

We trust that the data, results and recommendations deriving from this research report will be mutually beneficial for the national, international and local partners of the project and increase the understanding of the relationship between experiences in the criminal justice system, drivers and support for violent extremism in Kenya.

6 7

FCA_KenyaAnalysis.indd 7 24.1.2020 12.06 1. Introduction

Violent extremism continues to be an issue of global concern. The high number of fatalities arising from violent extremism has been catastrophic and the need for concerted international initiatives to counter violent extremism is evident.1 Kenya, in particular, has experienced numerous terrorist attacks in the last decade, which in addition to fatalities, have resulted in massive economic losses with a long term impact on foreign direct investment.2 The apparent continuing threat of violent extremism was reaffirmed during the al-Shabaab terrorist attack at the Dusit Hotel in Nairobi on 15 January 2019.3

The main objective of the research project is to advance a deeper understanding of ways in which violent extremist organizations use the breakdown of trust between local communities and state and criminal justice system to harness support for their activities and to recruit individuals at risk. To achieve this overarching objective interviews were conducted with respondents representing individuals who had been arrested, detained or convicted on terror charges in Kenya (referred to as the detained sample), their immediate families and friends, communities in the targeted counties as well as officials from the Criminal Justice System and Kenyan government. Questionnaires were developed to identify the triggering factors, including perceptions of unjust treatment and a history of mistrust and marginalization contributing to the erosion of relations and antagonistic attitudes between criminal justice authorities and local communities.

In light of the grave impact occasioned by violent extremism, coupled with changing dynamics of the strategies employed by the violent extremists, as well as the guaranteed continued attacks the designing of effective counter strategies is critical. In the absence of empirical research on all the different dimensions of violent extremism, but more important the consequences of countermeasures, counter strategies are at risk of being ineffective. In other words, an evidence-based understanding of the underlying factors propelling violent extremism is crucial.

Through analysing the experiences of the three target groups, the research results aim to improve the understanding of violent extremism in Kenya by providing primary data on how to better approach violence extremism. Second, it hopes to identify any policy shortfalls that impede the implementation of counter violent extremism policies. Third, this analysis hopes to identify policy shortcomings and provide policy recommendations to address contributing factors. These include the perceptions of unjust treatment as well as a history of mistrust and perceived marginalization that may contribute to the erosion of relations and antagonistic attitudes between the criminal justice system and local communities.

1 UNDP, Journey to Extremism (UNDP, 2017), vi. 2 UNDP, Articulating the Pathways of the Impact of Terrorism and Violent Extremism on the Kenyan Economy (UNDP, 2017) 1. 3 Jason Burke, “Nairobi Terror Attack: Gunfire Heard Hours After Minister Declares Scene Secure”, The Guardian, January 16, 2019; Macharia Munene, “Response to Terror Attack the Most Important Aspect”, The Standard, January 21, 2019.

8 9

FCA_KenyaAnalysis.indd 8 24.1.2020 12.06 2. Literature Review

Research conducted on radicalisation in Kenya until the time of writing attempted to identify both pull- and push factors into violent extremist organisations. Capturing some of the most prominent findings the following section will be divided into several sub-headings.

2.1 Underlying Catalysts for Radicalisation and Violent Extremist Operations Research reveals that diverse factors make individuals vulnerable to radicalisation. A key factor Bradbury and Kleinman (2010), Botha (2013 and 2014) and Kenya Muslim Youth Alliance (2016) highlighted was that individuals harbouring a sense of marginalisation embrace a narrative of victimisa- tion4. Marginalisation is based on most notably ethnic and religious identities. Therefore, individuals belonging to a certain ethnic or religious group may perceive themselves as marginalised, hence be driven to protect the groups they belong to, as identified by Botha (2014).5 This sense of marginalisation thrives in circumstances where a sense of national identity is lacking according to Botha (2013).6 According to Badurdeen (2012), in Kenya ethnic and religious identity have often been politicised to achieve political advantage. For instance, amongst the coastal communities narratives of political and economic marginalisation have been repeatedly disseminated thus providing an entry point for radicalisation.7 KMYA (2016) as well as Rink and Sharma (2018) explained that al-Shabaab extremists have used this narrative to lure Muslims who already considered themselves as having long been marginalised.8

Apart from the collective identity, Odhiambo (2016) explained that deficiencies in individual identity creates vulnerability;individuals in search of their personal identity have been seen as easy targets for radicalisation.9

The Open Society Foundation and MUHURI (2013), Botha (2014), Patterson (2015), Villa-Vicencio, Buchanan-Clarke and Humphrey (2016) and Abdikadir (2016) emphasised that discontent against State action, such as the response to violent extremism breeds a desire to retaliate thus creating

4 Anneli Botha, Assessing the Vulnerability of Kenyan Youths to Radicalisation and Extremism (ISS, 2013): 14; Anneli Botha, “Political Socialization and Terrorist Radicalization Among Individuals Who Joined Al- Shabaab in Kenya,” Studies in Conflict and Terrorism, 37 (November 2014): 913; HORN and CSCR, Map- ping Dynamics and Perceptions of Violent Extremism A Study of Nature, Drivers and Perceptions of Muslim Women and Girls Towards Violent Extremism in Kenya (HORN and CSCR) 14.International Alert and KYMYA, We Don’t Trust Anyone Strengthening Relationships As The Key To Reducing Violent Extremism In Kenya (International Alert and KYMYA, 2016) 11. Mark Bradbury and Michael Kleinman, Winning Hearts and Minds? Examining the Relationship between Aid and Security in Kenya (Feinstein International Centre, 2 0 1 0) 30. 5 Anneli Botha, “Political Socialization and Terrorist Radicalization Among Individuals Who Joined Al- Shabaab in Kenya,” Studies in Conflict and Terrorism, 37 (November 2014): 902. 6 Anneli Botha, Assessing the Vulnerability of Kenyan Youths to Radicalisation and Extremism (ISS, 2013): 19. 7 Fathima Azmiya Badurdeen, “Youth Radicalization in the Coast Province of Kenya,” Africa Peace And Con- flict Journal 5, no. 1 (June 2012): 56. 8 International Alert and KYMYA, We Don’t Trust Anyone Strengthening Relationships As The Key To Reducing Violent Extremism In Kenya (International Alert and KYMYA, 2016) 12; Anselm Rink and Kunaal Sharma, “The Determinants of Religious Radicalization: Evidence from Kenya, ” Journal of Conflict Resolution 62, no. 6 ( 2018) 1231 9 E.O.S Odhiambo et.al, “Domestic Radicalisation in Kenya,” Global Journal of Interdisciplinary Social Sciences 4, no. 3 (May-June, 2015): 50; Villa-Vicencio Charles, et al, 2016:20)

8 9

FCA_KenyaAnalysis.indd 9 24.1.2020 12.06 vulnerability to radicalisation.10 Similarly, Odhiambo (2015) and Abdikadir (2016) found that dissat- isfaction based on socio-economic difficulties creates vulnerability and enables violent extremists to recruit individuals by providing economic incentives.11 Socio-economic discontent further erodes a sense of belonging and thrives under the notion that, according to Villa-Vicencio et al (2016) “… when you have nothing, you have nothing to lose…”.12

According to Botha (2014), influential people in the community also play a major role in the radicalisation of individuals. For instance, religious leaders have been known to misuse religious texts to justify violent extremism thus encouraging followers to join extremist groups.13 Mazrui (2018)14 further highlighted that geopolitics have also contributed to violent extremism and that unresolved conflicts which sustain instability also create environments that are conducive for radicalisation.15

The transnational nature of violent extremism requires inter-state movement and porous borders such as between Kenya and Somalia to further facilitate extremist activities.16 Rosenau in 2005 and Patterson (2015) explained that insufficient border control and immigration management, often as a result of corruption, provide a conducive environment for violent extremist groups to operate.17 Thus, countries with high levels of corruption become easy targets for such groups.

The role of the family in a person’s early development was highlighted by Botha (2014) in her thesis on the role socialization plays in later vulnerability, as well as weakened family structures, that according to Villa-Vicencio et al (2016), further contribute to radicalisation, particularly amongst the youth. The lack of solid authority within families renders young people vulnerable to radicalisation and makes it difficult for families to intervene against radicalisation.18 Violent extremism is also dynamic and extremist groups aptly identify new opportunities that will further their agenda. One such opportunity relates to gender dynamics which present women as valuable players in violent extremism. Extremist groups are increasingly targeting women for recruitment owing to their lower visibility as terrorists, their strong influence on their sons and their role as wives of terrorists.19

10 Osman Ali Abdikadir, Youth Radicalization as a Tool for Terrorism in East Africa: A Case Study of Kenya. (MA in International Studies diss., University of Nairobi, 2016), 9; MUHURI and Open Society Foundations, “We’re Tired of Taking You to Court” (Open Society Foundations, 2013), 11; Villa-Vicencio Charles, et al, Community Perceptions of Violent Extremism in Kenya (IJR, 2016), 19; William R. Patterson, “Islamic Radi- calization in Kenya,” Joint Force Quarterly 78 (July 2015): 20. 11 Anneli Botha, Assessing the Vulnerability of Kenyan Youths to Radicalisation and Extremism (ISS, 2013): 5; Anneli Botha, “Political Socialization and Terrorist Radicalization Among Individuals Who Joined Al- Shabaab in Kenya,” Studies in Conflict and Terrorism, 37 (November 2014): 907; E.O.S Odhiambo et.al, “Domestic Radicalisation in Kenya,” Global Journal of Interdisciplinary Social Sciences 4, no. 3 (May-June, 2015): 50; Osman Ali Abdikadir, Youth Radicalization as a Tool for Terrorism in East Africa: A Case Study of Kenya. (MA in International Studies diss., University of Nairobi, 2016), 15. 12 Villa-Vicencio Charles, et al, Community Perceptions of Violent Extremism in Kenya (IJR, 2016), 18. 13 Anneli Botha, “Political Socialization and Terrorist Radicalization Among Individuals Who Joined Al- Shabaab in Kenya,” Studies in Conflict and Terrorism, 37 (November 2014): 904; HORN and CSCR, Map- ping Dynamics and Perceptions of Violent Extremism A Study of Nature, Drivers and Perceptions of Muslim Women and Girls Towards Violent Extremism in Kenya (HORN and CSCR) 12. 14 Anneli Botha, Assessing the Vulnerability of Kenyan Youths to Radicalisation and Extremism (ISS, 2013): 13; Alamin Mazrui et al, “Global and Local Contexts of Terrorism and Counterterrorism in Kenya,” in Counter- ing Violent Extremism in Kenya: Between the Rule of Law and the Quest for Security, ed. Alamin Mazrui et al (Nairobi: Twaweza Communications, 2018), 15. 15 Anneli Botha, Assessing the Vulnerability of Kenyan Youths to Radicalisation and Extremism (ISS, 2013), 13. 16 Anneli Botha, Assessing the Vulnerability of Kenyan Youths to Radicalisation and Extremism (ISS, 2013), 4. 17 William Rosenau, “Al Qaida Recruitment Trends in Kenya and Tanzania” Studies in Conflict & Terrorism 28 (2005): 2; William R. Patterson, “Islamic Radicalization in Kenya,” Joint Force Quarterly 78 (July 2015): 20. 18 Villa-Vicencio Charles, et al, Community Perceptions of Violent Extremism in Kenya (IJR, 2016), 20.

10 11

FCA_KenyaAnalysis.indd 10 24.1.2020 12.06 Enlisting in violent extremist groups has also been linked to general criminality. Individuals involved in criminal activities have been known to join the groups with the intention of disappearing and avoid- ing law enforcement.20 Further, the activities of violent extremist groups are facilitated within environments in which criminality is rife with ineffective law enforcement. According to Villa-Vicencio et al (2016), al-Shabaab in some cases recruit members through coercion and abduction.21

The advancement in technology has played a key role in furthering and actualising the agenda of violent extremist groups. For instance, the internet had been used to recruit individuals and coor- dinate extremist activities from other countries. According to Odhiambo (2015) with reference to Kenya, social media additionally facilitated easy access to young people enhancing their vunrability to be recruited into violent extremist organisations.22 According to Patterson (2015) in addition to the use of technology to radicalise and recruit new members, a good transportation and communication network furher facilitates the movement and communication of extremist groups, therefore enabling them to operate with ease.23

2.2 State Response to Violent Extremism In view of the imminence of violent extremist attacks and the continued recruitment of individuals into extremist groups, Kenya introduced a robust legal framework in response to violent extremism. This legal framework is set against the backdrop of critical institutions established by the Constitu- tion of Kenya with reference to the Kenya Defence Forces, the National Intelligence Service and the National Police Service. Other critical organs include the judiciary, the Kenya Prisons Service and the Probation and Aftercare Services. To specifically respond to violent extremist activities, the National Counter Terrorism Centre was established and consists of appointees from the National Security Council, National Intelligence Service; Kenya Defence Forces; the Attorney-General Directorate of Immigration and Registration; the National Police Service and such other national agencies, as may be determined by the National Security Council.

The Prevention of Terrorism Act was enacted in 2012 to specifically make provision for the prevention, detection and prosecution of terrorist activities. Further, the Proceeds of Crime and Anti–Money Laundering Act enacted in 2017, targets the financing of criminal activities which is a critical strategy against violent extremism. Apart from the general provisions empowering the security agencies to combat crime such as powers to search and arrest, the Prevention of Terrorism Act further provides powers to deal with specific terrorist activities. First, the Act criminalises a wide range of activities that are linked to terrorism. For instance, radicalisation is criminalised and is defined in very broad terms which capture a wide range of actions geared towards promoting ideologies sup- porting extremist violence.24 Secondly, punitive sanctions are prescribed for terrorist crimes.25 Third, the rights of individuals suspected of having committed terrorist activities may be limited. Section 35 of the Prevention of Terrorism Act provides that the right to bail, freedom of expression, freedom of security and the right to property may be limited. With respect to the right to bail, the cumulative extension by the court of periods in which suspected members of terrorist groups are remanded can extend to three hundred and sixty days.26 Fourth, the Act gives authority for the designation/

19 HORN and CSCR, Mapping Dynamics and Perceptions of Violent Extremism A Study of Nature, Drivers and Perceptions of Muslim Women and Girls Towards Violent Extremism in Kenya (HORN and CSCR) 23; Villa-Vicencio Charles, et al, Community Perceptions of Violent Extremism in Kenya (IJR, 2016), 25. 20 Villa-Vicencio Charles, et al, Community Perceptions of Violent Extremism in Kenya (IJR, 2016), 21. 21 Villa-Vicencio Charles, et al, Community Perceptions of Violent Extremism in Kenya (IJR, 2016), 21. 22 E.O.S Odhiambo et.al, “Domestic Radicalisation in Kenya,” Global Journal of Interdisciplinary Social Sci- ences 4, no. 3 (May-June, 2015): 50 23 William R. Patterson, “Islamic Radicalization in Kenya,” Joint Force Quarterly 78 (July 2015): 17. 24 Prevention of Terrorism Act, s.12D. 25 Prevention of Terrorism Act, part III.

10 11

FCA_KenyaAnalysis.indd 11 24.1.2020 12.06 listing of individuals or entities amenable to sanctions for their involvement in terrorist activities.27 Sanctions include freezing or seizing funds or property belonging to the designated individuals or entities involved in terrorist activities.28

To streamline its counter violent extremism initiatives and to provide a well-coordinated approach, Kenya launched its National Strategy to Counter Violent Extremism. Similarly, Mombasa, Kwale, Lamu and Kilifi Counties have launched County Plans to Counter Violent Extremism.

Apart from a robust legal framework designed to combat violent extremism, Kenya has also strength- ened community policing, referred to as the “Nyumba Kumi” initiative, with the purpose to enhance detection of extremist groups. However, according to International Alert and the KMYA (2016), mistrust of the State by affected communities has undermined the operations of community policing.29

Following terrorist attacks in Mombasa and Nairobi, a State operation referred to as “Usalama Watch” was undertaken with the intention of identifying individuals involved in terrorist activities. The outcome of the operation was the arrest of about 4000 individuals, the majority being of Somali origin.30 The disproportionate impact of the Usalama Watch on the Somali community created a sense of victimisation and was counterproductive to the efforts of the State in countering violent extremism.31

According to MUHURI and the Open Society Foundation (2013), Abdikadir (2016) and Van Metre (2016) to name a few, allegations of forced disappearances, arbitrary arrests and extrajudicial killings of individuals linked to terrorist activities (also to be referred to in this study); perceived victimisation of some communities and further diminished trust between the State – especially the security organs – and the said communities. This diminished trust is taken advantage of by violent extremist groups and fits into their victimisation narrative which they employ when recruiting individuals.32

26 Prevention of Terrorism Act, section 33. 27 Prevention of Terrorism (Implementation of the United Nations Security Council Resolutions on Suppression of Terrorism) Regulations, 2013, Part III. 28 Prevention of Terrorism (Implementation of the United Nations Security Council Resolutions on Suppression of Terrorism) Regulations, 2013, Part IV. 29 International Alert and KYMYA, We Don’t Trust Anyone Strengthening Relationships As The Key To Reducing Violent Extremism In Kenya (International Alert and KYMYA, 2016) 32. 30 Jeremy Lind, Patrick Mutahi and Marjoke Oosterom, Tangled Ties: Al-Shabaab and Political Volatility in Kenya (Institute of Development Studies, 2015), 27; KNCHR, The Error of Fighting Terror with Terror (KNCHR, 2015) 3. 31 Amnesty International, Somalis Are Scapegoats in Kenya’s Counterterror Crackdown (AI, 2014), 14. 32 Lauren Van Metre, Community Resilience to Violent Extremism in Kenya (USIP, 2016), 3; MUHURI and Open Society Foundations, “We’re Tired of Taking You to Court” (Open Society Foundations, 2013), 47; Osman Ali Abdikadir, Youth Radicalization as a Tool for Terrorism in East Africa: A Case Study of Kenya. (MA in International Studies diss., University of Nairobi, 2016), 9.

12 13

FCA_KenyaAnalysis.indd 12 24.1.2020 12.06 3. Research Methodology

Finn Church Aid (FCA) based in Finland, in collaboration with the Network for Religious and Traditional Peacemakers, Muslims for Human Rights (MUHURI) and Kenyan Muslim Youth Alliance (KMYA) as operating partners, conducted empirical research on the question whether there was a relationship between experiences associated with the criminal justice system and radicalisation into violent extremist organisations. The primary purpose of this research was to test the hypothesis that there was a link between counterterrorism strategies and recruitment into violent extremism. If there was a relationship, the report hoped to inform policy-makers on possible additions to current counterterrorism and counter violent extremist strategies.

Facilitating the collection of primary data, the study used structured questionnaires in which the majority of questions used a Likert scale to measure respondents’ attitudes and opinions on a range of topics (quantitative). Where necessary, these closed questions were followed by open-ended questions (quali- tative) to allow respondents to add more information or to clarify. Collected data was analyzed and presented in percentages.

After the questionnaires had been developed, all three were tested to allow corrections to be made where necessary before the fieldwork commenced. It was during this phase that overseers identified respondents’ inability to connect both numbers (1-7) and words (strongly disagree, disagree, neither agree nor disagree, agree and strongly agree) to their attitudes and opinions. Consequently, the traditional sliding scale was replaced by percentages (0% - 100%) that proved to be more effective.

In addition to testing the questionnaires during the pre-interview phase, a preselection questionnaire was used to ensure that the right respondents had been targeted and that consent forms were signed in the five targeted counties. This stage was particularly critical to introduce the project, build a relationship with the targeted communities and to attain better understanding of the local circumstances in preparation for the fieldwork. This short questionnaire explained the purpose of the study and ensured that respondents were made aware of measures to safeguard their identity, as well as measures to secure information they provided. Snowball sampling with very specific respondents in mind was used to ensure that the most relevant individuals were interviewed. With reference to the detained sample of individuals who went through the criminal justice system for terrorism-related offences, 161 one-on-one interviews were conducted (66 detained and 57 arrested and charged on terrorism-related offences). The second sample included 110 interviews conducted with family members and friends associated with individuals who went through the criminal justice system. The last set of interviews was conducted with 114 respondents (62 male and 52 female) who were part of the criminal justice system tasked with countering terrorism and violent extremism. These interviews were conducted in Mombasa, Kilifi, Kwale, Tana River, Lamu, Garissa, Wajir and Nairobi; these areas have proved to be vulnerable to radicalization and recruitment into violent extremism. 516 individuals were invited to participate in the study of whom 131 declined the invitation, bringing the overall total of people interviewed to 385.

In Kwale, the majority of extremism-related incidents had been recorded in Bongwe and Gombato ward in Msambweni sub-county. As a result, Msambweni sub-county became the focus area for the study. Interviews in Nairobi were conducted in Majengo and Kibra (areas that had borne the brunt of violent extremism in Kenya since the 1998 bombing of the U.S. Embassy). The study predominately focussed on Likoni and Mvita (especially Majengo) sub counties. Considering the changing social and contextual dynamics of violent extremism and radicalization, Kisauni sub county was also included. In addition to the spread of violent extremism, youths in these areas complained of being the target of heavy-handed security operations.

12 13

FCA_KenyaAnalysis.indd 13 24.1.2020 12.06 As part of the field research preparation, local teams of interviewers were identified and trained. Lever- aging on the existing pool of student researchers at the local universities, two research assistants were drawn from the University of Nairobi’s Mombasa Campus. Additionally, local teams of interviewers, including youth leaders representing both men and women were identified to assist in the research. Local youth and community mobilisers further acted as gate openers into the communities.

3.1 Preliminary Stage During the preliminary stage research permits and consent from relevant authorities were secured, most notably from the Office of the President. Introductory letters addressed to the respective county commissioners across the coastal region, the various offices of the Directorate of Public Prosecution (DPP), the Attorney General (AG), the judiciary and the Law Society of Kenya (LSK) were prepared and shared across the region. Permission to conduct interviews were also received from the Office of the County Director of Probation in Mombasa, the Regional Commander of Prisons in the Coast, as well as the principle magistrate at Shanzu court. Lastly, lead teams met with the local administration associated with the office of the county commissioner in the Ministry of Interior. Government officials, especially the local authorities at county level, were initially reluctant to embrace the research and required approvals from their leadership to participate in the study.

Although required, governmental permission to conduct the research made outreach to communities even more challenging as the project was seen by many as a government-sponsored study. Many could not see the research as being independent from government interference. Such association between the research and the government made the researchers to be percieved as government associates and therefore spies. In fact, some community members openly stated that the research team looked like government security intelligence officers. Often members of the communities were more suspicious since field workers were seen as government spies and that by participating in the project, respondents could become the target of extrajudicial assassinations or could disappear. As a result of this suspicion community members started to turn one against the other, as explained by one respondent in Kwale:

“We have been living in systemic fear; even amongst ourselves, no one trusts the other. You cannot trust even your close relative because he or she might be the reason as to either you disappear or are killed tomorrow”.

Another interviewee in Bongwe painted a similar picture of several families fleeing their homes in fear of being targeted by either al-Shabaab or security agents:

“You may be sharing security information not only to unknown individuals but even to close friends thinking that you are trying to solve a problem, but that information is finally misused putting your life or that of your loved ones in danger. That notwithstanding, once a neighbour’s son gets lost amidst such kind of processes (research) in which you are involved, then you are not only becoming a target to the terrorists but also face the general community’s reprisals.”

Having noticed these challenges, the research team decided to approach the matter differently. This time the teams began contacting influential community leaders with standing in the community as well as liasoned with local Community Based Organisations (CBOs) who had maintained a close relationship with the targeted communities through established networks and who were trusted by the people. With the assistance of these individuals and organizations, people’s perception of, and objections to, the study started to change. Individuals started to come forward to participate in the project.

3.2 Field preparation As part of the field research preparation, field teams were briefed to enhance their understanding of the topographical dynamics, climatic conditions as well as associated security situations. Based on this, information risk assessments and mitigation matrix were prepared and shared with the field team.

14 15

FCA_KenyaAnalysis.indd 14 24.1.2020 12.06 Preparations to initiate fieldwork were essential in addressing the following:

• The sensitive nature of the research topic and stigma associated with the terror label. It also called for ways to address perceptions on the side of respondents who went through the criminal justice system as well as their family and friends since the project had received official authori- zation from the relevant authorities.

• Appointments had to be made with the respective respondents. This also included coordination with prison officials in order to obtain the necessary clearance to conduct the interviews in the prisons under circumstances that would put both the interviewee and respondent at ease. Similarly, it was equally important to agree on a meeting point with other respondents to ensure that interviews could be conducted without any interferences.

3.3 Fieldwork and Data collection The data collection process commenced in July 2018. Most challenging during this period was that in some counties (especially in Nairobi) the fieldwork coincided with government security operations. It created suspicion among the local communities that these operations and the data collection were conveniently coincided. Primary data collection was done by means of the printed questionnaire for ease of recording and future reference. Once completed, the data was keyed into tablets to create an electronic copy that was password protected, and then sent to a central electronic repository for further analysis. Many respondents, especially those individuals who had been arrested, charged or convicted and their families were careful about their safety. Consequently, a number of interviews were conducted at venues outside the target areas where respondents felt comfortable.

At the end, the fieldwork started with interviewing those individuals who had been arrested, detailed or charged on violent extremism related charges followed by family members and criminal justice actors. The focus group discussions (FGDs) came to order during the second round of interviews which comprised of a mixture of families and immediate communities, including civil society members, the religious leaders, business leaders and criminal justice actors.

Field teams in especially Malindi and Kwale noticed that Kenyan security officers conducted surveillance operations on the field workers despite official permission, realizing that it could impact on the research. It was also as a result of this discovery that field teams became more concerned with the possibility that al-Shabaab’s sights were set on them. This situation was a reminder of how members of the community could have felt, being in the centre of two opposing forces (as will be referred to later in this report).

3.4 Focus Group Discussions Focus groups discussions (FGD) involved individuals from the three groups representing diverse religious, gender, political and socio-economic backgrounds. The topics discussed were similar to those used in the one-on-one interviews and were aimed at discussing and capturing the views and experiences of respondents. Between 15-25 people participated in each focus group discussion. In total, two FGD’s were conducted per county, bringing the total to 16.

FGDs participants maintained the main threat associated with violent extremism related to its ability to attract vulnerable youth. According to the participants, violent extremists tapped into grudges young people held against the state and its security forces, thereby turning the youth into militants. Young militants who seek to change the system through violent means not only pose a serious security threat, it is also an attack on the social fabric of the community. Participants highlighted the following factors that cause Kenyan citizens to be vulnerable to violent extremism: radical religious teachings that are not effectively challenged by Muslim clergy, as well as economic and social marginalization.

14 15

FCA_KenyaAnalysis.indd 15 24.1.2020 12.06 Considering that especially family members and friends had firsthand experience of the radicalization process, they identified the following signs: a sudden change in behaviour, especially being overly religious; individuals isolating themselves from the rest of the family and declaring that other people are infidels or not “proper” Muslims; identifying themselves with the causes of people perceived to be oppressed such as the Palestinians and a sudden dislike and condemnation of Western culture and institutions such as democracy, music, sport and fashion are some of the traits that could identify radicalized individuals.

On the role that could be played by community members to prevent recruitment into violent groups and getting radicalized, participants felt there was a need to engage in honest discussions among themselves, whereafter those communities needed to engage with other groups including religious groups and the government. Byworking with police and criminal justice actors, those communities would be in a better position to speak with one voice when engaging with challenging narratives used by extremist groups to radicalize vulnerable people. This would also facilitate participation in community initiatives to promote peace through inter-religious social and inter-cultural dialogue specifically regarded as counter-measures going a long way in shaping the behaviour of youth and other community members.

Focus group discussions highlighted issues of interest to the three groups who were the subjected of this study. Issues that came up constantly included:

• The stigma associated with the fight against violent extremism;

• Fear within the community;

• Poor coordination among key stakeholders involved in the criminal justice system and the lack of a common approach in fighting violent extremism. It was clear that there was a lot of finger pointing amongst the stakeholders;

• The lack of trust between the criminal justice actors and communities, especially in targeted counties. This issue was addressed in every FGD;

• Community perception of security forces when they require assistance. Kenyan security agencies have a reputation of being harsh, oppressive, insensitive and treating people as “guilty until proven otherwise” even when the former require the assistance of the community. Consequently in the mind of community members there was no need to cooperate with the police, even in cases where community members had been in possession of useful information about radicalization and violent extremism that were taking place in the community.

• Participants had similar views when it came to the criminal justice system whom they regarded as corrupt and only favouring the rich and politically powerful over the poor.

• Participants also felt that mistrust between the police and the community created the perception that the police was regarded as an external actor and therefore not being part of the community.

• When asked how to improve the relationship between the police and the community, participants proposed continuous community dialogue between security agencies and the community, capacity building and the retraining of the police.

On recommendations how to improve the justice system in order to prevent violent extremism and terrorism, participants proposed that the human rights of every person who came into contact with the criminal justice system should be observed and upheld at all times. The majority of Muslims detest radicalization and violent extremism and want the removal of bad elements from the community, but

16 17

FCA_KenyaAnalysis.indd 16 24.1.2020 12.06 within the legal framework. Stereotyping and collective punishment that often are exhibited within the criminal justice system when dealing with matters of radicalization and violent extremism cause many Muslims to withdraw their support for counter violent extremism (CVE) initiatives.

3.5 Analysis of data As already mentioned, researchers used structured questionnaires, which contained questions to allow for an ordinal scale of measurement (quantitative), and distinguishing open-ended questions (quali- tative), to collect data on perceptions and opinions. Collected data was analysed through referring to percentage and descriptive analysis. Although more interviews were conducted, 136 interviews as part of the ‘detained’ sample, 110 with family members and friends of individuals who went through the criminal justice system or who were family members or friends of individuals who had been killed or disappeared, and lastly 133 respondents representing the criminal justice system were used in the analysis. In other words, 385 one-on-one interviews will be used in this analysis. Questionnaires were substantialexcluded survey from this questions analysis asin casessome where respondents respondents indic onlyated provided that they their no biographic longer wanted data without to be partanswering of the study any of(for the whatever substantial reason). survey questions as some respondents indicated that they no longer wanted to be part of the study (for whatever reason). Criminal justice respondents predominately consisted of police officers (50%) followed by correctionalCriminal justice officers respondents (25.58%), predominately individuals consistedrepresenting of police the officersjudiciary (50%) (15.12%) followed and by correctional prosecutors (9.30%).officers When (25.58%), asked individuals to explain representing the relationship the judiciary with (15.12%) the anddetained prosecutors respondent(s) (9.30%). When who asked went throughto explain the criminal the relationship justice withsystem, the detainedthe majority respondent(s) (29%) indicated who went throughthat they the were criminal the justice spouse sys -of the tem,detained the majority, followed (29%) by indicated 24% who that theywere were sibling the sspouse of the of detained the detained, and followed 18% who by 24%were who the parentswere osiblingsf the detained of the detained. Only and 8% 18% of whorespondents were the parentsinterviewed of the weredetained. friends Only of8% individuals of respondents who wentinterviewed through the were criminal friends justiceof individuals system. who In wentother through words, the 92% criminal of respondents justice system. belonged In other words, to the same92% family of respondents as the detained belonged torespo the samendent family, a fact as the that detained would respondent, shed valuable a fact thatinsight would into shed the backgroundvaluable insight of respondents into the background who went of respondentsthrough the who criminal went throughjustice system.the criminal It also justice implied system. that It also implied that the interaction with the criminal justice system was extended to those individuals and the interaction with the criminal justice system was extended to those individuals and their family their family members, often making them the target of what some respondents referred to as ‘collective members, often making them the target of what some respondents referred to as ‘collective punishment’ after being treated as criminals themselves. punishment’ after being treated as criminals themselves.

Figure 2: Relation to the person who went through the criminal justice system

Friend Grandparents 8% 2% Parents to 18% Family 12%

Child to In laws 2% 5%

Sibling 24%

Married to 29% n=126

As not all questions in the questionnaires were answered, each question would present the total of respondentsFigure 1. Relation who to the answered person who wentthe throughrespective the criminal question justice systemof each category (n=). The questionnaire used in interviewing family members and friends included questions about the person who went through the criminal justice system as well as questions relevant to the person being interviewed. 16 In an attempt to prevent confusion, reference to ‘detained 2’ will refer to the person who went17 through the criminal justice system (CJS) and ‘family’ will refer to information provided by the person being interviewed and relevant to the respondent.

FCA_KenyaAnalysis.indd 174. Biographic background of respondents 24.1.2020 12.06

4.1 Gender In total 255 interviews were conducted with male and 128 with female respondents. It is however important to note that more women (83) were interviewed than men (42) as part of the sample representing family members and friends.

14 As not all questions in the questionnaires were answered, each question would present the total of respondents who answered the respective question of each category (n=). The questionnaire used in interviewing family members and friends included questions about the person who went through the criminal justice system as well as questions relevant to the person being interviewed. In an attempt to prevent confusion, reference to ‘detained 2’ will refer to the person who went through the criminal justice system (CJS) and ‘family’ will refer to information provided by the person being interviewed and relevant to the respondent.

18 19

FCA_KenyaAnalysis.indd 18 24.1.2020 12.06 4. Biographic background of respondents

4.1 Gender

In total 256 interviews were conducted with male and 129 with female respondents. It is important to note that more women (83) were interviewed than men (42) as part of the sample representing family members and friends.

4.2 Age

Respondents representing the criminal justice system were predominately older than 40 years of age, whereas the individuals that were part of the detained sample were predominately between the ages 24 and 35. Individuals who went through the criminal justice system (according to family members and friends) as well, also represented older individuals, even older than 50 years of age. Similarly, 39.39% of family members interviewed were older than 40 years of age (see Figure 3).

4.3 Marital status The majority of respondents interviewed were married, including respondents interviewed representing family members and friends of which 58.14% were married and 19.38% a widow or widower (often in relation to detained 2). Although not in the majority, 31.5% of the detained and 29.03% of the detained 2 samples were single.

Although the focus of this study was not to assess the impact that belonging (or the suspect’s sense of belonging) to an illegal organisation had on the person’s marital status, an interesting difference was that only the detained, detained 2 and family members samples included individuals who were divorced or the widow or widower of a person affected by the criminal justice system. In contrast, criminal justice respondents were either married or single. Although the interview did not focus on the circumstances leading up to the divorce or loss of his/her partner, respondents also included partners who disappeared and could therefore speak of the strain families living in vulnerable areas were under.

4.4 Number of children It was important to determine whether respondents had children and the ages of those children considering the potential medium- and long-term consequences to children – both positive and negative – of being in an environment where safety could not be guaranteed while being subjected to positive and negative messaging. Even more notable were the experiences that would play an important role in the formative years of the next generation.

Research on political socialisation identified parents as one of the primary socialisation agents of children, especially in the formative years before school and friends become more prominent. Therefore, from a very young age, children are susceptible to be influenced by the views of parents (or primary care givers) and their interpretation of the world around them.To reflect the serious nature of this influence, Assistant Commissioner Mark Rowley, who is the leading counter-terror officer in the United Kingdom (UK), said exposing children to extremist propaganda was “equally wicked” as keeping them in environments where there was sexual abuse and should therefore be removed from their parents as: ‘Extremists [reference to both Islamist and right-wing] have the same ambition to create hatred, intolerance and

18 19

FCA_KenyaAnalysis.indd 19 24.1.2020 12.06

Figure 3: Gender of respondents

100 95.04 85.19 90 80.8 80 70 60 50 40 Percentage 30 19.2 20Figure 3: Gender of respondents14.81 10 4.96 100 95.04 0 85.19 Criminal90 Justice (n=108) Detained 2 (n=121) Detained80.8 (n=125) 80 70 Male Female

60 50 4.2 Age 40 Percentage 30 Respondents representing the criminal justice system were predominately19.2 older than 40 years 20 14.81 of age, whereas the individuals that were part4.96 of the detained sample were predominately between the ages10 24 and 35. Individuals who went through the criminal justice system (according 0 to family membersCriminal and Justicefriends) (n=108) as well,Detained also 2represented (n=121) Detainedolder individuals, (n=125) even older than 50 years of age. Similarly, 39.39% of family members interviewed were older than 40 years of age (see Figure 3). Male Female

Figure 2. Gender of respondents Figure 4: Ages of respondents 4.2 Age

OlderRespondents than 50 years representing15.6 the criminal justice system16.54 were predominately 8.13older than 40 years of age,45-49 whereas the individuals14.68 that were 8.66part of the detained5.69 sample were predominately between40-44 the ages 24 and 35.19.27 Individuals who went through12.6 the criminal justice6.5 system (according 38-39 4.59 5.51 6.5 to family36-37 members6.42 and friends)3.94 as5.69 well, also represented older individuals, even older than 50 years 34-35of age. Similarly,7.34 39.39%7.87 of family members12.2 interviewed were older than 40 years of age (see32-33 Figure 3)8.26. 3.15 4.07 30-31 7.34 11.02 10.57 Figure 4:28-29 Ages of respondents6.42 7.87 11.38 26-27 5.5 2.36 8.94 24-25 3.67 7.09 8.13 Older22-23 than 50 years3.94 4.07 15.6 16.54 8.13 20-21 0.9245-493.94 2.44 14.68 8.66 5.69 Younger than 20 years 40-445.52 5.69 19.27 12.6 6.5 38-39 4.59 5.51 6.5 036-37 56.42 103.94 155.69 20 25 30 35 40 45 34-35 7.34 7.87 12.2 32-33 Criminal8.26 Justice (n=109)3.15 4.07Detained 2 (n=127) Detained (n=123) 30-31 7.34 11.02 10.57 Figure 3. Ages of respondents28-29 6.42 7.87 11.38 26-27 5.5 2.36 8.94 4 .3 Marital status24-25 3.67 7.09 8.13 100 22-23 3.94 4.07 90 20-21 0.92 86.363.94 2.44 The majorityYounger thanof 20respondents years 5.52 interviewed5.69 were married, including respondents interviewed 80 representing family members0 and5 friends10 of 15which 5208.14 were25 married30 and35 19.38%40 a widow45 or70 widower (often in relation to detained66.13 2). Although not in the majority, 31.5% of the detained Criminal58.14 Justice (n=109) Detained 2 (n=127) Detained (n=123) and60 29.03% of the detained 2 samples55.91 were single. 50 Figure 5: Marital status of respondents 40 4.3 Marital status 31.5 29.03 30 The majority of respondents interviewed were married, including respondents interviewed 17.02 19.38 20 13.64representing family members and friends of which 58.14 were married and 19.38% a widow 8.66 15 10 5.43 or widower (often in relation to detained 2). Although3.23 not in the majority, 31.5%1.61 3.94 of the detained 0 and 29.03% of the detained 2 samples were single. Single Married Divorced Widow Figure 5: Marital status of respondents Criminal Justice (n=110) Family (n=129) Detained 2 (n=124) Detained (n=127)

Figure 4. Marital status of respondents Although the focus of this study was not to assess the impact that belonging (or the suspect’s 15 sense of belonging) to an illegal organisation had on the person’s marital status, an interesting difference20 was that only the detained, detained 2 and family members samples included 21 individuals who were divorced or the widow or widower of a person affected by the criminal justice system. In contrast, criminal justice respondents were either married or single. Although the interview did not focus on the circumstances leading up to the divorce or loss of his/her partner, respondents also included partners who disappeared and could therefore speak of the FCA_KenyaAnalysis.inddstrain 20 families living in vulnerable areas were under. 24.1.2020 12.06

4.4 Number of children It was important to determine whether respondents had children and the ages of those children considering the potential medium- and long-term consequences to children – both positive and negative – of being in an environment where safety could not be guaranteed while being subjected to positive and negative messaging. Even more notable were the experiences that would play an important role in the formative years of the next generation.

Research on political socialisation identified parents as one of the primary socialisation agents of children, especially in the formative years before school and before friends become more prominent. Children are therefore susceptible to be influenced by the views of parents (or primary care givers) and their interpretation of the world around them from a very young age. To reflect the serious nature of this influence Assistant Commissioner Mark Rowley, who is the leading counter-terror officer in the United Kingdom (UK), said exposing children to extremist propaganda was "equally wicked" as keeping them in environments where there was sexual abuse and should therefore be removed from their parents as: ‘Extremists [reference to both Islamist and right-wing] have the same ambition to create hatred, intolerance and isolation.’33 Another concerning trend in the UK is that extremists remove their children from school and teach them at home where the threat is that they are being radicalised by their parents. According to Mr Rowley, in a survey police discovered that approximately half of all people convicted of terrorist offences in London had been found to be home schooling their children. The negative influence parents could have on the development of their children is a known subject under delinquent behaviour of children where parents were involved in criminal behaviour themselves. Although the focus of this study is not to determine and assess the relation between radicalisation and socialisation of children it is to be expected that children can be positively (children of criminal justice actors) and negatively (children who experienced the impact of the criminal justice system) influenced by the experiences of their parents.

Although 29.17% of detained, 23.58% of detained 2 and 11.11% of respondents representing family members and friends did not have children, the potential impact of children’s experience of the criminal justice system can be severe and requires attention, especially considering the ages of those children.

Figure 6: Number of children of respondents

0.83 15> 1.58 3.36 13-14 0.83 0.94 11-12 0.94 3.17 1.67 9-10 3.173.36 7-8 5.83 10.08 4.72 11.11 5-6 5 9.24 13.2115.08 3-4 20.17 22.5 26.98 35.85 1-2 30.25 34.17 27.78 33.96 None 23.53 29.17 10.3811.11 0 5 10 15 20 25 30 35 40 Percentage

Detained (n=120) Detained 2 (n=119) Family (n=106) Criminal Justice (n=106)

Figure 5. Number of children of respondents As to be expected, what is being experienced by adults have an effect (negative or positive) on their children. Borrowing from criminology it is accepted that parental imprisonment might cause an increase in a child’s antisocial and criminal behaviour. Within right-wing extremism, 33 scholarsisolation.’34 determinedAnother concerning that trend children in the UK growing is that extremists up in remove a family their children with fromextremist school and influences are particularlyteach them at vulnerablehome where the to threat becom is thate theyradicalised are being radicalised themselves by their due parents. to Accordingthe intergenerational to transmissionMr Rowley, in ofa survey ideology police discoveredas radicals that oftenapproximately share halfthe ofsame all people extreme convicted views of terrorist as their of- parents who fences in London had been found to be homeschooling their children. The negative influence parents servecould as have their on the role development models. of Therefore, their children w is hilea known most subject extremists under delinquent may not behaviour come offrom chil- an extremist family,dren where extremist parents familieswere involved do appearin criminal tobehaviour produce themselves. children Although with extremist the focus of views. this study35 is not to determine and assess the relation between radicalisation and socialisation of children, it is to be Figuexpectedre 7: Ages that of children ofcan criminal be positively justice respondents(children of criminal justice actors) and negatively (children who experienced the impact of the criminal justice system) influenced by the experiences of their parents.

As to be expected, what is being experienced by adults can have an effect (negative or positive) on their children. Borrowing from criminology, it is accepted that parental imprisonment might cause an increase in a child’s antisocial and criminal behaviour. Within right-wing extremism, scholars34 determined that children growing up in a family with extremist influences are particularly vulnerable to become radicalised themselves due to the intergenerational transmission of ideology, as radicals often share the same extreme views as their parents who serve as their role models. Therefore, while most extremists may not come from an extremist family, extremist families do appear to produce children with extremist views.35

33 Martin Evans. Terrorists should be treated like paedophiles and have their children removed, top cop sug- gests. The Telegraph, 26 February 2018. Available at https://www.telegraph.co.uk/news/2018/02/26/ terrorists-should-treated-like-paedophiles-have-children-removed/ (accessed on 28 December 2018). 34 Trees Pels & Doret J. de Ruyter (2012). ‘The Influence of Education and Socialization on Radicalization: An Exploration of Theoretical Presumptions and Empirical Research’, Child Youth Care Forum, 41(3): p. 311–325. 35 Sikkens, Elga, M. R. P. J. R. S. van San, S. M. A. Sieckelinck, and Micha de Winter. “Parental Influence on Radicalization and De-radicalization according to the Lived Experiences of Former Extremists and their Families.” Journal for Deradicalization 12 (2017): 192-226. 17 20 21

FCA_KenyaAnalysis.indd 21 24.1.2020 12.06

120

1.79 1.64 2.33 3.03 4.35 100 1.79 1.64 2.33 4.92 3.03 7.14 3.03 9.3 8.7 20 15.15 80 26.79 29.51 6.67 20.93 26.09 6.67

60 20

40 75.76 65.12 62.5 62.3 60.87

46.67 20

0 0-4 5-9 10-14 15-19 20-24 24>

1 child 2 children 3 children 4 children 5 children

Figure 6. Ages of children of criminal justice respondents Although scholars earlier assumed that very young children are not being affected by trauma for being too young to know what happened or not being able to remember what had happened,Although Osofsky scholars36 initially in an assumed article thatin 1995 very young explained children that are not even being in affectedthe earliest by trauma, phases as theyof infant and toddlerare too young development, to understand clear a tramaticassociations event orhave to remember been found the instance, between Osofsky exposure36 in an toarticle violence and postin 1995-traumatic explained symptoms that even in and the earliestdisorders phases (PTSD). of infant Starting and toddler with childevelopment,dren up toclear three associa years- of age, tionsstudies have showed been found that duebetween to the exposure very rapid to violence and complex and post-traumatic changes duringsymptoms the and first disorders three years of life,(PTSD). developmental Starting with childrenfactors upwill to influence three years the of age,young studies child's showed perception that due andto the experience very rapid and of the traumacomplex associated changes with during violence. the first threeInfants years show of life,increased developmental irritability factors and will sleepinfluence disturbances the young as child’s perception and experience of the trauma associated with violence. Infants who experienced well asa tramatic fear of event being show alone. increased Exposure irritability, to sleeptrauma disturbances interferes and with fear their of being normal alone. development Exposure to of trust traumaand theinterferes later with emergence their normal of development independence of trust andthrough the later explor emergenceing. Sleep of independence disturbances, nightmaresthrough and exploring. other Sleep manifestations disturbances, of nightmares increased and anxiety other manifestations are common of atincreased all age anxietys. Children are of pre-schoolcommon and at allschool ages. goingChildren ages of pre-school who are and exposed school going to violence ages who are are less exposed likely to toviolence explore are their physicalless environmentlikely to explore and their that physical may environment, lead to several which candifficulties. lead to severalOn the difficulties. one hand, On children the one may in laterhand, years children avoid may, behaviour in later years, associated avoid behaviour with associatedthe trauma; with or the on trauma the other or on thehand other intentionally hand, get involvedintentionally in getactivities involved associatedin activities associatedwith the withtrauma. the trauma. The relationship The relationship with with the the person person (for example(for example parent) parent) against against whom whom violence violence was was directed directed will will also influence influence the the path path the thechild child may may decidedecide upon upon later later in in life. Secondly, Secondly, when when assessing assessing the effects the ofeffects violence of on violencechildren, the on emotional children the emotionalstate ofstate the child’s of the parents child's or caregiversparents orshould caregivers also be taken should into consideration.also be taken When into parents consideration. and/ or caregivers are numbed, frightened and depressed themselves, the child cannot depend on the Whenemotional parents support,trust and/or caregivers or security thatare generallynumbed, come frightened from these an caregivers.d depressed Furthermore, themselves, children the at child cannotany depend age may onwithdraw the emotional and show disorganizedsupport or behaviour.trust on security The constant that barragecome fromof fear these and violence caregivers . Furthermorein the community, children may at steerany ageparents may to communicatewithdraw and helplessness show disorganized and hopelessness behaviour. to their children.The constant barrageBeing of exposed fear and to violence violence and in trauma the community directly impacts may onsteer the parentsability of theto communicatechild to have and helplessness show and hopelessnessempathy and other to their prosocial children. behaviour. Being Furthermore, exposed itto can violence result in and increased trauma aggression directly and impact vio- s on the abilitylence. ofIn the the instances child to wherehave theand parent show displayed empathy severe and verbalother and/orpro-social physical behavio outburstsur. Furthermore,of anger it canand result violence, in increased boys externalised aggression responses, and violencewhereas girls. In demonstratedthe instances more where internalising the parent responses. displayed severe verbal and/or physical outbursts of anger and violence, boys externalised responses, 36 Osofsky, Joy D. “The effect of exposure to violence on young children.” American Psychologist 50, no. 9 (1995): 782.

22 18 23

FCA_KenyaAnalysis.indd 22 24.1.2020 12.06

whereas girls demonstrated more internalising responses. On the positive side, Osofsky noted that children exposed to violence are better equipped to cope with violence if they have a supportive person, have a protected place in the neighbourhood that provides a safe haven from violence, and have the individual adaptable temperament or intelligence to make sense of their experiences.

Considering the experiences of adults throughout the study (in particular detained respondents, their families and especially where a parent or a close relative disappeared or was killed), it would be a mistake not to consider the psychological impact it had on children. This was particularly the case when noting that children of detained respondents (Figure 7) were much younger than those of criminal justice respondents (Figure 6).

Figure 8: Number of children of individuals who experienced the criminal justice system (detained)

120

100 2.04 3.51 3.57 6.25 3.51 12.24 10.71 16.67 12.5 25 21.05 80 20.41 25

37.5 25 60 50

40 71.93 65.31 60.71 50 20 43.75 33.33

0 0-4 5-9 10-14 15-19 20-24 24>

1 child 2 children 3 children 4 children

Figure 7. Number of children of individuals who experienced the criminal justice system (detained) The potential impact the incarceration of a parent might have on the development of a child is summarised in Table 1.

On the positive side, Osofsky noted that children exposed to violence are better equipped to cope Tablewith 1: violencePossible developmental if they have effectsa supportive on children person, of parents have arresteda protected and incarcerated place in the37 neighbourhood that provides a safe haven from violence and have the individual adaptable temperament or intelligence Developmentalto make sense of theirDevelopmental experiences. Developmental Influencing Factors Effects Stage Characteristics Tasks InfancyConsidering (0-2 years)the experiences Total dependency of adults throughout Attachment the study and (in particularParent detained-child respondents,Impaired their parent- families and especially where a parent or a closetrust relative disappearedseparation or was killed), it wouldchild be a mistake not to consider the psychological impact it had on children. This was particularly thebonding case Earlywhen childhood noting that (2 children- Increased of detained respondentsSense (Figure of autonomy, 7) were muchParent younger-child than those ofAnxiety, crimi- 6nal years) justice respondentsperception (Figure 6 ). independence and separation; developmental and mobility; initiative Trauma regression, acute The potential impact theincomplete incarceration of a parent might have on the development of a childtraumatic is sum- stress, marised in Table 1. individualization survivor’s guilt from Table 2 describes theparent type of behaviour associated with the relevant type of trauma a child might display. Although the impact of trauma or response to the incarceration of a parent differs from person-to-person and situation-to-situation, the overall fear is that the child is at risk for a number of negative behaviours that, in some instances, , especially in the absence of positive intervention, can lead to school failure, delinquency and intergenerational incarceration. 19

Although the impact of interactions with law enforcement and the criminal justice system has is not known, it is to be expected that these experiences could have a profound impact on the lives of children when their parents are arrested and incarcerated for terrorism-related offences. Considering that interaction with security forces plays an overwhelming role in the radicalisation process of adults, the fact is that there is no clear policy or standing operating procedures on how officials should respond and how to treat the children of individuals arrested or incarcerated on terrorism-related offences. This needs urgent attention to prevent second generation radicalisation;it should start with the conduct of security forces.

22 23

FCA_KenyaAnalysis.indd 23 24.1.2020 12.06 Table 1. Possible developmental effects on children of parents arrested and incarcerated37

Developmental Developmental Developmental Tasks Influencing Factors Effects Stage Characteristics Infancy (0-2 years) Total dependency Attachment and trust Parent-child Impaired parent-child separation bonding Early childhood Increased Sense of autonomy, Parent-child Anxiety, (2-6 years) perception and independence and separation; Trauma developmental mobility; incomplete initiative regression, acute individualization from traumatic stress, parent survivor’s guilt Middle childhood Increased Sense of industry, Parent-child Acute traumatic (7-10 years) independence, ability ability to work separation, enduring stress and reactive to reason, importance productively trauma behaviours of peers Early adolescence Increased abstract Ability to work Parent-child Rejection of limits on (11-14 years) thinking, future productively with separation, enduring behaviour, trauma- oriented behaviour, others, control of trauma reactive behaviours aggression, puberty emotions Late adolescence Emotional crisis and Achieves identity, Parent-child Premature termination (15-18 years) confusion, adult engages in adult work separation, enduring of parent-child sexual development, and relationships, trauma relationship; abstract thinking, resolves conflicts with intergenerational independence family and society crime and incarceration

Table 2. Intergenerational behaviours, crime and incarceration38

Childhood Trauma Emotional Response Reactive Behaviour Coping Pattern Criminal Activity Physical Abuse Anger Physical aggression Fighting with peers Assault Parent-child Sadness, grief Withdrawal Substance abuse Drug possession separation Witness to violence Anxiety Hypervigilance Group-violence Accessory to homicide activity

37 Simmons, Charlene Wear. “Children of Incarcerated Parents.” California Research Bureau, California State Library note 7, no. 2 (2000): p. 5. 38 Simmons, Charlene Wear. “Children of Incarcerated Parents.” California Research Bureau, California State Library note 7, no. 2 (2000): p. 5.

24 25

FCA_KenyaAnalysis.indd 24 24.1.2020 12.06 5. Education

The relationship between the level of education and radicalisation presented two scenarios: the first found a correlation between limited education (not completing secondary school) and radicalisation, while the second determined that individuals who received a tertiary education were more vulnerable to extremist ideology than those with limited education. 57.94% of respondents who represented the detained and 73.79% of the detained 2 sample did not complete secondary school. In comparison 18% of respondents representing the criminal justice system indicated that the highest level of secular education they received fell into this category, whereas 69.37% attained further education.

Post grad 18.02

Degree 1.94 40.54 11.21 Diploma 6.8 7.21 7.48 Certificate 3.88 3.6 Finish 23.36 13.59 secondary 12.61 11.21 Secondary 28.16 13.51 16.82 Finish primary school 15.53 1.8 29.91 Primary 30.1 2.7 0 5 10 15 20 25 30 35 40 45

Detained (n=107) Detained 2 (n=103) Criminal Justice (n=111)

Figure 8. Education Considering above differences, respondents were asked whether they felt that they had the same opportunities as others to receive an education. Although the majority of both criminal justiceConsidering and theabove detained differences, samples respondents indicated were that asked they whether were they not feltthe that victim theys ofhad discrimination, the same 47.93%opportunities of respondents as others to receive who went an education. through thAlthoughe criminal the majorityjustice system of both (criminaldetained justice sample) and the versus 26.61%detained samplesof the indicatedcriminal thatjustice they sample were not werethe victims under of discrimination,the impression 47.93% that of they respondents had been discriminatedwho went through against the criminal in not havingjustice systemthe same (detained opportunities sample) versusas others 26.61% in theof the community. criminal justice sample were under the impression that they had been discriminated against by not having the same Figureopportunities 10: Access asto theothers same ineducation the community. opportunities

In100% a follow-up question, the two samples were asked to identify the source of the discrimination they experienced;90% both identified economic circumstances. However, not having the financial resources to80% study further in a community where the majority found themselves in similar circumstances could 52.07 not70% be classified as discrimination per se. Gender, age, ethnicity, religion and disabilities being the source60% of discrimination73.39 were however of greater concern, although limited under both samples. 50% It 40%is however, not only receiving equal education opportunities, irrespective of religious or ethnic/ tribal30% orientation, but more importantly, having equal access to employment opportunities that will 47.93 counter20% perceptions of marginalization. For example, participants of a focus group discussion in Kilifi noted10% the evidence of26.61 ethnic and religious discrimination in the county which has led to a lack of employment.0% Cases had been reported of qualified Muslim youths who were denied jobs that were eventually givenCriminal to less Justice qualified (n=109) Christian youths. Detained (n=121)

Yes No 24 25

In a follow-up question the two samples were asked to identify the source of the discrimination they experienced, of which both identified economic circumstances. However, not having the financial resources to study further in a community where the majority found themselves in similar per se FCA_KenyaAnalysis.indd circumstances 25 could not be classified as discrimination . Gender, age, ethnicity, religion 24.1.2020 12.06 and disabilities being the source of discrimination were however of greater concern, although limited under both samples.

Figure 11: Reasons for not receiving the same education opportunities

Post grad 18.02

Detained (n=107) Detained 2 (n=103) Criminal Justice (n=111)

Considering above differences, respondents were asked whether they felt that they had the same opportunities as others to receive an education. Although the majority of both criminal justice and the detained samples indicated that they were not the victims of discrimination, 47.93% of respondents who went through the criminal justice system (detained sample) versus 26.61% of the criminal justice sample were under the impression that they had been discriminated against in not having the same opportunities as others in the community.

Figure 10: Access to the same education opportunities

100% 90% 80% 52.07 70% 73.39 60% 50% 40% 30% 47.93 20% 10% 26.61 0% Criminal Justice (n=109) Detained (n=121)

Yes No

Figure 9. Access to the same education opportunities In a follow-up question the two samples were asked to identify the source of the discrimination they experienced, of which both identified economic circumstances. However, not having the financial resources to study further in a community where the majority found themselves in similar circumstances could not be classified as discrimination per se. Gender, age, ethnicity, religion 63.16 andEconomic disabilities being the source of discrimination were however of greater68.75 concern, although limited under both samples. Disabilities 2.63 Figure 11: Reasons for not receiving the same education opportunities 10.53 Ethnic 12.5

7.89 Religion 12.5

Age 6.25 21 Gender 18.42

0 10 20 30 40 50 60 70 80

Detained (n=38) Criminal Justice (n=16)

Figure 10. Reasons for not receiving the same education opportunities It is however not only receiving equal education opportunities, irrespective of religious or ethnic/tribal orientation, but more important having equal access to employment opportunities that will counter perceptions of marginalization. For example, participants being part of a focus group discussion in Kilifi, noted that there was evidence of ethnic and religious discrimination in the county which has led to lack of employment. Cases had been reported of qualified Muslim youths who were denied jobs that were eventually given to less qualified Christian youths. 6. Religion

The majority of respondents across all three samples received very basic religious education. Although criminal justice respondents identified with being Christian and were not supposed to rely on religion to fulfil their responsibilities, it is important to note that the two community- related samples progressed further. However, considering the importance of a religious education to be able to counter the influences of an extremist ideology, the focus of the rest of this section 26will be on the respondents, as part of the ‘detained’ sample. 27

Figure 12: Level of religious education received

FCA_KenyaAnalysis.indd 26 24.1.2020 12.06

22 6. Religion

The majority of respondents across all three samples received very basic religious education. Al- though criminal justice respondents identified with being Christian and were not supposed to rely on religion to fulfil their responsibilities, it is important to note that the two community-related samples progressed further. However, considering the importance of a religious education to be able to counter the influences of an extremist ideology, the focus of the rest of this section will be on the respondents, as part of the ‘detained’ sample.

25.59% of the detained sample received religious education for longer than seven years;41% also studied the interpretation (Tafsir) of the Qur’an. However, despite receiving formal religious education, only 19% indicated that they understood what they read in the Qur’an and 26% more often than not, while 27% outright answered in the negative (see Figure 12).

Although only 35% internalised conflicting positions to those presented by a religious scholar, the majority was willing to discuss those matters with the particular religious scholar, a fact that was considered as a positive occurrence (see Figure 13). However, the concern rests with the credentials and interpretation of religious texts on the part of the particular religious scholar an individual may interact with. In other words, if the religious scholar was radicalised himself, he would relay his interpretation to a person searching for understanding and the latter would be convinced if he/she was not equipped to form his/her own opinion. There is a substantial portion of respondents ready to discuss questions as and when they arise with other sources (other scholars, friends, family).

It is not only the way adults deal with conflicting messages, but more importantly howchildren who are very susceptible to conflicting messaging handle the messages since they do not have the ability to differentiate between ‘correct’ or ‘incorrect’ interpretation, especially if it is coming from a person with authority. For example, during one focus group discussion in Kilifi, a reference was made to children as young as 14 years that were no longer safe as a class six pupil in Ganda Primary school was caught with a knife in class. The pupil wore a Kanzu that had Salafi Jihad written on the collar. When interrogated, he said that he had made a decision to defend his religion by eliminating any Christian pupil who spoke ill of Prophet Mohamed (SAW). He cited a verse in the Quran where the Prophet (SAW) was encouraging people to fight in the Jihad. There seems to be an abundance of alternative narratives available should the misinformed youth care to engage in discussions with other religious scholars, or family and friends. This seems to be the best venue for future campaigns proposing narratives alternative to extremist ones.

6.1 Religious Identity When respondents were asked to state their religious affiliation, 75% of criminal justice respondents indicated they were Christian, while the majority of respondents representing those who went through the criminal justice system (detained and detained 2) identified themselves as being Muslim (92.13% and 85.6%).

Assessing the importance respondents placed on their religious identity, those interviewed were asked to rate the importance of family, friends, religion, ethnicity and nationality. After family, respondents included in the detained and family and friends samples rated (expressed through percentages) their association with members of the same religion in their country highest. Criminal justice actors rated religion as third most important following their family and colleagues.

26 27

FCA_KenyaAnalysis.indd 27 24.1.2020 12.06

100% 3.85 8.14 8.93 1.92 3.85 90% 1.16 1.16 100% 7.15 7.69 3.85 8.14 8.93 1.92 15.12 3.57 80% 3.85 3.57 1.16 5.77 90% 1.16 1.79 7.15 5.77 7.69 7.14 70% 8.14 15.12 3.57 80% 3.57 5.77 11.54 2.33 7.14 1.79 5.77 60% 6.98 7.14 70% 8.14 10.71 9.62 2.33 11.54 10.47 7.14 50% 60% 6.98 10.71 16.07 9.62 40% 10.47 19.23 50%

30% 16.07 40% 19.23 46.51 20% 30% 33.93 30.77 46.51 10% 20% 33.93 Although only 35% internalised conflicting positions30.77 to those presented by a religious scholar, 0% Criminal Justice (n=86) 10% Detained 2 (n=56)theAlthough majority onlyDetained was 35% willing (n=64) internalise to discussd conflictingthose matters positions with the to particularthose presented religious by scholar a religious, a fact scholar, that wasthe consideredmajority was as willing a positive to discuss occurrence those matters(see Figure with 13)the . particularHowever, religiousthe concern scholar rests, a with fact the that Form 1 2 3 4 5 6 7 8 credentials9was10 consideredUniversity and interpretation as a positive of occurrence religious text(sees Figureon the 13)part. However,of the particular the concern religious rests scholarwith the 0% thecredentials person might and interainterpretationct with. In of other religious words, text if thes on religious the part scholar of the was particular radicalised religious himself, scholar he Criminal Justicewould (n=86) relay hisDetained interpretation 2 (n=56) to a personDetained searching (n=64) for understanding and the latter would be 25.59% of the detained sample received religious educationthe person for longer might interathan ctseven with. yearsIn other and words, if the religious scholar was radicalised himself, he Form convinced1would2 3 relay if4 he/she his5 interpretation6 was7 not8 equipped9 to10 a personUniversity to form searching his/her ownfor understandingopinion. There andis a substantialthe latter would portion be 41% also studied the interpretation (Tafsir) of the Qur’an.ofconvinced respondents However, if he/she despiteready was to receiving notdiscuss equipped questions formal to form as andhis/her when own they opinion. arise There with isother a substantial sources (otherportion religious education only 19% indicatedFigure 11. Level that of religiousthey underst educationscholars,oo receivedd what friends, they family).read in the Qur’an and of respondents ready to discuss questions as and when they arise with other sources (other 26% more often than not,25.59% while 27% of theoutright detained answered sample inscholars, thereceived negative friends, religious (seefamily). Figureeducation 12) . for longer than seven years and Figure 14: Reaction when confronted with conflicting preaching 41% also studied the interpretation (Tafsir) of the Qur’an. However, despite receiving formal Figure 13: Do you understand whatreligious you read education only 19% indicatedFigure 14: Reaction that theywhen confronted underst withoo dconflicting what preachingtheyChange read position in the Qur’an and 26% more often than not, while 27% outright answered in the negative (see2% FigureWalk 12)away. Educated person Change position 10% Keep to yourself 2% Walk away 2% 23% Figure 13: Do you understand what you readYes 10% 19% Keep to yourself 23% No Educated person 2% 27% Yes 19% No 27% Discuss with Discuss with other scholar religious scholar(s) Discuss33% with Discuss12% with other scholar religious scholar(s) 33% More often 12% 26% Discuss with family 6% Discuss with family Discuss with Less often 6% More oftenfriends 26%Discuss14% with 26% friends 14% It is however not only the way adults deal with conflicting messages, but even more important ly Less often childrenIt is26% however who are not very only susceptiblethe way adults to conflicting deal with messaging conflicting sincemessages, they do but not even have more the important ability toly Figure 12. Do you understand what differentiateyou read betweenFigure ‘correct’ 13. Reaction or ‘incorrect’when confronted interpretation, with conflicting especially preaching if it is coming from a children who are very susceptible to conflicting messaging since they do not have the ability to persondifferentiate with authority. between For ‘correct’ example, or ‘incorrect’during one interpretation, focus group discussionespecially in if Kilif it isi referencecoming from was a made to children as young as 14 years that were no longer safe as a class six pupil in Ganda 28 person with authority. For example, during one focus group discussion in Kilifi reference was 29 Primarymade toschool children was ascaught young with as 14a knife years 23in thatclass. were The no pupil longer wore safe a Kanzuas a class that sixhad pupil Salafi in JihadGanda writtenPrimary on schoolthe collar. was When caught interrogated, with a knife he in said class. tha Thet he hadpupil made wore a adecision Kanzu tothat defend had hisSalafi religion Jihad bywritten eliminating on the anycollar. Christian When pupilinterrogated, who spoke he saidill of tha Prophett he had Mohamed made a decision(SAW). toHe defend cited ahis verse religion in theby Quran eliminating where any the ChristianProphet (SAW)pupil who was spoke encouraging ill of Prophet people Mohamed to fight23 (SAW).in the Jihad. He cited There a verseseems in tothe be Quran an abundance where the ofProphet alternative (SAW) narratives was encouraging available people should to the fight misinformed in the Jihad. youth There care seems to FCA_KenyaAnalysis.indd 28 engageto be a inn abundancediscussions withof alternative other religious narratives scholars, available or family should and thefriends misinformed24.1.2020. This seems 12.06 youth to becare the to bestengage venue in for discussions future campaigns with other proposing religious scholars,narratives or alternative family and to friends extremist. This ones seems. to be the best venue for future campaigns proposing narratives alternative to extremist ones. 6 .1 Religious Identity 6.1 Religious Identity When respondents were asked to state their religious affiliation 75% of criminal justice respondentsWhen respondents indicated were that theyasked were to Christian,state their while religious the majority affiliation of respondents 75% of criminal representing justice thoserespondents who went indicated through that the they criminal were justiceChristian, syst whileem (thedetained majority and of respondentsdetained 2) representing identified themselvesthose who as went being through Muslim (92.13%the criminal and justice 85.6%). syst em (detained and detained 2) identified themselves as being Muslim (92.13% and 85.6%). Figure 15: Religious affiliation of respondents

Figure 15: Religious affiliation of respondents

24 24 After establishing the importance of religion in the lives of respondents, the same respondents were asked to assess the other or ‘them’. Criminal justice respondents overall expressed a stronger sense of ‘them’ in reference to other religions (see Figure 16). It is important to remind the reader that Christianity was predominately eminent in criminal justice respondents and in an area where Islam is the majority religion this sentiment may be expected considering the importance of religious identity. Secondly, it also reflects an existing threat perception amongst Christians living in vulnerable areas, being a section of the public that bears the brunt of attacks associated with al-Shabaab and a broader Salafi jihadi ideology. This is crucial as it is a known strategy of attackers to separate Chris- tians from Muslims. When the former refuse or are unable to recite the shahada (declaration of faith), they are executed. This can be exemplified through the Garissa University attack when more than 148 Christian students were killed, as well as in al-Shabaab ‘roadblocks’. Consequently, approximately 1,100 Christian teachers fled or had requested to be transferred to safer areas. According to the Kenya National Union of Teachers, non-local teachers make up approximately 60% of the teachers in the area (after being transferred due to a shortfall of teachers in the area), but they still experienced discrimination. In a statement made in 2018 by Wilson Sossion, the secretary-general of the union: “We have heard cases of locals pretending to be al-Shabaab or conspiring with the group to target non-local teachers. Students have attacked teachers through stoning, clobbering and issuing verbal attacks.” Cyntia Chepkemoi, a teacher, said she was forced to adopt customary Muslim dress and was told that Sharia law prohibited her from correcting her male students in class.39 Focus group discussions in Wajir and Garissa confirmed the rift between Muslims and Christians in the region was widening, with Christians being targeted by the terror groups and Muslims branded as sympathisers. As a result, education, economy, health and social life of the region have been negatively affected.

Christians are not the only group feeling persecuted by the ‘other’, as a detained respondent ex- Although only 35% internalised conflicting positions to those presented by a religious scholar, plained: “Muslims are presumed to be terrorists by security officers. Me being a Muslim will be the first the majority was willing to discuss those matters with the particular religious scholar, a fact that one to be held responsible when I report such a case to police [coming forward with information].” was considered as a positive occurrence (see Figure 13). However, the concern rests with the credentials and interpretation of religious texts on the part of the particular religious scholar Recognising the importance of religious identity does not necessarily contribute to conflict. To de- the person might interact with. In other words, if the religious scholar was radicalised himself, he termine the influence of religious divide in contributing to conflict, respondents were asked to rate would relay his interpretation to a person searching for understanding and the latter would be a series of potential sources of domestic instability, including religion. In that regard, the detained convinced if he/she was not equipped to form his/her own opinion. There is a substantial portion sample considered religion to play a more prominent role than ethnicity (to be discussed in a later of respondents ready to discuss questions as and when they arise with other sources (other section) in an area where Islam was practiced by the majority. This contributed to the question to scholars, friends, family). what extent religion plays a role in radicalisation. According to the family members sample, religious ideology was placed in the fourth position (60.78% after anger towards the police, financial incen- Figure 14: Reaction when confronted with conflicting preaching tives and anger towards government) and detained respondents in the fifth position (52.3% after Change position anger towards the police and government, financial incentives and treatment at the hands of security 2% Walk away 10% forces since arrest) contributing to radicalisation. In contrast, criminal justice respondents placed re- Keep to yourself ligious ideology second (65.34%) after financial incentives (see Table 4). Due to the large difference 23% between detained and criminal justice system representatives regarding the perception on the role religious ideology plays in contributing to domestic instability, religious education and how it differs from extremist ideologies should be considered.

Discuss with Discuss with other scholar religious scholar(s) 33% 12%

Discuss with family 6% Discuss with friends 14% 39 Doreen Ajiambo. In northern Kenya, al-Shabaab militants target Christian teachers. Sight Magazine, It is however not only the way adults deal with conflicting messages, but even more importantly 15 January 2019. Available at https://www.sightmagazine.com.au/news/11244-in-northern-kenya- al-shabab-militants-target-christian-teachers (accessed on 16 January 2019). children who are very susceptible to conflicting messaging since they do not have the ability to differentiate between ‘correct’ or ‘incorrect’ interpretation, especially if it is coming from a 28 person with authority. For example, during one focus group discussion in Kilifi reference was 29 made to children as young as 14 years that were no longer safe as a class six pupil in Ganda Primary school was caught with a knife in class. The pupil wore a Kanzu that had Salafi Jihad written on the collar. When interrogated, he said that he had made a decision to defend his religion by eliminating any Christian pupil who spoke ill of Prophet Mohamed (SAW). He cited a verse in the Quran where the Prophet (SAW) was encouraging people to fight in the Jihad. There seems to be an abundance of alternative narratives available should the misinformed youthFCA_KenyaAnalysis.indd care to 29 24.1.2020 12.06 engage in discussions with other religious scholars, or family and friends. This seems to be the best venue for future campaigns proposing narratives alternative to extremist ones.

6.1 Religious Identity

When respondents were asked to state their religious affiliation 75% of criminal justice respondents indicated that they were Christian, while the majority of respondents representing those who went through the criminal justice system (detained and detained 2) identified themselves as being Muslim (92.13% and 85.6%).

Figure 15: Religious affiliation of respondents

100% 0.89 0.8 7.87 90% 13.6 80% 70% 60% 75 50% 92.13 85.6 40%100% 0.89 0.8 7.87 30%90% 13.6 20%80% 10%70% 24.11 0%60% 75 Criminal Justice (n=112) Detained 2 (n=127) Detained (n=125) 50% 92.13 40% Muslim Christian Other 85.6

30% 20% Assessing the importance respondents24.11 placed on their religious identity, those interviewed were asked to rate the10% importance of family, friends, religion, ethnicity and nationality. After family, 0% respondents included inCriminal the detainedJustice (n=112) and familyDetained and 2 (n=127)friends samplesDetained rated (n=125) (expressed through percentages) their association with members of the same religion in their country highest. Criminal justice actors rated religion asMuslim thirdChristian most importantOther following their family and colleagues. Religious affiliation of respondents Figure 14. FigureAssessing 16: Importance the importance of religious identity respondents placed on their religious identity, those interviewed were nonasked-local toteachers rate the makeimportance up approximately of family, friends, 60% religion, of the ethnicityteachers andin the nationality. area (after After being family, transferredrespondents due included to a shortfall in the of detained teachers and in the family area), and but friends they experienced samples rated being (expressed discriminated through againstpercentages). In a statement their association made in 2018 with bymembers Wilson of Sossion, the same the secretaryreligion in-general their country of the uhighest.nion: “WeCriminal haveDetained heardjustice (n=139) cases actors4.39 of4.39 localsrated12.28 pretendingreligion as28.07 to third be almost-Shaba importantab 29.5or conspiring following with 19.42their the familygroup toand targetcolleagues. non-local teachers. Students have attacked teachers through stoning, clobbering and issuing verbal attacks.” Cyntia Chepkemoi, a teacher, said she was forced to adopt customary MuslimFigure dress16: Importance and was of religious told that identity Sharia law prohibited her from correcting her male students in Family (n=127) 3.94 9.45 11.81 14.17 23.35 32.28 class.39 Focus group discussions in Wajir and Garissa confirmed that the rift between Muslims and Christians in the region was widening with Christians being targeted by the terror groups and the Muslims branded as sympathisers. As a result, education, economy, health and social Detained (n=139) 4.394.39 12.28 28.07 29.5 19.42 life ofCriminal the Justiceregion (n=111) have3.6 been4.5 7.21 affected negatively.28.83 31.53 24.32

It is, however, not only0% Christians10% 20%feeling30% persecuted40% 50%by the60% ‘other’70%, as a 80%detained90% respondent100% explained: Family“Muslims (n=127) are3.94 presumed9.45 to11.81 be terrorists14.17 by security23.35 officers. Me being32.28 a Muslim will be the first one to be held responsibleNo 1-20% when30-40% I report50-60% such 70-80%a case to90-100% police [coming forward with information].” Figure 15. Importance of religious identity After establishing the importance of religion in the lives of respondents, the same respondents Figure 17Criminal: Identifying Justice (n=111)other religions3.6 4.5 as 7.21'them' 28.83 31.53 24.32 were asked to assess the other or ‘them’. Criminal justice respondents overall expressed a stronger sense of ‘them’ in reference to other religions (see Figure 16). It is important to remind the reader that Christianity0% 10%was predominately20% 30% eminent40% 50%in criminal60% justice70% respondents80% 90% and100% in an area whereDetained (n=134)Islam is 8.93the majorityNo 20.541-20% religion30-40% this22.32 sentiment50-60% 70-80% may25.89 be90-100% expected16.42 considering7.46 the importance of religious identity. Secondly, it also reflects an existing threat perception amongst Christians living in vulnerable areas, being a section of the public that bears the brunt of attacks associatedAfter establishingFamily with (n=129) al-Shabaab the6.98 importance13.95 and a broaderof religion26.36 Salafi in the jihadi lives19.38 ideology.of respondents, This16.28 is ethespecially same17.05 respondentssince it is awere known asked strategy to assessof attackers the other to separate or ‘them’. Christians Criminal from justice Muslims respondents and when overall the former expressed refuse a orstronger are unable sense to of recite ‘them’ the in shahada reference (declaration to other religions of faith) (seethey Figureare executed 16). It isas important witnessed to in remind the GarissatheCriminal reader JusticeUniversity that (n=109) Christianity attack8.26 when 14.68was more predominately than12.84 148 Christian eminent24.77 studentsin criminal were22.94 justice killed respondents, as16.51 well as and in al in- an Shabaabarea where ‘roadblocks’ Islam is . theConsequently, majority religion approximately this sentiment 1,100 may Christian be expected teachers consideringfled or had the requestedimportance to ofbe religioustransferred0% identity.10% to safer20% Secondly, areas.30% According it 40%also reflect50% to thes an60% Kenya existing70% National threat80% Union perception90% of Teachers100% amongst, Christians living in vulnerable areas, being a section of the public that bears the brunt of attacks No 1-20% 30-40% 50-60% 70-80% 90-100% associated with al-Shabaab and a broader Salafi jihadi ideology. This is especially since it is a known strategy of attackers to separate Christians from Muslims and when the former refuse Figure 16. Identifying other religions as ‘them’ 25 Recognisor areing unable the importanceto recite the of shahada religious (declaration identity does of faith) not necessarily they are executed contribute as towitnessed conflict. inTo the determineGarissa theUniversity influence attack of religious when more divide than in contributing148 Christian to conflict,students respondentswere killed, wereas well asked as in to al - rateShabaab a30 series ‘ roadblocks’of potential. Consequently,sources of domestic approximately instability 1,100, includ Christianing religion. teachers Seeing fled that or the had 31 detainedrequested sample to be consideredtransferred religionto safer areas.to play According a more toprominent the Kenya role National than ethnicityUnion of Teachers(to be , discussed in a later section) in an area where Islam was practiced by the majority, contributed to the question to what extent religion plays a role in radicalisation. According to the family members sample, religious ideology was placed in the fourth position (60.78% after anger FCA_KenyaAnalysis.inddtowards 30 the police, financial incentives and anger towards government) and detained2524.1.2020 12.06 respondents in the fifth position (52.3% after anger towards the police and government, financial incentives and treatment at the hands of security forces since arrest) contributing to radicalisation. In contrast, criminal justice respondents placed religious ideology second (65.34%) after financial incentives (see Table 4). Due to the large difference between detained and criminal justice system representatives regarding the perception on the role religious ideology plays in contributing to domestic instability, religious education and how it differs from extremist ideologies should be considered. 7. Government and National Identity

The National Strategy to Counter Violent Extremism that Kenya introduced in September 2016 listed initiatives to promote patriotism for Kenya’s nationhood as the second of nine priorities

26 7. Government and National Identity

The National Strategy to Counter Violent Extremism introduced by Kenya in September 2016 listed initiatives to promote patriotism for Kenya’s nationhood as the second of nine priorities set as outcomes of the strategy. Communal knowledge of this strategy will be presented in a later section; this section first intends to determine respondents’ sense of a Kenyan national identity. When respondents were asked to rate their association of being Kenyan as part of ‘us,’ it became clear that criminal justice actors had an established national identity as demonstrated by 66.35% respodants rated the importance of being Kenyan between 70-100% versus 43.42% amongst family and friends and 26.82% amongst individuals who went through the criminal justice system. The need to serve was confirmed when criminal justice respondents were asked why they became part of the criminal justice system: 89% rated serving his/her country and community and 82% rated protecting his/her country and community between 70 and 100%. As reflected in Figure 14, it is important to note that respondents (as part of the detained sample) were not necessarily not associating with being Kenyan as 30.7% rated their national identity between 50 and 60%. Of the three samples, respondents being part of the family sample associated the least with being Kenyan as 23.25% either answered ‘no’ or rated their association between 1 to 20%. Anger towards the Kenyan government and treatment at the hands of security forces explain this negative sentiment.

Following the previous question to assess national identity, a later question was put to respondents to indicate how proud the person was of being Kenyan. Confirming above assessment, 62.38% of criminal justice respondents indicated that they were most proud (10/10) of being Kenyan. 67.71% of family members rated being proud of being Kenyan between seven and ten, while 58.4% of detained respondents provided the same assessment. Despite the fact that most respondents representing all three samples were very proud of being Kenyan, 11.5% indicated that they were not at all proud of being Kenyan (1/10).

Despite the pride in being Kenyan and a strong national identity, 67.6% of criminal justice respondents identified the Kenyan government as ‘them’ or the out-group between 70% and 100%. In contrast, 34.33% of family members and friends and 25.55% of respondents in the detained sample reflected the same negative sentiment towards the government. It is Interesting to note the negative judgement about the government due to experiences with the criminal justice system is higher among families than in the detained sample, potentially indicating a kind of multifaceted negative effect following any authority’s misbehaviour.

Although expressing a strong sentiment in refering to the Kenyan government as ‘them,’ criminal justice respondents did not express the same level of distrust in the president, local and national governments. However, criminal justice respondents expressed very limited trust in politicians overall as 41.25% of respondents rated their trust between 10 and 20%.

30 31

FCA_KenyaAnalysis.indd 31 24.1.2020 12.06

set as outcomes of the strategy. Communal knowledge of this strategy will be presented in a later section, while this section first intends to determine respondents’ sense of a national identity being Kenyan. When respondents were asked to rate their association of being Kenyan as part of ‘us’ it became clear that criminal justice actors had an established national identity as 66.35% rated the importance of being Kenyan between 70-100% versus 43.42% amongst family and friends and 26.82% amongst individuals who went through the criminal justice system. The need to serve was confirmed when criminal justice respondents were asked why they became part of the criminal justice system: 89% rated serving his/her country and community and 82% rated protecting his/her country and community between 70 and 100%. As reflected in Figure 14 it is important to note that respondents (as part of the detained sample) weren’t necessarily not associating with being Kenyan as 30.7% rated their national identity between 50 and 60%. Of the three samples, respondents being part of the family sample associated the least with being Kenyan as 23.25% either answered ‘no’ or rated their association between 1 to 20%. Anger towards the Kenyan government and treatment at the hands of security forces explain this negative sentiment.

Figure 18: Respondents' association with being Kenyan (national identity)

Detained (n=138) 5.26 14.91 22.81 30.7 18.12 8.7

Family (n=129) 10.85 12.4 10.85 23.26 21.71 21.71

Criminal Justice (n=107) 5.61 7.48 19.63 39.25 27.1

0.93 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% No 1-20% 30-40% 50-60% 70-80% 90-100%

FollowingFigure 17. Respondents’the previous association question with beingto assess Kenyan (nationalnational identity) identity, a later question was put to Criminal Justice (n=101) 2 4 2 4 13 9 4 62 respondents to indicate how proud the person was of being Kenyan. Confirming above assessment, 62.38% of criminal justice respondents indicated that they were most proud (10/10) of beingDetained Kenyan. (n=113) 67.71%12 of2 fami7 ly6 members11 rated4 12being4 proud4 of being 38Kenyan between seven and ten, while 58.4% of detained respondents provided the same assessment. Despite

the fact thatFamily most (n-96) respondents1 4 3 4 representing13 7 all three13 samples11 were5 very proud39 of being Kenyan, 11.5%Criminal indicatedJustice (n=101) tha2 t 4they2 4 were13 not at all9 proud4 of being Kenyan (1/10).62 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Figure 19: Proud of being Kenyan Detained (n=113) 12 2 7 16 2 311 4 54 6 127 8 49 410 38

Despite Familythe (n-96)pride1 in4 being3 4 Kenyan13 and7 a strong13 national11 5 identity, 67.6%39 of criminal justice respondents identified the Kenyan government as ‘them’ or the out-group between 70% and 100%. In contrast,0% 34.33%10% of 20%family 30%members40% and 50%friends60% and 25.55%70% 80%of respondents90% 100% in the detained sample reflected the same1 2 negative3 4 5 sentiment6 7 8 towards9 10 the government. It is Interesting to note that the negative judgement about the government due to experiences with the criminal justice system is higher among families than in the detained sample, potentially indicating a kind Figure 18. Proud of being Kenyan Despiteof multi facetedthe pride negative in being effect Kenyan following and a any strong authority’s national misbehaviour identity, 67.6%. of criminal justice respondents identified the Kenyan government as ‘them’ or the out-group between 70% and 100%.Figure 20 In: Identifyingcontrast, the 34.33% Kenyan government of family as 'them'members and friends and 25.55% of respondents in the detained sample reflected the same negative sentiment towards the government. It is Interesting to note that the negative judgement about the government due to experiences with the criminal justice Detainedsystem (n=137) is higher7.83 among families20.87 than in the20 detained sample,26.96 potentially13.14 indicating12.41 a kind27 of multifaceted negative effect following any authority’s misbehaviour.

Figure 20: IdentifyingFamily (n=134) the Kenyan7.46 government13.43 as 'them'20.15 24.63 11.94 22.39

Criminal Justice (n=108) 7.41 6.48 5.56 12.96 35.19 32.41 Detained (n=137) 7.83 20.87 20 26.96 13.14 12.41

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Family (n=134) 7.46 13.43No 1-20% 20.1530-40% 50-60% 24.6370-80% 90-100%11.94 22.39

ACriminalFigurelthough Justice 19. Identifyingexpressing (n=108) the7.41 aKenyan strong6.48 government5.56 sentiment12.96 as ‘them’ in refering to35.19 the Kenyan government32.41 as ‘them’, criminal justice respondents did not express the same level of distrust in the president, local and national governments. However,0% 10% criminal20% justice30% respondents40% 50% expressed60% very70% limited80% trust90% in politicians100% overall as 41.25% of respondentsNo 1-20% rated30-40% their 50-60%trust between70-80% 1090-100% and 20%. 32 33 7.1 Ethnic Identity AKenya’slthough expressingpopulation ais strongdivided sentiment into more in thanrefer 40ing ethnic to the groups Kenyan (the government most prominent as ‘them’, presented criminal in justiceTable respondents 6) belonging did to notthree express linguistic the families: same level The of Bantu, distrust the in Cushitic, the president, and the local Nilotic. and Language national governments.traditionally hasHowever, been thecriminal primary justice characteristic respondents of ethnicexpressed identity. very Bantu limited-speaking trust in Kenyans politicians are overall as 41.25% of respondents rated their trust between 10 and 20%. FCA_KenyaAnalysis.indd divided 32 into three different groups: the western group (Luhya); the central, or highlands, group 24.1.2020 12.06 (including the Kikuyu, the Kamba, and other subgroups); and the coastal Bantu (Mijikenda). 7Among.1 Ethnic Kenya’s Identity Nilotic speakers, the major groups are the River-Lake, or Western group (Luo); Kenya’sthe Highlands, population or isSouthern divided groupinto more (Kalenjin); than 40 andethnic the groups Plains, (the or mostEastern prominent group presented(Masai). The in TableCushitic 6) -belongingspeaking groups to three include linguistic the families: Oromo andThe Bantu,the Somali. the Cushitic, The Kikuyu, and thewho Nilotic. make Languageup 22% of traditionallythe population, has beenare Kenya’s the primary largest characteristic ethnic group. of ethnicLuhya identity.is the second Bantu largest-speaking (14%), Kenyans followed are dividedby the intoLuo (13%),three different the Kalenjin groups: (12%) the andwestern the groupKamba (Luhya); (11%). theAlthough central, economic or highlands, and political group (includingdevelopment the Kikuyu,have increased the Kamba, mobility and otherand urbanisation subgroups); amongand the the coastal country’s Bantu inhabitants, (Mijikenda). the Amongmajority Kenya’s of Kikuyu Nilotic live speakers, in south the-central major Kenyagroups (Kiambu,are the River Muranga-Lake, orand Western Nyeri groupdistricts), (Luo); the the Highlands, or Southern group (Kalenjin); and the Plains, or Eastern group (Masai). The Cushitic-speaking groups include the Oromo and the Somali. The Kikuyu, who make up 22% of the population, are Kenya’s largest ethnic group. Luhya is the second largest (14%), followed28 by the Luo (13%), the Kalenjin (12%) and the Kamba (11%). Although economic and political development have increased mobility and urbanisation among the country’s inhabitants, the majority of Kikuyu live in south-central Kenya (Kiambu, Muranga and Nyeri districts), the

28 7.1 Ethnic Identity

Kenya’s population is divided into more than 40 ethnic groups (the most prominent presented in Table 6) belonging to three linguistic families: The Bantu, the Cushitic and the Nilotic. Language traditionally has been the primary characteristic of ethnic identity. Bantu-speaking Kenyans are divided into three different groups: the western group (Luhya), the central, or highlands, group (including the Kikuyu, the Kamba, and other subgroups), and the coastal Bantu (Mijikenda). Among Kenya’s Nilotic speakers, the major groups are the River-Lake, or Western group (Luo); the Highlands, or Southern group (Kalenjin); and the Plains, or Eastern group (Masai). The Cushitic-speaking groups include the Oromo and the Somali. The Kikuyu, who make up 22% of the population, are Kenya’s largest ethnic group. Luhya is the second largest (14%), followed by the Luo (13%), the Kalenjin (12%) and the Kamba (11%). Although economic and political development have increased mobility and urbanisation among the country’s inhabitants, the majority of Kikuyu live in south-central Kenya (Kiambu, Muranga and Nyeri districts), the majority of Luhya in western Kenya (Bugoma, Busia and Kakamega districts), the majority of Luo in south-western Kenya (consist of around 40 groups, each associated with an area), the majority of Kamba in east- central Kenya, and the majority of Kalenjin (who include the Nandi, Kipsigis, Eleyo, Marakwet, Pokot and Tugen) are predominantly based in west-central Kenya.40

Table 3. Ethnic composition4142

Ethnic group Population Predominantly based Kikuyu 6 622 576 Central and Nairobi (estimated at 47%) and the Rift Valley (15%) Luhya 5 338 666 Predominantly based in the Western region (80%) and 16% of Nairobi’s population is Luhya Kalenjin 4 967 328 Predominantly based in the Rift Valley (95%) Luo 4 044 440 Predominantly based in Nyanza (estimated at 87%) and Nairobi (15%) Kenyan Somali 2 385 572 Northeastern Province, Wajir area Kisii 2 205 669 Predominantly based in Nyanza (95%) Mijikenda 1 960 574 Coast Province Meru 1 658 108 Eastern Province, Meru District Turkana 988 592 Rift Valley Province, Turkana, Samburu, Trans-Nzoia, Laikipia, Isiolo districts Maasai 841 622 Rift Valley Province, Kajiado and Narok districts Teso 338 833 Western Province, Busia District Embu 324 092 Eastern Province, Embu District Taita 273 519 Coast Province, Taita District Kuria 260 401 Nyanza Province, Kuria District Samburu 237 179 Rift Valley, Baringo District Tharaka 175 905 Eastern Province, East Meru District, Embu District and Kitui District Mbeere 168 155 Eastern Province, Embu District Borana 161 399 Northern Province Basuba 139 271 Western Province, originally from the Lake Victoria islands of Rusinga and Mfangano Swahili 110 614 Coast Province Gabra 89 515 Northern Province Orma 66 275 North Eastern and Coast provinces, Garissa and Tana River districts Rendile 60 437 Eastern Province, Marsabit District, between Lake Turkana and Marsabit Mountain

40 Prinz, J. 2008. Determination of national identity in ambivalence of traditional means of communication and the medium radio in the case study of Kenya. Magister thesis, Wien University, Vienna, Austria. p. 28–32. 41 Oparanya, WA. 2010. ‘2009 Population & Housing Census Results’. Minister of State for Planning, National Development and Vision 2030. p. 34-35. 42 Elischer, S. 2008. Ethnic Coalitions of Convenience and Commitment: Political Parties and Party Systems in Kenya. German Institute of Global and Area Studies, No. 68, February. p. 11.

32 33

FCA_KenyaAnalysis.indd 33 24.1.2020 12.06 The Coast and North Eastern provinces are home to the Mijikenda (they also include the Giriama, Digo, Kauma, Duruma, Jibana, Kambe, Rabai, Ribe) the Pokomo, Taita, Taveta, the Makonde and the Swahili (not referring to the language). Kenya’s small ethnic minority groups, including the Borana, Burji, Ga- bra, Orma, Sakuye and Waata, also live in these two provinces. However, over the years other ethnic groups including the Kamba, Kikuyu and Luo migrated to the coast from other regions.43 The economic consequences of this migration contributed to the ethnic marginalisation of the original inhabitants.

Although criminal justice actors represent an array of tribal groups found in Kenya, the numbers are disproportional as far as the primary Somali clans (with specific reference to the Hawiye, Darod and Dir), Swahili, Kamba, Digo and Giryama found in the coastal and north-eastern regions are concerned. The Luo, Luhya, Kisii and Kamba are over-represented when compared to community regionsmembers are concerned interviewed.. The Luo, Luhya, Kisii and Kamba are over-represented when compared to community members interviewed. However, the ethnic/tribal identity of respondents are less important than their religious identity, with Figure 21reference: Ethnic/tribal to Figure representation 21 where of respondents still regarded it as being very important.

Rahanweyn 1 Hawiye 4 11 13 Dir clan 2 1 Darod 3 18 5 Turkana 1 Tugen 3 Taita 3 1 Swahili 1 8 18 Sabaot 1 Pokot 1 Orma 2 1 3 Mijikenda 1 7 Luo 13 4 1 Luhya 14 4 Kisii 6 1 Kikuyu 14 12 7 Kambe 11 Kamba 15 6 Kalenjin 2 Giryama 2 2 8 Embu 11 Digo 4 29 19 Chonyi 1 2 2 Borana 2 Ameru 5 3 1

0 10 20 30 40 50 60

Criminal Justice Family Detained

Figure 20. Ethnic/tribal representation of respondents However, the ethnic/tribal identity of respondents are less important than their religious identity with reference43 Makoloo, to MO. Figure 2005. 2Kenya:1 where Minorities, respondents Indigenous Peoplesstill regarded and Ethnic Diversity.it as being Minority very Rights important. Group In- ternational. p. 26. Figure 22: Importance of ethnic/tribal identity 34 35

Detained (n=139) 4.35 12.17 15.65 26.09 28.06 15.83

Family (n=131) 12.98 11.45 9.92 19.85 33.59 12.21 FCA_KenyaAnalysis.indd 34 24.1.2020 12.06

Criminal Justice (n=110) 7.27 10.91 11.82 26.36 25.45 18.18

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

regions are concerned. The Luo, Luhya, Kisii and Kamba are over-represented when compared to community members interviewed.

Figure 21: Ethnic/tribal representation of respondents

0 10 20 30 40 50 60

Criminal Justice Family Detained

However, the ethnic/tribal identity of respondents are less important than their religious identity with reference to Figure 21 where respondents still regarded it as being very important.

Figure 22: Importance of ethnic/tribal identity

Figure 23: Identifying other ethnic groups/tribes as 'them' Detained (n=139) 4.35 12.17 15.65 26.09 28.06 15.83

Family (n=131) 12.98 11.45 9.92 19.85 33.59 12.21 Detained (n=137) 6.09 19.13 20 26.96 21.17 7.3

Criminal Justice (n=110) 7.27 10.91 11.82 26.36 25.45 18.18 Family (n=130) 11.54 11.54 21.54 27.69 13.85 13.85 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100% Criminal Justice (n=109) 11.01 10.09 13.76 24.77 24.77 15.6 FigureFigure 23: 21.Identifying Importance other of ethnic ethnic/tribal groups/tribes identity as 'them'

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 30 No 1-20% 30-40% 50-60% 70-80% 90-100% Detained (n=137) 6.09 19.13 20 26.96 21.17 7.3

In an attempt to establish whether ethnic identity has the potential to be regarded as a threat to domesticFamily security(n=130) , respondents11.54 11.54 were asked21.54 to assess historic27.69 grievances between13.85 ethnic/tribal13.85 groups as a threat. Compared to religion, both criminal justice and family respondents regarded ethnicity as a more severe threat to domestic stability than the detained sample. According to oneCriminal focus Justice group (n=109) discussion11.01 10.09in Kilifi, 13.76there is ethnic24.77 discrimination between24.77 the Swahilis15.6 , the Mijikenda and Giriama communities. The county government had been providing more opportunities to the Mijikenda than the Swahilis who are a significant population in Kilifi County. 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% This then justifies why most of the youths who had been linked to terror related crimes were either Muslims or Swahilis. NoLimited1-20% opportunities30-40% 50-60% there contribute70-80% 90-100% to vulnerability.

Figure 22. Identifying other ethnic groups/tribes as ‘them’ Figure 24: Historic grievances based on ethnicity as a threat to domestic security In an attempt to establish whether ethnic identity has the potential to be regarded as a threat to domestic security, respondents were asked to assess historic grievances between ethnic/tribal groups as a threat. Compared to religion, both criminal justice and family respondents regarded ethnicityDetained as a(n=121) more severe17.09 threat 13.68to domestic17.09 stability than11.97 the detained25.64 sample. Ac14.53cording to one focus group discussion in Kilifi, there is ethnic discrimination between the Swahilis, the MijikendaFamilies and (n=106) Giriama19.05 communities.4.76 10.48 The 8.57county government26.67 had been 30.48providing more opportunities to the Mijikenda than the Swahilis who are a significant population in Kilifi County. This then justifies why most of the youths who had been linked to terror related crimes were eitherCriminal Muslims Justice (n=104) or Swahilis.12.67 Limited4.85 5.83 opportunities16.5 there contribute34.95 to vulnerability.25.24

Figure 24: Historic grievances0% based10% on ethnicity20% as a30% threat to 40%domestic 50%security 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

FigureDetained 23. Historic (n=121) grievances17.09 based on ethnicity13.68 as a threat17.09 to domestic11.97 security 25.64 14.53 As an answer to challenges associated with Kenya being a diverse country, 89.43% of criminal justice respondents (n=106), 68.52% of respondents representing family and friends (n=110) andIn an62.29%Families attempt (n=106) ofto establishrespondents19.05 whether part 4.76ethnic of10.48 identitythe detained8.57 has the potential sample26.67 to (n=128) be regarded called as 30.48a threaton the to doKenyan- governmentmestic security, to utiliserespondents more were resources asked toto assess build historic a Kenya grievances beyond between ethnic ethnic/tribal differences. groups This is significantas a threat. as Compared there seems to religion, to be both consensus criminal justicetowards and unifyingfamily respondents non-ethnic regarded-based ethnicity develo aspment strategiesCriminala more Justice severe rather (n=104) threat than 12.67to a domesticdemand4.85 5.83 stability for privileging 16.5than the detainedthe respondent’s sample.34.95 According own ethnic to oneidentity.25.24 focus group discussion in Kilifi, there is ethnic discrimination between the Swahilis, the Mijikenda and Giriama 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 7.1.1communities. Access to The land county government had been providing more opportunities to the Mijikenda than the Swahilis, who are a significant population in Kilifi County. This then justifies why most of the youths No 1-20% 30-40% 50-60% 70-80% 90-100% who had been linked to terror related crimes were either Muslims or Swahilis. Limited opportunities Although access to basic services will be discussed later in this section, families and friends of in the county contributed to vulnerability. Asthose an whoanswer went to throughchallenges the associated criminal justic withe Kenyasystem being and respondentsa diverse country, who experienced 89.43% of criminalit were justice respondents (n=106), 68.52% of respondents representing family and friends (n=110) 34 and 62.29% of respondents part of the detained sample (n=128) called on the Kenyan35 government to utilise more resources to build a Kenya beyond ethnic differences. This31 is significant as there seems to be consensus towards unifying non-ethnic-based development strategies rather than a demand for privileging the respondent’s own ethnic identity.

FCA_KenyaAnalysis.indd7 .1.1 35 Access to land 24.1.2020 12.06

Although access to basic services will be discussed later in this section, families and friends of those who went through the criminal justice system and respondents who experienced it were

31 As an answer to challenges associated with Kenya being a diverse country, 89.43% of criminal justice respondents (n=106), 68.52% of respondents representing family and friends (n=110) and 62.29% of respondents part of the detained sample (n=128) called on the Kenyan government to utilise more resources to build a Kenya beyond ethnic differences. This is significant as there seems to be consensus towards unifying non-ethnic-based development strategies rather than a demand for privileging the respondent’s own ethnic identity.

7.2 Access to land askedAlthough to rate access their to access basic servicesto land will, in bewhich discussed the majority later in this of section,both samples families rated and friends their ofaccess those to landwho between went through 1-30%. the Land criminal ownership justice system and landand respondentsreform has whobecome experienced a very itsensitive were asked topic to in recentrate historytheir access as ownin to land;theg land majority is perceived of both samplesto be linked rated theirto wealth access andto land not between owning 1-30%.land is regardedLand ownership as a form and landof exclusion,reform has becomeespecially a very when sensitive perceived topic in recentto be historylinked as to owning ethnicity. land isAs explainedperceived by to Robert be linked Gilman, to wealth private while ownershipnot owning landenhances is regarded personal as a freedomform of exclusion, (for those especially who are owners),when perceivedbut frequently to be linked leads to toethnicity. vast concentrations As explained by ofRobert wealth Gilman, (even private in the ownership U.S., 75% enhances of the privatelypersonal held freedom land (foris owned those who by are5% owners), of the butprivate frequently landholders), leads to vast and concentrations the effective of denial wealth of freedom(even inand the U.S.,power 75% to of those the privately without held great land wealth.is owned Consequently, by 5% of the private the Mombasa landholders), Republ and theican Councileffective (MRC) denial focuses of freedom on land and powergrievances to those and without the greatfact wealth.that outsiders Consequently, dominate the Mombasa the local economy,Republican which Council consists (MRC) predominantly focuses on land of grievancestourism.44 and the fact that outsiders dominate the local economy, which consists predominantly of tourism.44 As mentioned above, both families and friends as well as detained respondents considered unequalAs mentioned access toabove, land both and families natural and resources friends, as as well the as primary detained threat respondents, to domestic considered security unequal when comparedaccess to to land other and threats. natural It resources is however as theimportant primary threatto note to that domestic especially security families, when compared followed to by criminalother justicethreats. actors It is however and respondents important to representing note that especially the detained families, samplefollowed, recognisedby criminal justice the serious ac- naturetors ofand this respondents threat as representingpresented inthe Figure detained 25 .sample, recognised the serious nature of this threat as presented in Figure 25.

Figure 25: Unequal access to land and natural resources as a security challenge

Detained (n=121) 12.82 10.26 6.84 17.95 26.5 25.64

Families (n=106) 14.71 3.92 2.94 10.78 24.51 43.14

Criminal Justice (n=104) 6.8 6.8 10.86 12.62 33.98 29.13

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

FigureFigure 25 24. confirmed Unequal access the to overall land and seriousness natural resources of asthe a securityland issue challenge in Kenya as presented in Figure 26, but this graph went further and reflected the potential for conflict over land in the mind of both criminal justice respondents and the sample representing family members and friends.

FigureFigure 26: Potential 25 confirmed conflict over the land overall seriousness of the land issue in Kenya as presented in Figure 26, but this graph went further and reflected the potential for conflict over land in the mind of both criminal justice respondents and the sample representing family members and friends.

Detained (n=123) 11.02 10.17 11.86 19.49 22.88 24.58

44 Gatsiounis, I. 2012. ‘After Al-Shabaab’. Current Trends in Islamist Ideology, 14: 74-89. Families (n=111) 10.19 7.41 2.78 10.19 27.78 41.67

36 37

Criminal Justice (n=105) 3.85 5.77 7.69 9.62 43.27 29.81

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

FCA_KenyaAnalysis.indd 36 No 1-20% 30-40% 50-60% 70-80% 90-100% 24.1.2020 12.06

asked to rate their access to land, in which the majority of both samples rated their access to land between 1-30%. Land ownership and land reform has become a very sensitive topic in recent history as owning land is perceived to be linked to wealth and not owning land is regarded as a form of exclusion, especially when perceived to be linked to ethnicity. As explained by Robert Gilman, private ownership enhances personal freedom (for those who are owners), but frequently leads to vast concentrations of wealth (even in the U.S., 75% of the privately held land is owned by 5% of the private landholders), and the effective denial of freedom and power to those without great wealth. Consequently, the Mombasa Republican Council (MRC) focuses on land grievances and the fact that outsiders dominate the local economy, which consists predominantly of tourism.44

As mentioned above, both families and friends as well as detained respondents considered unequal access to land and natural resources as the primary threat to domestic security when compared to other threats. It is however important to note that especially families, followed by criminal justice actors and respondents representing the detained sample, recognised the serious nature of this threat as presented in Figure 25.

Figure 25: Unequal access to land and natural resources as a security challenge

Detained (n=121) 12.82 10.26 6.84 17.95 26.5 25.64

Families (n=106) 14.71 3.92 2.94 10.78 24.51 43.14

Criminal Justice (n=104) 6.8 6.8 10.86 12.62 33.98 29.13

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 25 confirmed the overall seriousness of the land issue in Kenya as presented in Figure 26, but this graph went further and reflected the potential for conflict over land in the mind of both criminal justice respondents and the sample representing family members and friends.

Figure 26: Potential conflict over land

Detained (n=123) 11.02 10.17 11.86 19.49 22.88 24.58

Families (n=111) 10.19 7.41 2.78 10.19 27.78 41.67

Criminal Justice (n=105) 3.85 5.77 7.69 9.62 43.27 29.81

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Unequal access to land canNo be seen1-20% as 30-40%a major 50-60%challenge70-80% to domestic90-100% security, compared to the need for land rating between 90-100% (see Figure 26). The need for land reform as growing concern urgently has to be satisfied as government’s success to provide access to land is rated betweenFigure 25. 1Potential-30% conflict(see Figure over land 27).

32 Figure 27: Need for land

90 80 76.64 70 60 54.24 50 40 30 22.03 20 22.03 15.89 10 5.93 6.78 10.17 0 0.85 0.93 2.8 3.74 No 1-20% 30-40% 50-60% 70-80% 90-100%

Family (n=108) Detained (n=120)

Figure 26. Need for land Figure 28: Government's success in providing access to land

45 41.46 40 35 36.71 30 28.05 25 20 18.99 13.92 15 15.85 16.46 13.92 13.92 10 8.54 5 6.1 0 0 1-30 40-60 70-80 90-100

Family (n=82) Detained (n=89) Figure 27. Government’s success in providing access to land

After having established the religious and ethnic/tribal affiliations of respondents and determining its importance to the identity of respondents, the following section will discuss the importance of religious and ethnic diversity in the police as an important factor towards police legitimacy.

36 In dealing with religious grievances, respondents were asked whether religious dialogue will37 be a viable strategy to deal with security challenges (Figure 28). The family and friends sample was particularly supportive of religious dialogue, followed by criminal justice and detained respondents. However, when discussing the best counter strategy, the question should be what the role of religious differences in security challenges is: Is it a motivating factor in contributing FCA_KenyaAnalysis.indd to 37 radicalisation into violent extremist organisations and is it a topic that needs to be discussed? 24.1.2020 12.06 According to criminal justice respondents (n=103), religious ideology is the second most prominent reason (65.34%) why individuals join extremist organisations after financial incentives (80.8%). Family members and friends (n=105) on the other hand placed religious ideology at

33 Unequal access to land can be seen as a major challenge to domestic security, compared to the need for land rating between 90-100% (see Figure 26). The need for land reform as growing concern urgently has to be satisfied as government’s success to provide access to land is rated between 1-30% (see Figure 27).

After having established the religious and ethnic/tribal affiliations of respondents and determining its importance to the identity of respondents, the following section will discuss the importance of religious and ethnic diversity of the police as an important factor towards police legitimacy.

In dealing with religious grievances, respondents were asked whether religious dialogue will be a viable strategy to deal with security challenges (Figure 28). The family and friends sample was particularly supportive of religious dialogue, followed by criminal justice and detained respondents. However, when discussing the best counter strategy, the question should be what the role of religious differences in security challenges is: Is it a motivating factor in contributing to radicalisation into violent extremist organisations and is it a topic that needs to be discussed? According to criminal justice respondents (n=103), religious ideology is the second most prominent reason (65.34%) why individuals join extremist organisations after financial incentives (80.8%). Family members and friends (n=105) on the other hand placed religious ideology at fifth place (60.78%) on of a list of twenty potential reasons after anger towards the police (75.97%), employment (71.29%), financial incentives (68.32%) and anger towards government (62.5%). In contrast, respondents part of the detained sample (n=117) placed religious ideology at number six (52.3%), after anger towards the police (66.96%), anger towards government (60.18%), financial incentives (59.46%) and treatment from security forces since arrest (56.88%).

As an answer to challenges associated with Kenya being a diverse country, 83.81% of criminal justice fifthrespondents place (60.78%) (n=106), 68.22% on of aof listrespondents of twenty representing potential familyreasons and after friends anger (n=110) towards and 56.55% the policeof (75.97%),respondents employment part of the (71.29%),detained sample financial (n=128) incentives called (68.32%)on the Kenyan and government anger towards to utilise government more (62.5%).resources In to contrast,build a Kenya respondents beyond religious part divides.of the detained sample (n=117) placed religious ideology at number six (52.3%), after anger towards the police (66.96%), anger towards government (60.18%), financial incentives (59.46%) and treatment from security forces since arrest (56.88%).

Figure 29: Religious dialogue as a strategy to deal with security challenges

1.71 Detained (n=122) 6.84 10.26 23.08 31.62 26.5

0.93 Families (n=107) 6.54 19.63 36.45 34.58 1.87

1.94 Criminal Justice (n=104) 4.85 6.8 17.48 40.78 28.16

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

AsFigure an answer 28. Religious to challenges dialogue as a associated strategy to deal with with Kenya security beingchallenges a diverse country, 83.81% of criminal justice respondents (n=106), 68.22% of respondents representing family and friends (n=110) and 56.55% of respondents part of the detained sample (n=128) called on the Kenyan government to utilise more resources to build a Kenya beyond religious differences.

7.3 Create a sense of belonging in Kenya Most38 challenging for any religious and ethnic diverse country is to establish and maintain a sense 39 of belonging, especially amongst minorities. As established above, both criminal justice actors and families and friends of those who went through the criminal justice system and respondents who experienced it, include both Islam and Christianity and represent a multitude of tribal groups. However, despite an overlap there exists a clear distinction between the three samples.

FCA_KenyaAnalysis.inddWith 38 this in mind, criminal justice respondents indicated a higher sense of belonging and being 24.1.2020 12.06 proud of being Kenyan.

Figure 30: Government's success in creating a sense of belonging

1.19 Detained (n=89) 4.76 27.38 35.71 30.95

Family (n=82) 16 20 34.67 18.67 10.67

1.28 Criminal Justice (n=79) 12.82 39.74 33.33 12.82

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

0 1-30 40-60 70-80 90-100

fifth place (60.78%) on of a list of twenty potential reasons after anger towards the police (75.97%), employment (71.29%), financial incentives (68.32%) and anger towards government (62.5%). In contrast, respondents part of the detained sample (n=117) placed religious ideology at number six (52.3%), after anger towards the police (66.96%), anger towards government (60.18%), financial incentives (59.46%) and treatment from security forces since arrest (56.88%).

Figure 29: Religious dialogue as a strategy to deal with security challenges

1.71 Detained (n=122) 6.84 10.26 23.08 31.62 26.5

0.93 Families (n=107) 6.54 19.63 36.45 34.58 1.87

1.94 Criminal Justice (n=104) 4.85 6.8 17.48 40.78 28.16

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

As an answer to challenges associated with Kenya being a diverse country, 83.81% of criminal justice respondents (n=106), 68.22% of respondents representing family and friends (n=110) and 56.55% of respondents part of the detained sample (n=128) called on the Kenyan government to utilise more resources to build a Kenya beyond religious differences.

7.3 Create a sense of belonging in Kenya Most challenging for any religious and ethnic diverse country is to establish and maintain a sense of7.3 belonging, Create especially a sense ofamongst belonging minorities. in Kenya As established above, both criminal justice actors and families and friends of those who went through the criminal justice system and respondents whoThe experienced most challenging it, aspectinclude for both any religiousIslam and and Chris ethnictian diverseity and country represent is to establish a multitude and maintain of tribal a groups.sense ofHowever, belonging, despite especially an overlapamongst thereminorities. exists As a established clear distinction above, between both criminal the threejustice samples. actors Withand thisfamilies in mind, and friendscriminal of justicethose whorespondents went through indicated the criminal a higher justice sense system of andbelonging respondents and whobeing proudexperienced of being it, Kenyaincluden. both Islam and Christianity and represent a multitude of tribal groups. Despite an overlap, there exists a clear distinction between the three samples. With this in mind, criminal justice Figurerespondents 30: Government's indicated success a inhigher creating sense a sense of of belonging belonging and being proud of being Kenyan.

1.19 Detained (n=89) 4.76 27.38 35.71 30.95

Family (n=82) 16 20 34.67 18.67 10.67

1.28 Criminal Justice (n=79) 12.82 39.74 33.33 12.82

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

0 1-30 40-60 70-80 90-100

Figure 29. Government’s success in creating a sense of belonging

34 After establishing the influence of diversity in Kenyan society, respondents also emphasised the need to build an inclusive society to be able to successfully address security challenges Kenya currently is being confronted with.

All three samples – especially criminal justice respondents – recognised the importance of initiatives to build national cohesion to deal with security challenges (see Figure 31).

In July 2018, French president Emmanuel Macron announced that his government introduced a national service requirement for all 16-year-olds, to be presented in two phases. The first phase focussing on civil culture is mandatory; participants will go through a month-long placement when young French citizens can look into teaching or work with charities, or take part in traditional military training with the police, fire service or defence force. The second phase is optional and can be between three months to a year. This gives young people the opportunity to work in defence and security or alternatively in social care, the environment, or heritage. The decision was taken to ‘promote social cohesion and foster and sustain a more active sense of citizenship’.45 Although national service can be expensive, its benefits in establishing a sense of belonging and duty to the country and its people beyond differences can on the medium- to long-term outweigh the costs. Family and criminal justice respondents also supported this proposal (see Figure 32).

Building on national cohesion, the majority of all three samples, especially criminal justice representatives, supported community involvement as a counter to security challenges. This support enforces the importance of community policing to be discussed in a later section. Although community involvement is critical to address any challenge affecting it, without trust in government and its institutions, the community will be reluctant to get involved in any initiatives either in word or deed. When discussing radicalisation and initiatives to prevent and counter recruitment into violent extremist organisations this matter will again be raised.

45 Aamna Mohdin. European countries are quietly bringing back the draft. Quartz, 3 July 2018. Available at https://qz.com/1318379/france-joins-sweden-and-lithuania-in-bringing-back-mandatory-national- service/ (accessed on 10 January 2019).

38 39

FCA_KenyaAnalysis.indd 39 24.1.2020 12.06

After establishing the influence of diversity in Kenyan society, respondents also emphasised the need to build an inclusive society to be able to successfully address security challenges the country currently is being confronted with.

All three samples – especially criminal justice respondents – recognised the importance of initiatives to build national cohesion to deal with security challenges (see Figure 31).

Figure 31: Initiatives to build national cohesion to deal with security challenges

1.69

Detained (n=122) 9.32 12.71 18.64 27.97 29.66 After establishing the influence of diversity in Kenyan society, respondents also emphasised the need to build an inclusive society to be able to successfully address security challenges the

countryFamilies currently (n=107) is being4.72 4.72 confronted13.21 with.12.26 28.3 36.79

All three samples0.96 – especially1.92 criminal justice respondents – recognised the importance of initiativesCriminal Justice to build(n=104) national2.88 13.46cohesion to deal with38.46 security challenges (see Figure42.31 31).

Figure 31: Initiatives to build national cohesion to deal with security challenges 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100% 1.69 Detained (n=122) 9.32 12.71 18.64 27.97 29.66 In July 2018, French president Emmanuel Macron announced that his government introduced a national service requirement for all 16-year-olds to be presented in two phases. The first phase focussingFamilies on (n=107) civil 4.72culture4.72 is13.21 mandatory12.26 in which participants28.3 will go through36.79 a month-long

placement when0.96 young1.92 French citizens can look into teaching or work with charities, or take part in traditional military training with the police, fire service or defence force. The second phase is optionalCriminal Justice and (n=104) can be2.88 between13.46 three months 38.46to a year giving young people42.31 the opportunity to work in defence and security or alternatively in social care, the environment, or heritage. The decision was taken0% to ‘promote10% 20% social 30%cohesion40% and 50%foster and60% sustain70% a more80% active90% sense100% of citizenship’.45 Although nationalNo 1-20% service30-40% can be 50-60%expensive,70-80% its benefits90-100% in establishing a sense of belonging and duty to the country and its people beyond differences can on the medium- to

longFigure-term 30. Initiativesoutweigh to build the national costs. cohesion Family to dealand with criminal security challengesjustice respondents also supported this Inproposal July 2018, (see French Figure president 32). Emmanuel Macron announced that his government introduced a national service requirement for all 16-year-olds to be presented in two phases. The first phase focussingFigure 32: Nationalon civil service culture of the youthis mandatory in which participants will go through a month-long placement when young French citizens can look into teaching or work with charities, or take part in traditional military training with the police, fire service or defence force. The second phase is optional and can be between three months to a year giving young people the opportunity to work inDetained defence (n=122) and4.27 security13.68 or alternatively11.11 21.37in social care, the26.5 environment, or 23.08heritage. The decision was taken to ‘promote social cohesion and foster and sustain a more active sense of citizenship’.45 Although national service can be expensive, its benefits in establishing a sense of belongingFamilies and (n=107) duty to9.43 the 7.55country7.55 and its 22.64people beyond 22.64differences can on30.19 the medium- to

long-term outweigh the0.68 costs. Family and criminal justice respondents also supported this Buildingproposal on (see national Figure 32cohesion,). the majority of all three samples, especially criminal justice Criminal Justice (n=104) representatives supported4.85 8.74 community16.5 involvement as 33.01a counter to security challenges.26.21 This supportFigure 32: Nationalenforces service the of importancthe youth e of community policing to be discussed in a later section. Although community0% involvement10% 20%is critical30% to address40% any50% challenge60% affecting70% 80% it, without90% trust100% in government and its institutionsNo the1-20% community30-40% will50-60% be reluctant70-80% to 90-100%get involved in any initiatives either Detainedin word (n=122) or deed.4.27 When13.68 discussing11.11 radicalisation21.37 and initiatives26.5 to prevent23.08 and counter

recruitmentFigure 31. National into violent service ofextremist the youth organisations this matter will again be raised.

Figure 33: Community involvement to address security challenges Families (n=107) 9.43 7.55 7.55 22.64 22.64 30.19

0.68 35 0.85 Criminal Justice (n=104) 4.85 8.74 16.5 33.01 26.21 Detained (n=122) 5.084.24 17.8 32.2 39.83

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2.8 No 1-20% 30-40% 50-60% 70-80% 90-100% Families (n=107) 3.74 16.82 26.17 50.47

Criminal Justice (n=104) 9.71 38.83 50.49 0.97 35

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 32. Community involvement to address security challenges 8. Trust in Government

Measuring40 the performance of government overall before requesting respondents to rate the 41 delivery of specific services it is responsible for, the questionnaire asked the three samples to rate their trust in the president, national government, local government and politicians overall. From the onset it is important to note that criminal justice respondents expressed greater trust in these institutions than the other two samples. The obvious explanation is that criminal justice respondents as civil servants should have a greater level of trust than members of the public. FCA_KenyaAnalysis.inddAnother 40 explanation may also relate to the fact that criminal justice respondents were not on 24.1.2020 12.06 the receiving end of the criminal justice system as an extension of government and that will make them less sceptical of government and what it represents.

Although trust in government and its institutions will be discussed throughout this report, Table 4 will summarise the position of these institutions as expressed by respondents. This rating is based on 70-100% trust.

Table 4: Trust in government and its institutions

Position Detained (n=91) Family (n=84) Criminal Justice (n=80) 1 Presidency Presidency President (45.55%) (32.1%) (81.25%) 2 Judiciary Judiciary National government (27.38%) (23.07%) (70%) 3 National government Military Military (22.35%) (20.51%) (62.5%) 4 Military National government Intelligence agencies (62.5%)

36 8. Trust in Government

Measuring the performance of government overall before requesting respondents to rate the delivery of specific services it is responsible for, the questionnaire asked the three samples to rate their trust in the president, national government, local government and politicians overall. It is important to note that criminal justice respondents expressed greater trust in these institutions than the other two samples. The obvious explanation is that criminal justice respondents as civil servants should have a greater level of trust than members of the public. Another explanation may also relate to the fact that criminal justice respondents were not on the receiving end of the criminal justice system as an extension of government and that will make them less sceptical of government and what it represents.

Table 4. Trust in government and its institutions

Position Detained (n=91) Family (n=84) Criminal Justice (n=80) 1 Presidency (45.55%) Presidency (32.1%) President (81.25%) 2 Judiciary (27.38%) Judiciary (23.07%) National government (70%) 3 National government (22.35%) Military (20.51%) Military (62.5%) 4 Military (20.73%) National government (19.75%) Intelligence agencies (62.5%) 5 Local government (16.67%) Intelligence agencies (14.67%) Religious leaders (62.5%) 6 Prison authorities (15.39%) Prison authorities (12.99%) Prison authorities (61.25%) 7 Intelligence agencies (13.09%) Politicians overall (11.39%) Police (55.7%) 8 Politicians overall (10.59%) Local government (10.25%) Local government (46.84%) 9 Police (5.88%) Police (7.6%) Politicians overall (11.25%)

Starting with the president, respondents – when compared to the national and local government and politicians overall – expressed most trust in President Uhuru Kenyatta. President Kenyatta received the highest approval from criminal justice respondents with 46.25% rating their trust in him between 70-80%, followed by 35% that rated their trust in him between 90-100%. Only 1.25% rated their trust in him between 1-10% and another 1.25% between 10-20%. Families and friends of those who went through the criminal justice system expressed the least trust in President Kenyatta with 14.81% rating trust between 1-10% and another 22.22% between 10-20%. Despite receiving very low approval from one spectrum, 18.52% rated their trust in him between 70-80% followed by 13.58% that rated their trust in him between 90-100%. The majority, 30.86% rated their trust between 30-60%. Respondents who went through the criminal justice system expressed far more trust with only 5.56% rating trust between 1-10% and 17.78% between 10-20%. The majority in this sample, 33.33% rated their trust between 70-80%, followed by 31.11% who rated trust between 30-60%. 12.22% rated their trust in President Kenyatta between 90-100%.

Criminal justice respondents expressed very high levels of trust – yet lower than that in the president – in the national government with 42.5% rating trust between 70-80% and a further 27.5% between 90-100%. The category of between 30 to 60% however increased to 26.25% when compared to 16.25% in reference to the president. Following the same trend as with the president, the family and

40 41

FCA_KenyaAnalysis.indd 41 24.1.2020 12.06

(20.73%) (19.75%) 5 Local government Intelligence agencies Religious leaders (16.67%) (14.67%) (62.5%) 6 Prison authorities (15.39%) Prison authorities Prison authorities (12.99%) (61.25%) 7 Intelligence agencies Politicians overall Police (13.09%) (11.39%) (55.7%) 8 Politicians overall Local government Local government (10.59%) (10.25%) (46.84%) 9 Police Police Politicians overall (5.88%) (7.6%) (11.25%)

Starting with the president, respondents – when compared to the national and local government and politicians overall – expressed most trust in President Uhuru Kenyatta. President Kenyatta received the highest approval from criminal justice respondents with 46.25% rating their trust in him between 70-80%, followed by 35% that rated their trust in him between 90-100%. Only 1.25% rated their trust in him between 1-10% and another 1.25% between 10-20%. Families and friends of those who went through the criminal justice system expressed the least trust in President Kenyatta with 14.81% rating trust between 1-10% and another 22.22% between 10-20%. Despite receiving very low approval from one spectrum, 18.52% rated their trust in him between 70-80% followed by 13.58% that rated their trust in him between 90-100%. The majority, 30.86% rated their trust between 30-60%. Respondents who went through the criminal justice system expressed far more trust with only 5.56% rating trust between 1-10% and 17.78% between 10-20%. The majority in this sample, 33.33% rated their trust between 70- 80%, followed by 31.11% who rated trust between 30-60%. 12.22% rated their trust in President Kenyatta between 90-100%.

Figure 34: Trust in the presidency

Detained (n=91) 5.56 17.78 31.11 33.33 12.22

Family (n=84) 14.81 22.22 30.86 18.52 13.58

Detained (n=91) 5.88 27.06 44.71 16.47 5.88 1.25 Criminal Justice (n=80) 1.25 16.25 46.25 35 Family (n=84) 13.58 30.86 35.8 11.11 8.64

1.250% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Criminal Justice (n=80) 26.250-10 10-20 30-60 70-8042.5 90-100 27.5

Figure 33. Trust in the presidency2.5 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Criminal justice respondents expressed very high levels of trust – yet lower than that in the president – in the national government0-10 10-20with 42.5%30-60 rating70-80 trust90-100 between 70-80% and a further

27.5%Detained between (n=91) 905.88-100%. The27.06 category of between 30 44.71to 60% however increased16.47 to 26.255.88 % Localwhen comparedgovernment to lost 16.25% substantial in reference trust amongst to the president. all as the Followingmajority ofthe all same three trend samples as with rat theed president, the family and friends sample expressed far less trust in national government than trust betweenFamily (n=84) 30-60%13.58 (45.24% amongst30.86 the detained, 43.03%35.8 representing11.11 criminal8.64 justice andthe detained40.51% sample.within the Notwithstanding, family and friend extremes sample). trust in32.14% the national of the governmentdetained sample decreased and 29.11%substantially amongst in the1.25 family other andtwo friendssamples rated when trust compared in local governmentto that in the between president 10. -The20%. majority Criminal in justicebothCriminal samples, representatives Justice (n=80) 44.71% expressedin the26.25 detained more and trust 35.8% as 34.18% in the42.5 familyrated trustand friendsbetween sample 7027.5-80% rated and trust a between 30-60%. further 12.66% between2.5 90-100%. Amongst family and friends 8.64% rated trust in national government between0% 9010%-100%,20% in comparison30% 40% 6.33%50% expressed60% the70% same level80% of 90%trust in 100%local Figure 35: Trust in national government government. 0-10 10-20 30-60 70-80 90-100

Figure Figure 36: 34.Trust Trust in local in national government government Local government lost substantial trust amongst all as the majority of all three samples rat37ed trust between 30-60% (45.24% amongst the detained, 43.03% representing criminal justice2.38 and 40.51%Detained (n=91) within5.95 the family 32.14and friends sample). 32.14%45.24 of the detained 14.29sample and 29.11% amongst family and friends rated trust in local government between 10-20%. Criminal justice representatives expressed more trust as 34.18% rated trust between 70-80% and a further Family12.66% (n=84) between10.13 90-100%.29.11 Amongst family and friends40.51 8.64% rated trust13.92 in national6.33 government between2.53 90-100%, in comparison 6.33% expressed the same level of trust in local government.Criminal Justice (n=80) 7.59 43.03 34.18 12.66

Figure 36: Trust in local0% government10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

0-10 10-20 30-60 70-80 90-100 2.38 Figure 35. Trust in local government Detained (n=91) 5.95 32.14 45.24 14.29 Respondents within all three samples expressed the least trust in politicians overall as 48.1% of 2.35 family members and friends rated their trust between 0-10%, while 28.24% within the detained Family (n=84) 10.13 29.11 40.51 13.92 6.33 and 10%Primary amongst (n=91) the criminal28.24 justice sample expressed31.76 a similar view.29.41 The majority8.24 within the criminal justice (41.25%)2.53 and detained sample (31.76%) rated their trust between 10-20%. Overall,Criminal Justice representatives (n=80) 7.59 within all three 43.03samples expressed very similar34.18 views in relation12.66 to trust betweenFamily 70 -(n=84)100% (see Figure 3748.1). 24.05 16.46 6.33 5.06 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Figure 37: Trust in politicians 0-10 10-20 30-60 70-80 90-100 Criminal Justice (n=80) 10 41.25 37.5 6.25 5

Respondents within all three samples expressed the least trust in politicians overall as 48.1% of family members and0% friends10% rated20% their30% trust between40% 50% 0-10%60%, while70% 28.24%80% within90% the 100%detained and 10% amongst the criminal 0-10justice10-20 sample30-60 expressed70-80 a90-100 similar view. The majority within the criminalFigure 36. justice Trust in (41.25%) politicians and detained sample (31.76%) rated their trust between 10-20%. Overall, representatives within all three samples expressed very similar views in relation to trust between8.1 Participating 70-100% in (see the Figure political 37 ).process 42 43 Participating in the political process is most probably the best way to determine the legitimacy Figureof any 37 : government.Trust in politicians In short, the public will not participate in elections if they do not see the value of voting. Although the majority of all three samples identified the value of participating in the political process, criminal justice respondents were particularly positive. 38 FCA_KenyaAnalysis.inddFigure 42 38: Participation in the political process 24.1.2020 12.06

Criminal Justice (n=100) 75 25

Detained (n=122) 52.46 47.54

38 Family (n=108) 55.56 44.44

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Yes No

Respondents did not see the value of participating in the political process since the perception existed that the political process was not free and fair, followed by the sentiment that politicians only represented a small minority. To a lesser extent amongst the three samples the perception that politicians were corrupt was the most prominent reason not to participate.

Figure 39: Reasons for not participating in the political process

39 friends sample expressed far less trust in national government than the detained sample. Notwith- standing, extreme trust in the national government decreased substantially in the other two samples when compared to that in the president. The majority in both samples, 44.71% in the detained and 35.8% in the family and friends sample rated trust between 30-60%.

Local government lost substantial trust amongst all as the majority of all three samples rated trust between 30-60% (45.24% amongst the detained, 43.03% representing criminal justice and 40.51% within the family and friends sample). 32.14% of the detained sample and 29.11% amongst family and friends rated trust in local government between 10-20%. Criminal justice representatives expressed more trust as 34.18% rated trust between 70-80% and a further 12.66% between 90-100%. Amongst family and friends, 8.64% rated trust in national government between 90-100%. In comparison 6.33% expressed the same level of trust in local government. 2.35 RespondentsPrimary (n=91)within all three28.24 samples expressed the31.76 least trust in politicians29.41 overall as 48.1%8.24 of family members and friends rated their trust between 0-10%, while 28.24% within the detained and 10% amongst the criminal justice sample expressed a similar view. The majority within the criminal justice (41.25%)Family and (n=84) detained sample (31.76%)48.1 rated their trust between 10-20%.24.05 Overall, representatives16.46 6.33 5.06 within all three samples expressed very similar views in relation to trust between 70-100% (see Figure 37).

8.1Criminal Participating Justice (n=80) in10 the political41.25 process 37.5 6.25 5 Participating in the political process is probably the best way to determine the legitimacy of any government. In short,0% the public10% will20% not participate30% 40% in elections50% if60% they do70% not see80% the value90% of voting.100% Although the majority of all three samples identified the value of participating in the political process, 0-10 10-20 30-60 70-80 90-100 criminal justice respondents were particularly positive.

Respondents did not see the value of participating in the political process since the perception 8existed.1 Participating that the political in the process political was process not free and fair, followed by the sentiment that politicians Participatingonly represented in thea small political minority. process To a lesseris most extent probably amongst the the best three way samples to determine the perception the legitimacy that ofpoliticians any government. were corrupt In wasshort, the the most public prominent will notreason participate not to participate. in elections if they do not see the value of voting. Although the majority of all three samples identified the value of participating in the political process, criminal justice respondents were particularly positive.

Figure 38: Participation in the political process

Criminal Justice (n=100) 75 25

Detained (n=122) 52.46 47.54

Family (n=108) 55.56 44.44

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

Yes No

Figure 37. Participation in the political process Respondents did not see the value of participating in the political process since the perception existed that the political process was not free and fair, followed by the sentiment that politicians only represented a small minority. To a lesser extent amongst the three samples the perception that politicians were corrupt was the most prominent reason not to participate.

Figure 39: Reasons for not participating in the political process

42 43

FCA_KenyaAnalysis.indd 43 24.1.2020 12.06

Do not feel 'Kenyan' 8.16 1.79 Not allowed to register political party 3.57 4.08 Political parties not representing ideas 14.29 7.14

Did not recognise the political process 8.93

All politicians are corrupt 42.86 32.14 14.29

Political process is not free and fair 51.02 42.86 57.14

Politicians represent small elite 51.02 37.5 46.43

0 20 40 60 80 100 120 140 160

Family (n=49) Detained (n=56) Criminal Justice (n=28)

Figure8.2 38. Access Reasons to for national not participating identity in the documents political process Critical in creating a sense of belonging, but also a source of marginalisation has been the perceived unequal access to national identity documents (ID). An ID document or card is technically the key to the privileges of being a citizen of that particular country, from being 8.2allowed Access to participateto national in identitythe political documents process (vote in elections), access a government building, to open a bank account. Not being in possession of an ID has a tendency to facilitate arrest in CriticalKenya in. creatingFor example, a sense following of belonging, two but attacks also a sourcein March of marginalisation, 2014 in Nairobi has andbeen Mombasa the perceived that killed unequal12 people access, Joseph to national Ole identityLenku, thedocuments Kenyan (ID). Minister An ID ofdocument Interior or announced card is technically that more the than key 4 000 to peoplethe privileges suspected of being of beinga citizen Somali of that h adparticular been arrested,country, from some being held allowed at Kasarani to participate soccer in stadium . theThe political mass process arrests (vote were in elections), part of access a crackd a governmentown on building,Somalis to suspected open a bank of account. belonging Not to or beingsympathizing in possession with of an al ID-Shabaab. has a tendency In another to facilitate example, arrest ina Kenya. participant For example, part followingof a focus two group attacksdiscussion in March gave 2014 the in following Nairobi and testimony: Mombasa E thatight killedmonths 12 people,ago, I wasJoseph arrested Ole Lenku, in Wajir the Kenyan town without Ministeran ID ofand Interior for alleged announced involvem that moreent with than al 4- Shabaab000 people and suspected I was placed of being under Somali remand had been for sixar- months, rested,interrogated, some held tortured at Kasarani and soccer injected stadium. with chemicalsThe mass arrests that changed were part my of lifea crackdown forever. Ion am Somalis now trapped suspectedin the body of belonging of a disable to ord sympathizingman with a harrowing with al-Shabaab. tale of Inextremism another example, in the hands a participant of the same part security of agentsa focus thatgroup were discussion supposed gave to the protect following me. testimony: My alleged Eight involvement months ago, claim I was witharrested al-Shabaab in Wajir didn’t townbear without any fruit an ID since and fornone alleged of their involvement advanced with interrogation al-Shabaab andskills I was could placed work. under I was remand innocent for and to sixdate months, I didn’t interrogated, get justice tortured or compensation and injected fromwith chemicals anyone. Ithat was changed even threatened my life forever. by the I am anti now-terrorism trappedunit in the body of a disabled man with a harrowing tale of extremism in the hands of the same securitynot to agents pursue that the werematter supposed. to protect me. My alleged involvement claim with al-Shabaab didn’t bear any fruit since none of their advanced interrogation skills could work. I was innocent and to Stopdate Iand didn’t asking get justice people or compensation to produce froman IDanyone. can reasonably I was even threatened be applied by the, but anti-terrorism then it should be unitnotapplied to pursue equally the tomatter. all while recognising that the person has 24-hours to produce the document. According to the Kenyan Constitution, the freedom of movement (Article 39) can only be limited Stopingafter andall possibilities asking people to tolimit produce any basican ID rightcan reasonably were taken be intoapplied, consideration but it should (Article be applied 24). Article equally21 set to alltwo while principles recognising in the that implementation the individual in question of rights has and 24-hours fundamental to produce freedoms, the document. namely: All AccordingState organs to the Kenyanand all Constitution, public officers the freedom have the of movementduty to address (Article 39)the can needs only beof limitedvulnerable after groups all possibilities to limit any basic right were taken into consideration (Article 24). Article 21 set two within society, including women, older members of society, persons with disabilities, children, principles in the implementation of rights and fundamental freedoms, namely: All State organs and youth, members of minority or marginalised communities, and members of particular ethnic, all public officers have the duty to address the needs of vulnerable groups within society, including women,religious older or members cultural ofcommunities society, persons; and withsecondly, disabilities, the State children, shall youth, enact members and implement of minority legislat or ion marginalisedto fulfil its communities,international and obligations members in of respect particular of ethnic,human religiousrights and or culturalfundamental communities; freedoms and. secondly, the State shall enact and implement legislation to fulfil its international obligations in re- Figure 40: Stopped and asked for ID spect of human rights and fundamental freedoms.

44 45 40

FCA_KenyaAnalysis.indd 44 24.1.2020 12.07

70 62 60

50 46.67

30 17.78 20 13.33 13 13.33 11 8.89 9 10 5

0 1 2 3 4 5

Family (n=90) Detained (n=94) Figure 39. Stopped and asked for ID

The Kenyan Constitution46, under Article 39 guarantees the freedom of movement of every person. Article 21 under the implementation of rights and fundamental freedoms: (1) It is a fundamental duty of the State and every State organ to observe, respect, protect, The Kenyanpromote Constitution and fulfil46 the, under rights Article and fundamental 39 guarantees freedoms the freedom in the ofBill movement of Rights. of every person.(2) The Article State 21 shall under take the legislative, implementation policy ofand rights other and measures, fundamental including freedoms: the setting of standards, to achieve the progressive realisation of the rights guaranteed under Article 43. (3)1. ItAll is aState fundamental organs andduty allof thepublic State officers and every have State the organduty to observe,address respect,the needs protect, of vulnerable promotegroups within and fulfil society, the rights including and fundamentalwomen, older freedoms members in ofthe society, Bill of Rights. persons with disabilities, children, youths, members of minority or marginalised communities, and members of particular 2. Theethnic, State reli shallgious take or legislative,cultural communities. policy and other measures, including the setting of stand- (4) ards,The State to achieve shall enactthe progressive and implement realisation legislation of the torights fulfil guaranteed its international under obligationsArticle 43. in respect of human rights and fundamental freedoms. The Constitution also stipulated the following limitation of rights and fundamental freedoms under 3. All State organs and all public officers have the duty to address the needs of vulner- Article 24: able groups within society, including women, older members of society, persons with (1) A right or fundamental freedom in the Bill of Rights shall not be limited except by law, and disabilities,then only tochildren, the extent youths, that members the limitation of minority is reasonable or marginalised and justifiable communities, in an and open and membersdemocratic of particularsociety based ethnic, on religious human dignity,or cultural eq communities.uality and freedom, taking into account all relevant factors, including— 4. The State(a) the shall nature enact of and the implementright or fundamental legislation freedom;to fulfil its international obligations in respect(b) ofthe human importance rights andof the fundamental purpose of freedoms. the limitation; (c) the nature and extent of the limitation; The Constitution(d) the also need stipulated to ensure the that following the enjoyment limitation of of rightsrights and fundamentalfundamental freedoms freedoms by any under Article 24:individual does not prejudice the rights and fundamental freedoms of others; and (e) the relation between the limitation and its purpose and whether there are less 1. A right orrestrictive fundamental means freedom to achieve in the the Bill purposeof Rights shall not be limited except by law, and then only to the extent that the limitation is reasonable and justifiable in an open and democratic society based on human dignity, equality and freedom, taking into Sensitivityaccount around all producing relevant factors, an ID including—can be traced back to the colonial period when the pass was produced only on demand by a police officer or an authorised person in accordance with the law. This was a.seen the as nature a measure of the right to prevent or fundamental the free freedom; movement of non-Europeans in the country of native Kenyans.b. the Years importance after ofindependence the purpose of militias the limitation; from certain communities were demanding a display of IDc. cards the nature during and the extent 1992 of tribalthe limitation; clashes to assist in identifying their victims. A similar formula was appliedd. the need in the to ensure1994 Rwandathat the enjoymentgenocide. of47 rights and fundamental freedoms by any individual does not prejudice the rights and fundamental freedoms of The second sensitiveothers; topic and associated with ID documents is the perceived unequal access to these documents. Memberse. the relation of the between community the limitationliving in vulnerableand its purpose areas and expressed whether there that are when less applying for national identityrestrictive cards means and to passportsachieve the Musl purposeims especially felt discriminated against. For example, when applying for a passport, Muslims are required to produce additional documentary evidence of citizenship. Whereas ‘Christian applicants only needed two birth 46certificates, The Republic their of Kenya. own Theand Constitution, of one of 2010.their Available parents, at applicantsfile:///Users/Anneli/Desktop/Constitu with Islamic names were- required to produce,tionofKenya%202010.pdf in addition, (accessed the birth on 10certificate January 2011). of one of the grandparents’. This was again

44 45 41

FCA_KenyaAnalysis.indd 45 24.1.2020 12.07

confirmed in a focus group discussion in Wajir as participants blamed this occurrence on corruption in government agencies (corruption will be discussed in greater detail below) such as immigration department and Office of Registrar of Persons. Participants referred to this state of affairs as a cause of an increased rate in radicalisation. Assessing this perception, respondentsSensitivity around as part producing of the an IDdetained can be traced and backfamily to the and colonial friends period samples when the w passere wasasked produced to rate government’sonly on demand success by a police in providing officer or anidentity authorised documents person in and accordance passports. with Ofthe law.the 89This rwasespondents seen as a as partmeasure of the to detainedprevent the sample free movement 43.37% of non-Europeansrated government’s in the country ability of to native provide Kenyans. identity Years documentsafter in- anddependence, passports militias 40.51% from above certain average.communities In werecontrast, demanding only 29.87% a display of IDfamily cards membersduring the 1992and tribalfriends 47 (n=82)clashes expressed to assist in identifyinga similar viewtheir victims.on the Aissuing similar of formula identity was documents applied in the and 1994 22.66% Rwanda with genocide. reference to passports, than respondents part of the detained sample (see Figures 41 and 42). The second sensitive topic associated with ID documents is the perceived unequal access to these docu- Figurements. 41 : MembersGovernment's of successthe community in providing living identity in documentsvulnerable areas expressed that when applying for national identity cards and passports they, Muslims especially, felt discriminated against. For example, when ap- confirmedplying for ain passport, a focus Muslims group are discussion required toin produce Wajir additionalas participants documentary blame evidenced this ofoccurrence citizenship. on corruptionWhereas ‘Christian in government applicants agencies only needed (corruption two birth will certificates, be discussed their inown greater and of detailone of below)their parents. such as immigrationApplicants with department Islamic names and were Office required of Registrar to produce, of inPersons. addition, Participants the birth certificate referred of oneto this of thestate Detained (n=89) 4.82 21.69 30.12 34.94 8.43 ofgrandparents’. affairs as Thisa causwas againe of confirmedan increased in a focus rate group in discussionradicalisation. in Wajir Assessingas participants this blamed perception this , respondentsoccurrence on as corruption part of inthe government detained agencies and family (corruption and will friends be discussed samples in greaterwere askeddetail below) to rate government’ssuch as immigration success department in providing and Officeidentity of documentsRegistrar of Persons.and passports. Participants Of referredthe 89 torespondents this state of as partaffairs of asthe a detainedcause of an sample increased 43.37% rate in radicalisation.rated government’s Assessing ability this perception, to provide respondents identity as documents part of andthe passportsdetained and 40.51% family and above friends average. samples wereIn contrast, asked to only rate 29.87% government’s of family success members in providing and identity friends documentsFamily (n=82) and6.49 passports. Of 28.57the 89 respondents as part of35.06 the detained sample 18.1843.37% rated11.69 govern- (n=82) expressed a similar view on the issuing of identity documents and 22.66% with reference ment’s ability to provide identity documents and passports 40.51% above average. In contrast, only 29.87% toof pas familysports members, than andrespondents friends (n=82) part expressed of the detained a similar viewsample on the (see issuing Figures of identity 41 and documents 42). and 22.66% with reference0% 10% to passports,20% than30% respondents40% part50% of the detained60% sample70% (see80% Figures90% 41 and 42100%). Figure 41: Government's success in providing identity documents 0 1-30 40-60 70-80 90-100

As explained in one focus group discussion, the Somali community has been subjected to profiling,Detained (n=89) harassment4.82 and21.69 massive corruption30.12 in the issuing of identity34.94 documents. According8.43 to one testimony: “It’s easier for an elderly Somali born in Mogadishu loaded with cash to access those key documents than a teenager born and raised in Wajir county hospital armed with birth certificate to get the same. And this will hinder their access to jobs and further studies adding to the vulnerabilities.” According to another respondent: “I have two families; one in Eldoret and the otherFamily one (n=82) in Wajir6.49 and I was28.57 blessed with children from35.06 both ends. High18.18 level discrimination11.69 is observed in the birth certificates issued to children in both places. As for the one born in Eldoret, the certificate indicates ‘this is a proof of citizenship’ while clearly marked on the one issued in Wajir county 0%indicates10% ‘this is20% not a pro30%of of 40%citizenship’50% a clear60% picture70% of ‘one80% family,90% one nation100% and different identities.” 0 1-30 40-60 70-80 90-100

Figure Figure 42 40.: Government's Government’s success success in providing in providing passports identity documents As explained in one focus group discussion, the Somali community has been subjected to profiling, harassment and massive corruption in the issuing of identity documents. According to one testimony: “It’s easier for an elderly Somali born in Mogadishu loaded with cash to access thoseDetained key (n=89) documents7.59 than a 24.05teenager born and raised27.85 in Wajir county 30.38hospital armed with10.13 birth certificate to get the same. And this will hinder their access to jobs and further studies adding to the vulnerabilities.” According to another respondent: “I have two families; one in Eldoret and the other one in Wajir and I was blessed with children from both ends. High level discrimination is

observedFamily (n=82) in the birth17.33 certificates issued 33.33to children in both places.26.67 As for the one13.33 born in9.33 Eldoret, the certificate indicates ‘this is a proof of citizenship’ while clearly marked on the one issued in Wajir county indicates ‘this is not a proof of citizenship’ a clear picture of ‘one family, one nation and different 0%identities10%.” 20% 30% 40% 50% 60% 70% 80% 90% 100%

Figure 42: Government's success in providing passports0 1-30 40-60 70-80 90-100

Figure 41. Government’s success in providing passports

Detained47 Gitonga (n=89) Muriuki.7.59 Demands by24.05 officers to see ID cards27.85 illegal. Daily Nation, 6 February30.38 2006. Available10.13 at 42 https://www.nation.co.ke/lifestyle/1190-106304-gkwrqkz/index.html (accessed on 15 January 2014).

46 47

Family (n=82) 17.33 33.33 26.67 13.33 9.33

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% FCA_KenyaAnalysis.indd 46 24.1.2020 12.07 0 1-30 40-60 70-80 90-100

As explained in one focus group discussion, the Somali community has been subjected to profiling, 8harassment.3 Service and delivery: massive corruptionEfficiency in the issuing of identity documents. According to one testimony: “It’s easier for an elderly Somali born in Mogadishu loaded with cash to access those key documents Athan key a teenagerfunction ofborn government and raised in that Wajir will county ultimately hospital influence armed with its birthlegitimacy certificate is which to get governmentthe same. is ableAnd this to willaddress hinder thetheir needs access of to jobsall its and citizens further studiesand provide adding topublic the vulnerabilities.” services, such According as health to care, another respondent: “I have two families; one in Eldoret and the other one in Wajir and I was blessed education, water and electricity etc. Through providing public services government interacts with thewith public children which from bothshape ends.s people’s High level trust discrimination in and expectations is observed inof the government. birth certificates The issuedproviding to of publicchildren services in both places.also contribute As for thes oneto welfare born in Eldoret,and economic the certificate growth. indicates It is therefore ‘this is a proof critical of that citizenship’ while clearly marked on the one issued in Wajir county indicates ‘this is not a proof of government responds to the needs of its citizens and that the delivery of these services addresses citizenship’ a clear picture of ‘one family, one nation and different identities.” 8the.3 Serviceneeds of delivery: the most vulnerableEfficiency .

AStarting 8.3key Service function with education, ofdelivery: government the Efficiency majority that will of ultimately all three samplesinfluence expressed its legitimacy a very is which high needgovernment to receive is ableeducationA key to function address (see of Figure thegovernment needs 43), alsoofthat all onewill its ultimately ofcitizens the main andinfluence areas provide its respondents legitimacy public services, is felt if itis discriminated able such to as address health against the care, as education,aneeds result of of all water financial its citizens and difficulties. electricit and providey etc public. Through services, providing such as phealthublic care,services education, government water interactand elecs- with the tricity public etc. Throughwhich shape providings people’s public services, trust in government and expectations interacts withof government. the public which The then providing shapes of Figure 43: Need to receive education publicpeople’s services trust in also and expectationscontributes toof government.welfare and The economic providing growth.of public Itservices is therefore also contributes critical that governmentto welfare and respond economics to thegrowth. needs It is of therefore its citizens critical and that that government the delivery responds of thes eto services the needs address of its es thecitizens needs and of thatthe mostthe delivery 4.2vulnerable of these. services addresses the needs of the most vulnerable. Detained (n=120) 1.68 7.56 35.29 51.26 StartingStarting with with education,education, thethe majority majority of ofall allthree three samples samples expressed expressed a very ahigh very need high to needreceive to edu receive- 0.93 educationcation (see (see Figure Figure 431.87), 43 also), alsoone of one the of main the areas main respondentsareas respondents felt discriminated felt discriminated against as aagainst result as a ofresult financial ofFamily financial difficulties. (n=108) difficulties. 11.21 27.1 57.94 0.93 Figure 43: Need to receive education Criminal Justice (n=105) 9.62 40.38 50

0%4.2 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Detained (n=120) 1.68 7.56 35.29 51.26 No 1-20% 30-40% 50-60% 70-80% 90-100% 0.93 1.87 Family (n=108) 11.21 27.1 57.94 In response to the0.93 overwhelming need to receive education, the majority of all three samples rated government’s ability to provide education between 70-80%, while a further 24.36% of familiesCriminal Justice and (n=105) friends 9.62and 21.52% amongst40.38 the criminal justice sample rated50 government’s success between 90-100%. This is a positive reflection on the Kenyan government and a crucial step towards building0% a country10% that20% is informed30% 40% and able50% to 60%pursue 70%a better80% life. 90%However,100% the

critical question is whetherNo government1-20% 30-40% and the50-60% business70-80% community90-100% create sufficient employment opportunities to accommodate especially the youth. Figure 42. Need to receive education InFigure response 44: Ability to ofthe government overwhelming to provide neededucation to receive education, the majority of all three samples rated government’s ability to provide education between 70-80%, while a further 24.36% of families and friends and 21.52% amongst the criminal justice sample rated government’s success betweenDetained 90- 100%(n=89) 3.61. This 13.25is a positive reflection27.71 on the Kenyan government45.78 and a crucial9.64 step towards building a country that is informed and able to pursue a better life. However, the critical question is whether government and the business community create sufficient employment opportunitiesFamily (n=82)to accommodate3.85 14.1 especially25.64 the youth. 32.05 24.36

Figure 44: Ability of government to provide education

Criminal Justice (n=79) 7.59 26.58 44.3 21.52

Detained (n=89) 3.610% 13.2510% 20% 27.7130% 40% 50% 60% 45.7870% 80% 90%9.64 100%

0 1-30 40-60 70-80 90-100 Figure 43. Ability of government to provide education Family (n=82) 3.85 14.1 25.64 32.05 24.36

46 47 Criminal Justice (n=79) 7.59 26.58 44.3 21.52 43

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

0 1-30 40-60 70-80 90-100 FCA_KenyaAnalysis.indd 47 24.1.2020 12.07

43 In response to the overwhelming need to receive education, the majority of all three samples rated a pull quote here? lorem government’s ability to provide education between 70-80%, while a further 24.36% of families and friends and 21.52% amongst the criminal justice sample rated government’s success between 90- ipsum dolor sit.Women’s 100%. This is a positive reflection on the Kenyan government and a crucial step towards building a Bank made its fundraising country that is informed and able to pursue a better life. However, the critical question is whether government and the business community create sufficient employment opportunities to accommo- record and raised over date especially the youth. 1.8 MEUR to support women in the developing In addition to education, respondents rated their need (Figure 45) for healthcare higher than what they actually received (Figure 46). countries. Although respondents did not experience the same level of success providing in their healthcare needs, it was less than the success they experienced in the provision for education. Overall, education proved to be another sensitive topic considering the opening of opportunities and further lessening In additionthe associated to education, probability respondents of being manipulated rated their due need to illiteracy. (Figure 45) for healthcare higher than what they actually received (Figure 46).

Figure 45: Need for healthcare

2.5 Detained (n=120) 4.17 6.67 24.17 62.5

0.95 0.95 Family (n=108) 5.71 20 71.43 0.95

Criminal Justice 0.966.73 29.81 62.5 (n=105)

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 44. Need for healthcare Figure 46: Government's success in providing healthcare

Detained (n=89) 4.65 20.93 31.4 33.72 9.3

1.3 Family (n=82) 9.09 32.47 37.66 19.48

Criminal Justice (n=79) 16.46 25.32 46.84 11.39

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

0 1-30 40-60 70-80 90-100

AlthoughFigure respondents45. Government’s did success not exp in providingerience healthcare the same level of success providing in their healthcare needs it was less than the success they experienced in the provision for education. Overall, education proved to be another sensitive topic considering the opening of opportunities and further lessening the associated probability of being manipulated due to illiteracy.

8.4 Safety from foreign enemies An internationally recognised responsibility placed on governments through its military is to protect48 the country against foreign intervention. Al-Shabaab incursions from Somalia into Kenya 49 is probably the country’s main threat from foreign enemies. It is also this threat that prompted the Kenyan government to deploy troops to Somalia in October 2011 after al-Shabaab was implicated in a series of kidnappings along the Kenyan coast. The deployment of the Kenyan Defence Force unfortunately did not prevent Kenya from being the target of attacks, most notably the Garissa University attacks and Westgate that had been planned and executed FCA_KenyaAnalysis.inddfrom 48 Somalia in response to Kenyan presence in that country. 24.1.2020 12.07

44 8.4 Safety from foreign enemies

An internationally recognised responsibility placed on governments through its military is to protect the country against foreign intervention. Al-Shabaab incursions from Somalia into Kenya is arguably the country’s main threat from foreign enemies. It is also this threat that prompted the Kenyan government to deploy troops to Somalia in October 2011 after al-Shabaab was implicated in a series of kidnappings along the Kenyan coast. The deployment of the Kenyan Defence Force unfortunately did not prevent Kenya from being the target of attacks, most notably the Garissa University attacks and Westgate, that had been planned and executed from Somalia in response to Kenyan presence in that country.

Before assessing the success of the Kenyan government to protect its citizens, respondents were asked to rate the need to be protected from a foreign enemy. It is important to note that all three samples expressed a very similar need to be protected against a foreign enemy, that can be interpreted as a similar threat perception. Figure 47: Need to be protected against a foreign enemy Whereas the majority (49.37%) of criminal justice respondents rated government’s ability to protect the country against 2.54foreign enemies between 70-80% the majority of families (39.19%) and 35.63% amongstDetained the detained(n=120) rated4.24 9.23success between15.25 40-60%. It31.36 is however important to note37.29 that both samples’ second classification was between 70-80%, while 21.52% of criminal justice respondents rated government’s success at the same level. Figure 47: Need to be protected against a foreign enemy Family (n=108) 5.71 6.67 16.19 25.71 42.86 2.86

2.54 CriminalDetained Justice (n=120) (n=105) 1.964.245.889.23 13.7315.25 35.2931.36 41.1837.29 1.96 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Family (n=108) 5.71 6.67 16.19 25.71 42.86 No 1-20% 30-40% 50-60% 70-80% 90-100% 2.86

WhereasCriminal Justice the (n=105) majority1.965.88 (49.37%)13.73 of criminal 35.29justice respondents rated government’s41.18 ability to protect the country1.96 against foreign enemies between 70-80% the majority of families (39.19%) and 35.63% amongst0% the10% detained20% rated30% success40% between50% 60%40-60%.70% It is however80% 90% important100% to

note that both samples’ secondNo 1-20% classification30-40% was50-60% between70-80% 70-90-100%80%, while 21.52% of criminal justice respondents rated government’s success at the same level. Figure 46. Need to be protected against a foreign enemy WhereasFigure 48: Successthe majority of the Kenyan (49.37%) government of to criminalprotect citizens justice against respondents foreign enemies rated government’s ability to protect the country against foreign enemies between 70-80% the majority of families (39.19%) and 35.63% amongst the detained rated success between 40-60%. It is however important to note thatDetained both (n=89) samples6.9 ’ second20.69 classification was 35.63between 70-80%, while28.74 21.52% of8.05 criminal justice respondents rated government’s success at the same level.

Figure 48: Success of the Kenyan government to protect citizens against foreign enemies Family (n=82) 13.51 16.22 39.19 21.62 9.46

CriminalDetained Justice (n=89) (n=79) 6.910.13 20.69 21.52 35.63 49.37 28.74 17.728.05 1.27 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Family (n=82) 13.51 16.22 39.19 21.62 9.46 0 1-30 40-60 70-80 90-100

Figure 47. Success of the Kenyan government to protect citizens against foreign enemies 8Criminal.5 Safety Justice (n=79) from domestic10.13 enemies21.52 49.37 17.72 1.27 Respondents were0% asked10% to assess20% the30% most pressing40% 50%security60% challenges70% in 80%Kenya. 90%Although100% these individual threat perceptions will0 be1-30 presented40-60 under70-80 the90-100 relevant headings it is important to 48 note that according to both families and friends as well as detained respondents, unequal49 access to land and natural resources was regarded as the primary threat to domestic security. This was 8followed.5 Safety by from terrorism, domestic historic enemies grievances based on ethnicity and religion, criminal activities and political grievances. Criminal justice respondents rated the latter as the overall most pressing Respondentssecurity threat, were followed asked to by assess unequal the most access pressing to land security and natural challenges resources, in Kenya historic. Although grievances these FCA_KenyaAnalysis.indd individual 49based on threat ethnicity, perception criminals willactivities be presented and terrorism. under Onlythe relevant a small headings minority itof is respondents important to – 24.1.2020 12.07 noterepresenting that according all three to both samples families – consideredand friends conflictas well as detaineda result of respondents climate change, unequal and access cattle torustling land and as anatural serious resources threat. was regarded as the primary threat to domestic security. This was followed by terrorism, historic grievances based on ethnicity and religion, criminal activities and politicalStarting grievances. with terrorism Criminal as a main justice threat respo tondents domestic rated security, the latter it is interestingas the overall to note most that pressing instead securityof criminal threat, justice followed actors, by it unequalwas respondents access to beingland andpart natural of family resources, members historic and grievancesfriends who based on ethnicity, criminal activities and terrorism. Only a small minority of respondents – representing all three samples – considered conflict as a result of climate change and cattle rustling as a serious threat. 45

Starting with terrorism as a main threat to domestic security, it is interesting to note that instead of criminal justice actors, it was respondents being part of family members and friends who

45 8.5 Safety from domestic enemies

Respondents were asked to assess the most pressing security challenges in Kenya. Although these individual threat perceptions will be presented under the relevant headings, it is important to note that according to both families and friends as well as detained respondents, unequal access to land and natural resources was regarded as the primary threat to domestic security. This was followed by terrorism, historic grievances based on ethnicity and religion, criminal activities and political grievances. Criminal justice respondents rated the latter as the overall most pressing security threat, followed by unequal access to land and natural resources, historic grievances based on ethnicity, criminal activities and terrorism. Only a small minority of respondents – representing all three samples – considered conflict as a result of climate change and cattle rustling as a serious threat.

Starting with terrorism as a main threat to domestic security, it is interesting to note that instead of ratedcriminal the threat justice of actors, terrorism it was the respondents highest. A being possible part explanationof family members for this and high friends threat who perception rated the on thethreat part oficular terrorism group the highest.was found A possible in one explanation focus group for thisdiscussion high threat when perception participants on the noted particular that group was found in one focus group discussion when participants noted that threats from terrorists, threats from terrorists, due to its indiscriminate nature was as serious as indiscriminate due to its indiscriminate nature was as serious as indiscriminate countermeasures by security agents. countermeasures by security agents. rated In athe follow-up threat questionof terrorism in the the questionnaire highest. A respondentspossible explanation were asked forto rate this the high threat threat of al-Shabaab perception Figure 49: Terrorism as a main threat to domestic security on theand part Islamicicular State group attacks was in foundKenya. in one focus group discussion when participants noted that threats from terrorists, due to its indiscriminate nature was as serious as indiscriminate countermeasuresIn contrast to al-Shabaab, by1.92 security theagents. majority of respondents representing all three samples did not recog- Criminalnise Justice the possibility(n=104) of8.65 Islamic State18.27 executing15.38 attacks in Kenya. 36.54 19.23 Figure 49: Terrorism as a main threat to domestic security

Detained (n=121) 1.928.55 13.68 8.55 23.93 17.95 27.35 Criminal Justice (n=104) 8.65 18.27 15.38 36.54 19.23 2.91 Family (n=106) 3.88 15.53 15.59 23.3 40.78 Detained (n=121) 8.55 13.68 8.55 23.93 17.95 27.35 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 2.91 No 1-20% 30-40% 50-60% 70-80% 90-100% Family (n=106) 3.88 15.53 15.59 23.3 40.78

In a follow-up question0% in10% the questionnaire20% 30% respondents40% 50% were60% asked70% to rate80% the 90%threat 100%of al- Shabaab and Islamic State attacks in Kenya. No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 50: Threat of al-Shabaab attacks in Kenya Figure 48. Terrorism as a main threat to domestic security In a follow-up question in the questionnaire respondents were asked to rate the threat of al- Shabaab and Islamic State attacks in Kenya. Criminal Justice (n=105) 5.83 6.8 22.33 38.83 26.21 Figure 50: Threat of al-Shabaab attacks in Kenya

Detained (n=123) 4.1 12.3 18.85 19.67 21.31 23.77 Criminal Justice (n=105) 5.83 6.8 22.33 38.83 26.21

Family (n=111) 3.67 6.42 15.6 19.27 24.77 30.28 Detained (n=123) 4.1 12.3 18.85 19.67 21.31 23.77 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100% Family (n=111) 3.67 6.42 15.6 19.27 24.77 30.28 Figure 49. Threat of al-Shabaab attacks in Kenya In contrast to al-Shabaab,0% 10% the majority20% 30% of respondents40% 50% representing60% 70% all three80% samples90% did100% not recognise the possibility of Islamic State executing attacks in Kenya. 50 No 1-20% 30-40% 50-60% 70-80% 90-100% 51

Figure 51: Threat of Islamic State attacks in Kenya In contrast to al-Shabaab, the majority of respondents representing all three samples did not recognise the possibility of Islamic State executing attacks in Kenya.

FCA_KenyaAnalysis.inddFigure 50 51: Threat of Islamic State attacks in Kenya 24.1.2020 12.07

CriminalCriminal Justice Justice (n=105) (n=105) 32.6932.69 24.0424.04 15.3815.38 10.5810.58 8.658.65 8.658.65 Criminal Justice (n=105) 32.69 24.04 15.38 10.58 8.65 8.65 1.671.67 1.671.67 1.67 1.67 DetainedDetained (n=123) (n=123) 35.8335.83 22.522.5 2525 13.3313.33 Detained (n=123) 35.83 22.5 25 13.33

FamilyFamily (n=111) (n=111) 44.3444.34 12.2612.26 16.0416.04 9.439.43 10.3810.38 7.557.55 Family (n=111) 44.34 12.26 16.04 9.43 10.38 7.55

0%0% 10%10% 20%20% 30%30% 40%40% 50%50% 60%60% 70%70% 80%80% 90%90% 100%100% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% NoNo 1-20%1-20% 30-40%30-40% 50-60%50-60% 70-80%70-80% 90-100%90-100% No 1-20% 30-40% 50-60% 70-80% 90-100% Figure 50. Threat of Islamic State attacks in Kenya With With reference reference to to other other criminal criminal activities activities when when compared compared to to terrorism, terrorism, criminal criminal justice justice respondentsWithrespondents reference rated rated to this thisother threat threat criminal to to domestic domestic activities security security when higher. higher. compared This This sentiment sentiment to terrorism, was was however however criminal not not sharedjustice shared byrespondentsby respondents respondents rated being thisbeing threat part part toof ofdomestic the the detained detained security and higher.and family family This andsentiment and friends friends was samples. howeversamples. A not Apossible sharedpossible byexplanation respondentsWith reference for beingth isto different other part criminal of threat the activities perceptiondetained when and may compared family be that to and interrorism, the friends mind criminal ofsamples. criminal justice justiceA respondents possible actors explanationrated this for threat this different to domestic threat security perception higher. Thismay sentiment be that wasin the however mind of not criminal shared justiceby respondents actors alexplanational-Shabaab-Shabaab forwas was th unableis unable different to to execute threatexecute perceptiona amajor major terrorist terroristmay be attack thatattack in ( on the(on the mindthe same same of criminal level level as asjustice Westgate Westgate actors), ), al-Shabaabbeing partwas of unable the detained to execute and familya major and terrorist friends samples. attack (Aon possible the same explanation level as forWestgate this differ), - whilewhile membersent members threat perceptionof of the the community community may be that(including (including in the mindsecurity security of criminal officials) officials) justice ha ha dactors dbeen been al-Shabaab the the victim victim swas sof of unablesmaller smaller to attackswhileattacks members executed executed of by theby al al-communityShabaab-Shabaab operatives(including operatives securityin in remote remote officials) areas. areas. ha d been the victims of smaller attacksexecute executed a major by al terrorist-Shabaab attack operatives (on the same in remote level as areas.Westgate), while members of the community (including security officials) had been the victims of smaller attacks executed by al-Shabaab opera- Figure Figure 52 52: Crime: Crime as asa threata threat to todomestic domestic security security Figure 52tives: Crime in remoteas a threat areas. to domestic security

DetainedDetained (n=121) (n=121) 7.027.02 13.1613.16 21.9321.93 18.4218.42 17.5417.54 21.9321.93 Detained (n=121) 7.02 13.16 21.93 18.42 17.54 21.93

FamiliesFamilies (n=106) (n=106) 9.629.62 9.629.62 20.1920.19 21.1521.15 19.2319.23 20.1920.19 Families (n=106) 9.62 9.62 20.19 21.15 19.23 20.19 2.942.94 2.94 CriminalCriminal Justice Justice (n=104) (n=104) 5.885.88 18.2718.27 32.3532.35 31.3731.37 8.628.62 Criminal Justice (n=104) 5.88 18.27 32.35 31.37 8.62

Figure 53: Serious nature of criminal activities: Criminal justice respondents Figure 53Figure: Serious 51. natureCrime ofas criminal a threat activities: to domestic Criminal security justice respondents Figure 53: Serious nature of criminal activities: Criminal justice respondents

2.882.88 2.88 DrugDrug abuse abuse 7.697.69 12.512.5 20.1920.19 31.7331.73 2525 Drug abuse 7.69 12.5 20.19 31.73 25

DrugDrug smuggling smuggling 5.775.77 8.658.65 12.512.5 24.0424.04 30.7730.77 18.2718.27 Drug smuggling 5.77 8.65 12.5 24.04 30.77 18.27

OrganisedOrganised crime crime 5.775.77 13.4613.46 13.4613.46 26.9226.92 29.8129.81 10.5810.58 Organised crime 5.77 13.46 13.46 26.92 29.81 10.58

ViolentViolent crime crime4.854.85 8.748.74 16.516.5 28.1628.16 31.0731.07 10.6810.68 Violent crime 4.85 8.74 16.5 28.16 31.07 10.68 0%0% 10%10% 20%20% 30%30% 40%40% 50%50% 60%60% 70%70% 80%80% 90%90% 100%100% 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% NoNo 1-20%1-20% 30-40%30-40% 50-60%50-60% 70-80%70-80% 90-100%90-100% No 1-20% 30-40% 50-60% 70-80% 90-100% Figure 52. Serious nature of criminal activities: Criminal justice respondents

4747 50 4751

FCA_KenyaAnalysis.indd 51 24.1.2020 12.07

Figure 54: Serious nature of criminal activities: Family and friends respondents

Drug abuse 3.7 5.56 7.41 14.81 28.7 39.81

3.74 Drug smuggling 7.48 10.28 23.36 22.43 32.71

Organised crime 7.62 10.48 14.29 22.86 21.9 22.86

Violent crime 7.69 7.69 23.08 27.88 18.27 15.38

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 53. Serious nature of criminal activities: Family and friends respondents Figure 55: Serious nature of criminal activities: Respondents history of being detained

Drug abuse 4.24 6.78 9.32 24.58 27.12 27.97

Drug smuggling 4.2 8.4 10.92 24.37 26.05 26.05

Organised crime 6.19 9.73 18.58 29.2 17.7 18.58

Violent crime 6.03 12.07 20.69 30.17 12.93 18.1

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 54. Serious nature of criminal activities: Respondents history of being detained The above prompted the question whether threats presented by illegal organisations were being accurately assessed. It is important to note that while the majority of criminal justice actors (43%) and family members and friends (38.68%) regarded the threat presented by illegal organisationsThe above prompted as being the questionunderestimated, whether threats the majority presented within by illegal the organisations detained sample were being (43.44%) ac- consideredcurately assessed. the threat It is presented important to by note illegal that organisationswhile the majority as ofbeing criminal overestimated. justice actors Th(43%)e fact and that thefamily majority members of respondents and friends (38.68%) as part regardedof the detained the threat sample presented we reby beingillegal organisationsimplicated in as terrorism being - relatedunderestimated, offences the(and majority not withinregarded the detained as ‘victims’ sample of (43.44%) al-Shabaab) considered may the shedthreat lightpresented on this perception.by illegal organisations Despite these as beingdifferences overestimated. 41% amongst The fact criminal that the justicemajority actors, of respondents 36.79% as amongst part familyof the members detained andsample friends were andbeing 36.07% implicated amongst in terrorism-related the detained offences sample (and considered not regarded the threat as presented‘victims’ of by al-Shabaab) illegal organisations may shed light as beingon this assessedperception. accurately. Despite these differences 41% amongst criminal justice actors, 36.79% amongst family members and friends and 36.07% amongst the de- Figuretained 56: Securitysample threat considered overestimated the threat presented by illegal organisations as being assessed accurately.

52 53

FCA_KenyaAnalysis.indd 52 24.1.2020 12.07

100% 16 90% 24.53 80% 43.44 70% 41 60% 36.79 50%

40% 36.07 30%

20% 38.68 43 10% 20.49

0% Families (n=106) Detained (n=122) Criminal Justice (n=100)

Underestimate Accutate Overestimate

Figure 55. Security threat overestimated 8.5.1 Radicalisation and recruitment into illegal organisations

Under8.5.1 Section Radicalisation 12D of andthe Preventionrecruitment of into Terrorism illegal organisationsAct, radicalisation is defined as ‘a person who adoptsUnder or Section promotes 12D anof theextreme Prevention belief of system Terrorism for Act,the purposeradicalisation of facilitating is defined ideologicallyas ‘a person whobased violenceadopts to or advancepromotes political,an extreme religious belief system or social for thechange purpose commits of facilitating an offence ideologically and is liablebased on conviction,violence toto advance imprisonment political, for religious a term or socialnot exceedingchange commits thirty an years.’ offence Considering and is liable onthe convic serious- consequencestion, to imprisonment of radicalisation, for a term not recruitment exceeding thirtyand years.’the laterConsidering possible the seriousinvolvement consequences of those radicalisedof radicalisation, in acts recruitmentof terrorism and in theand later outside possible Kenya involvement, especially of those respondents radicalised representing in acts of ter -the familiesrorism andin and friends outside who Kenya, went especially through respondents the criminal representing justice system the familiesrecognised and friendsthe serious who went nature of throughKenyans the being criminal recruited justice systeminto illegal recognised organisations the serious such nature as al of- ShabaabKenyans being. 67.6% recruited of the into said groupillegal classif organisationsied the serious such as natureal-Shabaab. of the 67.6% recruitment of the said of groupKenyan classified nationals the seriousabove natureaverage of the (70 - 100%)recruitment (see Figure of Kenyan 57) .nationals above average (70-100%) (see Figure 57).

ThisThis section section can can be be divideddivided into into two two primary primary sub-sections: sub-sections: in the In first, the the first, objective the objective will be to willidentify be to identifywhy individuals why individuals may be vulnerablemay be vulnerable to be recruited to be into recru illegalited organisations. into illegal Answersorganisations. from all Answersthree fromsamples all three were samples based on were perceptions based onto guardperceptions against to potential guard againstconcern potentialon the part concern of the ontwo the community-based samples that might incriminate themselves or a person close to them. The second part of the two community-based samples that might incriminate themselves or a person close to them.part will The focus second on whether part will respondents focus on werewhether aware respondents of initiatives were to prevent aware radicalisation of initiatives and to recruit prevent- radicalisationment and how and it was recruitment being perceived. and how it was being perceived.

Figure 57: Seriousness of Kenyan2.54 nationals being recruited into illegal organisations

Detained (n=123) 11.02 14.41 23.73 21.19 27.12

3.7 Families (n=111) 1.85 12.04 14.81 30.56 37.04

0.97

Criminal Justice (n=105) 5.83 4.85 18.45 41.75 28.16

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 56. Seriousness of Kenyan nationals being recruited into illegal organisations Considering the above average concern surrounding the potential recruitment of Kenyan nationals into illegal organisations, respondents were asked to identify the primary reasons that 52 might motivate individuals to join these organisations. The most critical reasons – most notably5349 anger towards government and its security forces – will be discussed separately. Table 5 compares perceptions amongst the three samples in relation to above average indicators. What is particular noticeable is the different perceptions on the part of criminal justice actors in relation to the other two samples, considered to be closer to understanding the driving factors

FCA_KenyaAnalysis.indd 53that facilitate radicalisation. Although financial incentives play an important role (according to 24.1.2020 12.07 detained respondents rated as fourth most important (59.46%)), it is a long way from being the primary incentive (80.8%) according to criminal justice respondents.

Table 5: Reasons for joining illegal organisations based on priority

Priority Detained (n=117) Family (n=105) Criminal Justice (n=103) 1 Anger towards the police Anger towards the police Financial incentives (80.8%) (66.96%) (75.97%0

2 Anger towards the Employment (71.29%) Religious ideology (65.34%) government (60.18%)

3 Employment1 (59.63%) Financial incentives Political ideology (63.63%) (68.32%) 4 Financial incentives Anger towards the Anger towards the (59.46%) government (62.5%) government (63.36%)

5 Treatment from security Religious ideology Anger towards the police forces since arrest (60.78%) (59.4%) (56.88%)

6 Religious ideology (52.3) Treatment from security Frustration over situation at forces since arrest home (57%) (52.47%)

7 Anger towards AMISOM Frustration over situation Anger towards the military (43.39%) at home (50%) (56%)

1 Not included in Criminal Justice questionnaire that referred to financial incentives

50 Considering the above average concern surrounding the potential recruitment of Kenyan nationals into illegal organisations, respondents were asked to identify the primary reasons that might motivate individuals to join these organisations. The most critical reasons – most notably anger towards government and its security forces – will be discussed separately. Table 5 compares perceptions amongst the three samples in relation to above average indicators. What is particular noticeable is the different perceptions on the part of criminal justice actors in relation to the other two samples, considered to be closer to understanding the driving factors that facilitate radicalisation. Although financial incentives play an important role (according to detained respondents rated as fourth most important (59.46%), it is a long way from being the primary incentive (80.8%) according to criminal justice respondents.

Table 5. Reasons for joining illegal organisations based on priority

Priority Detained (n=117) Family (n=105) Criminal Justice (n=103) 1 Anger towards the police Anger towards the police (75.97%) Financial incentives (80.8%) (66.96%) 2 Anger towards the government Employment (71.29%) Religious ideology (65.34%) (60.18%) 3 Employment48 (59.63%) Financial incentives (68.32%) Political ideology (63.63%) 4 Financial incentives (59.46%) Anger towards the government Anger towards the government (62.5%) (63.36%) 5 Treatment from security forces Religious ideology (60.78%) Anger towards the police (59.4%) since arrest (56.88%) 6 Religious ideology (52.3) Treatment from security forces since Frustration over situation at home arrest (52.47%) (57%) 7 Anger towards AMISOM Frustration over situation at home Anger towards the military (56%) (43.39%) (50%) 8 Anger towards the military Anger towards intelligence agencies Anger towards AMISOM (54.45%) (41.96%) (47.47%) 9 Means to settle personal Anger towards AMISOM (44.9%) Becoming friends with existing disputes (40.37%) members (52.04%) 10 Anger towards intelligence Anger towards the military (44.55%) Treatment from security forces agencies (39.82%) since arrest (52%) 11 Political ideology (36.7%) Political ideology (43.56%) Anger towards intelligence agencies (49.5%) 12 Frustration over situation at Becoming friends with existing Being isolated from other opinions home (36.36%) members (38%) (46%) 13 Conditions in prison (34.61%) Being isolated from other opinions Means to settle personal scores (33.33%) (36.28%) 14 Being isolated from other Means to settle personal disputes Protection from the group while in opinions (34.29) (31%) custody (35%) 15 Becoming friends with existing Sense of belonging (27.55%) Protection to family - by the group members (30.84%) - while in custody (35%) 16 Sense of belonging49 (30.56%) Protection from the group while in Respect members of the custody (24.24%) organisation (31%) 17 Protection from the group while Conditions in prison (24.24%) Conditions in prison (27.27%) in custody (30.19%) 18 Need to be respected (27.1%) Need to be respected (22.77%) Need to be respected (19.19%) 19 Protection to family - by the Protection to family - by the group - group - while in custody (25%) while in custody (22.55%) 20 Respect members of the Respect members of the organisation organisation (21.3%) (22.45%) 20 Respect members of the Respect members of the organisation organisation (21.3%) (22.45%)

48 Not included in Criminal Justice questionnaire that referred to financial incentives 49 Not included in Criminal Justice questionnaire, easier to determine a sense of belonging from a community perspective. 54 55

FCA_KenyaAnalysis.indd 54 24.1.2020 12.07

Priority Family (n=105) Detained (n=116) Criminal Justice (n=103) 1 Social media Between friends Internet Since respondents being part of the detained sample who went through the criminal justice system 2 were not directlyInternet asked how they had been radicalised,Social media Table 6 summarises theSocial facilitators media respond - 3 ents perceiveVideos to be / furtheringCDs radicalisation andInternet recruitment. Between friends 4 Between friends Videos / CDs Mosque 5 Interaction with foreigners in Interaction with Videos / CDs Table 6. RadicalisationKenya facilitated through: foreigners in Kenya

6 Priority FamilyRefugee (n=105) Camps DetainedPoor (n=116) neighbourhoods CriminalMadrassa Justice (n=103) 7 1 SocialMadrassa media Between Refugeefriends Camps Internet University / Colleges 2 Internet Social media Social media 8 3 VideosMosque / CDs Internet Within families BetweenRefugee friends Camps 9 4 BetweenPoor neighbourhoods friends Videos / MadrassaCDs Mosque Poor neighbourhoods 10 5 InteractionUniversity with / foreigners Colleges in Kenya InteractionMosque with foreigners in Kenya Videos /Interaction CDs with foreigners 6 Refugee Camps Poor neighbourhoods Madrassain Kenya 7 Madrassa Refugee Camps University / Colleges 11 8 MosquePrison Within familiesCommunity Centers RefugeeSecondary Camps school 12 9 PoorSecondary neighbourhoods school MadrassaUniversity / Colleges Poor neighbourhoodsCommunity Centers 10 University / Colleges Mosque Interaction with foreigners in Kenya 13 11 PrisonWithin families CommunityPrison Centers SecondaryPrison school 14 12 SecondaryCommunity school Centers UniversitySecondary / Colleges school CommunityWithin Centers families 15 13 WithinPrimary families school Prison Primary school Prison Primary school 14 Community Centers Secondary school Within families 15 Primary school Primary school Primary school

Figure 58: Most vulnerable to be recruited into an illegal organisation

80 67.32 67.31 70 66.34 55.05 60 53.4 54.46 51.81 49.02 47.71 50 50 36.69 40 34.23 30 20 10 0 Individuals from People who Poor Uneducated High school University students broken families originally wanted graduates to join the police/military

Detained (n=113) Family (n=105)

Figure 57. Most vulnerable to be recruited into an illegal organisation In the following part of the discussion the focus will be on counter strategies. To be effective, people, especially those who are vulnerable, need to be aware of its existence. Although the majorityIn the offollowing all three part ofsamples the discussion were theaware focus ofwill counter be on counter initiatives strategies. at the To be time effective, of the people, interviews, 32.06especially% of detained those who respondents are vulnerable, were need not to be aware aware of of counterits existence. initiatives. Although the majority of all three samples were aware of counter initiatives at the time of the interviews, 32.06% of detained Figurerespondents 59: Aware of wereinitiatives not toaware prevent of people counter from initiatives. joining illegal organisations

54 55

FCA_KenyaAnalysis.indd 55 24.1.2020 12.07

90 82.69 8090 74.58 82.69 67.94 7080 74.58 67.94 6070

5060

4050 32.06 40 30 25.42 32.06 25.42 17.31 2030 17.31 1020

010 Families (n=118) Detained (n=131) Criminal Justice (n=104) 0 Families (n=118) Detained (n=131) Criminal Justice (n=104) Yes No Yes No

The majorityFigure 58. Awareof both of initiatives samples to hadprevent been people aware from joining of counter illegal organisations initiatives for a year and longer. The majority of both samples had been aware of counter initiatives for a year and longer. Figure 60: Duration being aware of counter measures Figure 60: Duration being aware of counter measures 30 30 24.1 24.1 25 23.86 24.1 24.1 25 20.45 23.86 19.32 19.32 20 20.45 18.07 19.32 19.32 20 18.07 14.46 15 13.64 14.46 12.05 15 13.64 12.05 10 8.43 10 8.43 5 2.27 5 1.14 2.27 0 1.14 Less than a month 1 - 6 months 7 - 12 months 1 - 2 years 2 - 3 years 4 - 5 years Longer than 5 0 years Less than a month 1 - 6 months 7 - 12 months 1 - 2 years 2 - 3 years 4 - 5 years Longer than 5 years Families (n=88) Detained (n=83) Families (n=88) Detained (n=83)

Respondents Figure 59. Durationrepresenting being aware families of counter and measures friends as well as the detained sample identified non- governmentalRespondents andrepresenting religious organisationsfamilies and friendsas taking as thewell lead, as the while detained criminal sample justice identified respondents non - referredgovernmental to the Kenyanand religious government, organisations NGOs andas taking religious the organisationslead, while criminal (see Figure justice 61 respondents). referredThe majority to the of Kenyan both samples government, had been NGOs aware andof counter religious initiatives organisations for a year (see and Figurelonger. 61). Figure 61: Leading agency responsible for counter initiatives Figure 61: Leading agency responsible for counter initiatives Respondents representing families and friends as well as the detained sample identified non-governmental and religious organisations as taking the lead, while criminal justice respondents referred to the Kenyan government, NGOs and religious organisations (see Figure 61). All three samples referred to community dialogue as the leading initiative, followed by religious dialogue and education and skills development.

All three samples referred to community dialogue as the leading initiative, followed by religious dialogue and education and skills development.

Of above initiatives, criminal justice respondents (n=95) participated in 65.26% of initiatives, followed by 61.68% of family and friends (n=107) and 53.33% of detained respondents (n=120).

56 53 57 53

FCA_KenyaAnalysis.indd 56 24.1.2020 12.07

Women 21.84 6.9 14.12

International 18.39 6.9 18.82 organisation

NGO 58.62 47.13 58.82 1.15 Another 11.76 country

Youth 25.29 32.18 24.71 organisation

Religious 49.43 45.98 40 organisations 1.15 Prison 10.59

School 12.64 8.05 14.12

Local Political Leaders 13.79 17.24

Government 36.78 31.03 75.29

0 20 40 60 80 100 120 140 160 180

Families (n=87) Detained (n=87) Criminal Justice (n=85)

AllFigure three 60. samples Leading agency referred responsible to community for counter dialogueinitiatives as the leading initiative, followed by religious dialogue and education and skills development.

Figure 62: Type of initiatives

2.3 Service 11.63 23.26 delivery

Sport 38.37 35.63 43.02

Religious 52.33 54.02 50 discussion

Employment 20.93 18.39 40.7 programmes Education & skills 34.88 25.29 53.49 development 1.15 Prison programs 19.77 1.16

Amnesty 15.12 20.69 33.72

Community 84.88 75.86 75.58 dialogue

0 50 100 150 200 250

Families (n=86) Detained (n=87) Criminal Justice (n=86)

Figure 61. Type of initiatives

56 57 54

FCA_KenyaAnalysis.indd 57 24.1.2020 12.07

Of above initiatives, criminal justice respondents (n=95) participated in 65.26% of initiatives, followed by 61.68% of family and friends (n=107) and 53.33% of detained respondents (n=120).

Criminal justice respondents overall expressed more trust in counter initiatives than respondents being part of the family and friends sample. The latter particularly had suspicions of the success of community dialogue initiatives (14.74%), whereas prison programs (81.82%) and the amnesty process (74.07%) received the most support. In contrast criminal justice actors expressed the least trust in the success of amnesty and prison projects. Detained respondents expressed limited trust in community dialogue (23.96%) and religious discussion. These results are particularly important considering the emphasis placed on these initiatives. It therefore calls for an evaluation into the reasons why those most at risk did not consider the said initiatives.

Figure 63: Assess the success of counter-initiatives (70-100% success rate)

80 81.82 74.07 70 69.14 63.86 60.71 60.71 61.84 60 58.75 58.44 56.32 56.32 49.4 50 51.9 51.81 51.36 45.98 43.68 41.38 40.45 41.33 40 42.22 36.67 30 30.93 23.96 23.68 20 20.51 14.74 10

0 Community Amnesty Prison Education and Employment Religious Sport events Youth-led Women-led dialogue programs skills programs discussions initiatives initiatives development

Family (n=98) Detained (n=102) Criminal Justice (n=85)

ConsideringFigure 62. theAssess limited the success perception of counter-initiatives of success, (70-100% respondents success rate) were asked why these counter measures were not successful. According to both the detained (61.46%) and family and friends (49.04%) samples, friends were more convincing countering the arguments presented by those offering counter-messaging. Although not supported by the majority, 54.54% of criminal justice respondentsCriminal justicesupported respondents this perception. overall expressed Criminal more justice trust respondents in counter initiatives were largely than respondents under the being part of the family and friends sample. The latter particularly had suspicions of the success of impression that the ideals of the illegal organisation (59.09%) were more convincing, a community dialogue initiatives (14.74%), whereas prison programs (81.82%) and the amnesty process perception(74.07%) that received was notthe mostcompletely support. supported In contrast criminalby the otherjustice two actors samples expressed. Therefore, the least trustinstead in the of relyingsuccess on counter of amnesty-messaging and prison through projects. the Detained media, respondentspeer pressure expressed is far limitedmore ctrustonvincing. in community What makesdialogue counter (23.96%)-messaging and religiousthrough discussion. peers particularly These results challenging are particularly relates important to the consideringbasic principle the of socialisationemphasis placed as people on these (irrespective initiatives. Itof therefore age) associate calls for withan evaluation others who into share the reasons similar why opinions. those Consequently,most at risk asdid anot result consider of thegroupthink said initiatives. that often emerges, the possibility of a different perspective being presented to the rest of the group and others being open to this different opinionConsidering will become the limited less the perception more ‘controversial’ of success, respondents the counter were opinion. asked In why other these words, counter since measures friends sharewere a particular not successful. opinion According (for example to both the more detained extreme (61.46%) views) and ,family the possibility and friends of(49.04%) one or sam two- peopleples, being friends accommodated were more convincing with counteringmoderate theopinions arguments becomes presented less. by It thoseis especially offering counter- for this reasonmessaging. that the Although changing not supportedof friends by is the a majority,relatively 54.54% common of criminal occurrence justice respondentsin the radicalisation supported process.this perception. Criminal justice respondents were largely under the impression that the ideals of the illegal organisation (59.09%) were more convincing, a perception that was not completely supported by the other two samples. Therefore, instead of relying on counter-messaging through the media, peer pressure is far more convincing. What makes counter-messaging through peers particularly55 challenging relates to the basic principle of socialisation as people (irrespective of age) associate with others who share similar opinions. Consequently, as a result of groupthink that often emerges, the possibility of a different perspective being presented to the rest of the group and others being open to this different opinion will become ‘controversial’ in the counter opinion. In other words, since friends share a particular opinion (for example more extreme views), the possibility of one or two people being accommodated with moderate opinions becomes less. It is especially for this reason that the changing of friends is a relatively common occurrence in the radicalisation process.

58 59

FCA_KenyaAnalysis.indd 58 24.1.2020 12.07

Figure 64: Reasons why counter initiatives may not be successful Figure 64: Reasons why counter initiatives may not be successful

Distrust those who presented these initiatives 28.42 25 54.02

Distrust those who presented these initiatives 28.42 25 54.02 Friends were more convincing 61.46 49.04 54.54

Friends were more convincing 61.46 49.04 54.54 The group was more successful in providing answers 37.37 33.01 51.72

The group was more successful in providing answers 37.37 33.01 51.72 The ideals of the group were more attractive 40.81 31.43 59.09

The ideals of the group were more attractive 40.81 31.43 59.09 People did not agree with their opinions 31.31 23.3 40.45

People did not agree with their opinions0 31.3120 4023.3 60 40.4580 100 120 140 160 180

Family (n=102) 0 Detained20 (n=107)40 Criminal60 Justice80 (n=89)100 120 140 160 180

Family (n=102) Detained (n=107) Criminal Justice (n=89) Figure 63. Reasons why counter initiatives may not be successful Figure 65: Aware of the Kenyan Government's Strategy against violent extremism

Figure 65: Aware100% of the Kenyan Government's Strategy against violent extremism 12.26 90% 100%80% 38.39 49.22 12.26 70%90% 60%80% 38.39 49.22 50%70% 87.74 40%60% 30%50% 61.61 50.78 87.74 20%40% 10%30% 61.61 50.78 20%0% Family (n-112) Detained (n=128) Criminal Justice (n=106) 10% 0% Yes No Family (n-112) Detained (n=128) Criminal Justice (n=106)

The Kenyan government informed the majorityYes of criminalNo justice respondents of the strategy, whereas the majority of the community-based samples were informed through radio and Figure 64. Aware of the Kenyan Government’s Strategy against violent extremism Thetelevision. Kenyan government informed the majority of criminal justice respondents of the strategy,

whereasFigure 66: Informed the majority of Kenya's counterof the strategy community -based samples were informed through radio and television.

Figure 66: Informed of Kenya's counter strategy The Kenyan government informed the majority of criminal justice respondents of the strategy, whereas the majority of the community-based samples were informed through radio and television.

In addition to being informed of the strategy, the question should rather be whether respondents were informed of its contents. While the majority of criminal justice respondents were aware of the contents of the strategy, the majority of the remaining two samples were not.

58 5956

FCA_KenyaAnalysis.indd 59 24.1.2020 12.07

NGO 4 2 6

Social media 9 5 5

Radio NGO 4 202 6 5 21

TelevisionSocial media 922 5 5 13 19

Internet Radio3 9 206 5 21

NewspapersTelevision 12 22 13 1413 19

Religious leaderInternet7 3 5 9 7 6

SchoolNewspapers1 2 2 12 13 14

GovernmentReligious leader 12 7 5 7 43 12

FamilySchool4 11 2 2

FriendsGovernment6 3 127 43 12

Family0 4 110 20 30 40 50 60 70 80

Family (n=72) Criminal Justice (n=92) Detained (n=62) Friends 6 3 7

In addition to being0 informed10 of the strategy,20 the30 question40 should rather50 be whether60 respondent70 s 80 were informed of its contents.Family While (n=72) the majorityCriminal of Justice criminal (n=92) justiceDetained respondents (n=62) were aware of the contents of the strategy, the majority of the remaining two samples were not. Figure 65. Aware of the Kenyan Government’s Strategy against violent extremism Figure In67 addition: Aware of theto contentbeing of informed government's of strategy the strategy, the question should rather be whether respondents were informed of its contents. While the majority of criminal justice respondents were aware of the contents100% of the strategy, the majority of the remaining two samples were not. 90% 21.65 Figure 67: Aware80% of the content of government's strategy 70% 59.41 100% 66.36 60% 90% 50% 21.65 80% 40% 78.35 70% 59.41 30% 66.36 60% 20% 40.59 50% 33.64 10% 40% 0% 78.35 30% Family (n=101) Detained (n=110) Criminal Justice (n=97) 20% 40.59 Yes No 33.64 10% Figure 66. Aware of the0% content of government’s strategy When respondents were requestedFamily (n=101) to rate the importanceDetained (n=110) of counterCriminal measures Justice (n=97), criminal justice respondents held a high regard for the majority ofYes possibleNo counter strategies (see Table 7). When respondents were requested to rate the importance of counter measures, criminal justice re-

Tablespondents 7: Priority held of counter a high strategies regard for the majority of possible counter strategies (see Table 7). When respondents were requested to rate the importance of counter measures, criminal justice Importancerespondents Family held (n=114)a high regard for Detainedthe majority (n=121) of possible counterCriminal strategies Justice (see (n=104) Table 7). Asking respondents to identify who should take the lead, both the detained and family samples identi- fiedTable the 7family: Priority and of counter community strategies leaders to take the lead, and that government and security agencies should stay away. The family and friends sample also expressed more trust in international donors 57 than Importancewhat was expressed Family (n=114)by criminal justice Detainedactors (see (n=121) Table 8). Criminal Justice (n=104)

Respondents were also asked to prioritise on what government should spend more money and (human) resources. Although data covered in Table 9 are being presented throughout this report, this table57 presents priorities in relation to others. Some proposed changes have serious human rights implications and respondents were asked to determine the acceptance (70-100%) of these negative measures.

60 61

FCA_KenyaAnalysis.indd 60 24.1.2020 12.07 Table 7. Priority of counter strategies

Importance Detained (n=117) Family (n=105) Criminal Justice (n=103) 1 Community policing (68.22%) Research into violent extremism Promote patriotism in Kenya (89.42%) (66.08%) 2 Building trust between security Building trust between security Community policing (89.42%) forces and public (66.97%) forces and public (58.62%) 3 Government support to local Government support to local More research on radicalisation communities (63.97%) communities (53%) (87.38%) 4 Research into violent extremism Community policing (50.86%) Economic development (85.58%) (61.68%) 5 Promote patriotism in Kenya Promote patriotism in Kenya Countering extremist ideologies (59.63%) (50%) (81.55%) 6 Government-led early warning Countering extremist ideologies Rehabilitation and reintegration (57.27%) (48.3%) (78.21%) 7 Counter extremist ideologies Prosecute radicalisers (46.96%) Train religious leaders to identify those (51.37%) at risk (73.08%) 8 Community-led early warning Rehabilitation and reintegration Train teachers to identify those at risk (49.08%) (45.69%) (67.31%) 9 Law enforcement against Law enforcement against radi- Harsh punishment (63.1%) radicalisation (49.07%) calisation (42.74%) 10 Rehabilitation and reintegration Government-led early warning Focus on families of radicalised (47.75%) (39.82%) (61.16%) 11 Prosecute radicalisers (44.55%) Community-led early warning Hold communities responsible (39.82%) (24.28%)

Table 8. Who should take the lead implementing the strategy

Priority Families (n=119) Detained (n=129) Criminal Justice (n=85) 1 Families (81.74%) Families (67.2%) Local religious organisations (85.85%) 2 Community leaders (79.82%) Community leaders (66.67%) Local youth organisations (83.81%) 3 Media (75.86%) Local religious organisations (66.4%) Local women organisations (78.09%) 4 Local religious organisations (71.79%) Media (59.02%) Intelligence (70.76%) 5 International donors (63.47%) Local youth organisations (58.07%) Department of Education (64.76%) 6 Local women organisations (61.21%) Local women organisations (50.83%) Police (58.49%) 7 Local youth organisations (60.71%) Department of Education (47.16%) Military (47.17%) 8 Department of Education (55.46%) International donors (40.67%) International donors (43.81%) 9 Police (49.57%) NCTC (38.02%) 10 Intelligence (47.83%) Intelligence (36.98%) 11 NCTC (44.74%) Police (31.14%) 12 Military (37.17%) Military (26.83%) 13 Political leaders (34.48%) Political leaders (23.57%)

As will be presented later in the report, respondents recognised the need to better train and equip, especially for the police to develop an improved relationship with the public and successfully fulfil its mandated tasks. It is equally necessary to recognise the importance (percentages) respondents as part of the criminal justice sample placed on changes and needs associated with security agencies. It is however particularly interesting to note the level of support respondents gave to very intrusive counter measures, most notably the monitoring of religious institutions (which can be interpreted as an infringement on religious freedom and freedom of expression) and the monitoring of all telephone calls. Both counter measures have serious implications on privacy and currently require legal permission (Section 36 of the Prevention of Terrorism Act). Monitoring of the internet and social media is seen as a priority when read with Table 6 in which all three samples identified social media and the internet as radicalisation facilitators.

60 61

FCA_KenyaAnalysis.indd 61 24.1.2020 12.07 Table 9. Priority of proposed changes to enhance the effectiveness of counter measures

Priority Families (n=111) Detained (n=119) Criminal Justice (n=105) 1 Train police to better conduct Train police to better conduct Provide better equipment to police investigations (86.24%) investigations (79.31%) officers (89.52%) 2 Train police to serve and protect the Provide training on human rights Enhance intelligence capabilities community (85.32%) (74.36%) (89.32%) 3 Provide training on human rights Train police to serve and protect Train police to better conduct (83.34%) the community (73.27%) investigations (88.46%) 4 Provide better equipment to police Enhance border security Provide better equipment to officers (69.44%) (70.28%) military officers (87.5%) 5 Enhance intelligence capabilities Provide better equipment to Train police to serve and protect (68.22%) police officers (67.83%) the community (85.44%) 6 Enhance border security (66.66%) Improve conditions in police Enhance the training of police stations (65.22%) officers (83.81%) 7 Enhance the training of police Enhance intelligence capabilities Enhance border security (82.69%) officers (64.81%) (61.95%) 8 Improve conditions in police stations Enhance the training of police Provide training on human rights (62.39%) officers (61.4%) (74.76%) 9 Change the criteria for recruitment Monitor the movement of Enhance the training of military of police officers (55.96%) foreigners (53.98%) officers (73.08%) 10 Provide better equipment to military Improve conditions in prisons Increase the monitoring of the officers (55.14%) (53.16%) Internet (72.11%) 11 Improve conditions in prisons50 Change the criteria for Monitor the movement of (54.9%) recruitment of police officers foreigners (71.43%) (52.14%) 12 Increase the monitoring of all foreign Enhance counter messaging Increase the monitoring of all financial transactions (54.2%) (48.67%) foreign financial transactions (70.87%) 13 Enhance the training of military Increase the monitoring of the Increase the monitoring of social officers (53.34%) Internet (47.37%) media (69.23%) 14 Monitor the activities of religious Increase the monitoring of all Install more CCTV cameras institutions (53.33%) foreign financial transactions (67.96%) (45.61%) 15 Enhance counter messaging51 (50%) Provide better equipment to Monitor the activities of religious military officers (43.36%) institutions (51.96%) 16 Increase the monitoring of the Enhance the training of military Monitor the activities of NGOs Internet (49.06%) officers (42.11%) (48.54%) 17 Increase the monitoring of social Increase the monitoring of social Increase the monitoring of all media (49.06%) media (41.11%) telephone calls (45.63%) 18 Monitor the movement of foreigners Install more CCTV cameras Change the criteria for recruitment (48.6%) (36.52%) of police officers (44.23%) 19 Install more CCTV cameras Change the criteria for Change the criteria for (46.79%) recruitment of military officers recruitment of military officers (35.97%) (38.84%) 20 Change the criteria for recruitment Monitor the activities of NGOs Recruit police officers from the of military officers (45.19%) (35.97%) region (25.49%) 21 Recruit police officers from the Monitor the activities of religious Change police stations (20.19%) region (43.12%) institutions (33.62%) 22 Monitor the activities of NGOs Close refugee camps (29.2%) Close refugee camps (19.23%) (42.05%) 23 Increase the monitoring of all Recruit police officers from the telephone calls (41.13%) region (27.82%) 24 Close refugee camps (27.1%) Increase the monitoring of all telephone calls (26.79%)

50 Exclusive to primary and family and friends samples 51 Exclusive to primary and family and friends samples

62 63

FCA_KenyaAnalysis.indd 62 24.1.2020 12.07 8.6 Corruption

Corruption – perceived and actual – is one of the most prominent factors determining legitimacy of any institution. Consequently, the lack of trust, reduced legitimacy and lack of confidence in public institutions can both be a cause and an effect of corruption, leading to a vicious cycle of decline.

Assessing the prominence of corruption in the mind of respondents, all three samples were asked to rate the level of corruption across the different levels of government and critical institutions. This question was asked to establish perceived corruption and the following discussion will identify the five most perceived to be corrupt. According to respondents being part of the family and friends sample, employees of national government are the most corrupt (87.03%), followed by police officers (86.24%), traffic police (84.4%), local government (79.09%) and Home Affairs, with specific reference to official documentation (70.76%). To respondents as part of the detained sample, police officers were the most corrupt (87.39%), followed by national government (83.62%), traffic police officers (83.48%), and lastly Home Affairs (53.91%). Finally, criminal justice respondents perceived traffic police officers to be most corrupt (76.24%), followed by local government employees (76.23%), national government (64.7%), police officers (60.78%) and Home Affairs (49.5%).

After determining who respondents perceived to be corrupt, the two non-government samples were asked whether they ever had to pay a bribe, of which 58.87% of the detained sample (n=124) and 57.94% of family members and friends answered in the affirmative. Asking respondents to indicate if they had to pay a bribe to the already identified institutions while at the same time indicating the number of times they had to pay a bribe, an interesting trend emerged.

300

250

200 64.7 76.23 60.78 76.24

150 49.5 83.62 87.39 82.76 83.48 100 53.91

14.85 22.55 50 26.95 87.03 86.24 84.4 23 5.88 79.09 70.76 38.89 33.96 12.28 15.89 0 National Local Police officer Military Traffic police Education Home Affairs Religious government government officer officer Department (passport leader official / official / application) employee employee

Family (n=111) Detained (n=119) Criminal Justice (n=102)

AfterFigure determining 67. Perception who ofrespondents corruption per perceived level of government, to be corrupt, department the or two institution non-government samples were asked if they ever had to pay a bribe, of which 58.87% of the detained sample (n=124) and 57.94% of family members and friends answered in the affirmative. Starting with the 86 respondents who indicated that they had to pay a bribe, the majority of respndents (57.83% ) had to pay a bribe to a police officer and even more than five times. Despite being fifth on the list of being perceived to be corrupt, 44.57% had to bribe a Home Affairs official, 43.21% 62 had to bribe a national government official, while 25.92% had to bribe a local government63 official and only 24.7% had to pay a bribe to a traffic police officer (see Figure 69 for more information).

Figure 69: Actual corruption - family and friends sample

FCA_KenyaAnalysis.indd 63 1.23 24.1.2020 12.07 100% 3.61 3.7 5.06 2.47 3.7 8.64 4.94 7.41 1.27 1.23 90% 16.87 7.23 8.64 7.41 11.39 80% 17.28 9.88 15.66 70% 33.73 25.93 60% 25.3 50% 95.06 96.3 40% 82.28 74.07 75.31 30% 55.42 56.79 20% 42.17 10% 0% Police officer Home Affairs National Local Traffic police Education Military officer Religious Government Government officer Department leader

No 1-2 3-4 5>

62 After determining who respondents perceived to be corrupt, the two non-government samples were asked if they ever had to pay a bribe, of which 58.87% of the detained sample (n=124) and 57.94% of family members and friends answered in the affirmative. Starting with the 86 respondents who indicated that they had to pay a bribe, the majority of respndents (57.83% ) had to pay a bribe to a police officer and even more than five times. Despite being fifth on the list of being perceived to be corrupt, 44.57% had to bribe a Home Affairs official, 43.21% had to bribe a national government official, while 25.92% had to bribe a local government official and only 24.7% had to pay a bribe to a traffic police officer (see Figure 69 for more information).

Although the actors implicated by family members and friends are very similar to the detained sample, the numbers are different, especially with reference to the police. Of the 97 respondents who had to pay a bribe, the majority (64.9%) had to bribe a police officer of which 30.85% had to pay a bribe five times and more. In reinforcing the sentiment that religious and ethnic minorities are being further marginalised, ethnic Somali Kenyans complained in the aftermath of the Westgate attack regarding their treatment at the hand of the police: we suffered at home in Somalia at the hands of al-Shabaab, while in Kenya, the security forces became like another ‘al-Shabaab’ in the way they treated us. Police fully understand that most of us have nothing to do with al-Shabaab but we are the cash cow, like a bank ATM. Once you are arrested, money is your solution and your innocence is nothing.52

Similar to family and friends, despite being fifth on the list of being perceived to be corrupt 48.91% had to bribe a Home Affairs official and 42.7% had to bribe a national government official. However, according to the detained sample, 36.67% had to pay a bribe to a traffic police officer while 27.27% had to bribe a local government official (see Figure 70 for more information).

Although some actors – most notably traffic police officers – are being perceived to be more corrupt than they actually are, others were expected to do better, but expectations were misplaced with reference to, for example, Home Affairs. At the same time, it is important to also take note of the percentages of respondents who were not required to pay a bribe to officials representing the mentioned institutions.

After determining some of the realities surrounding corruption in Kenya, especially the overestimation of perceptions associated with corruption, it did not come as a surprise that government’s attention was called for when respondents were asked where government should divert more resources in an attempt to address security challenges. Both respondents part of family members and friends (n=110) and the detained sample (n=128) placed a call for resources in the security forces (specifically the police) on the top of the list with 85.18% of family and friends and 80.33% of the detained sample placing this need at ‘major’ and ‘severe’. Although criminal justice respondents placed resources addressing corruption in security forces at seventh place after the training of security personnel (95.19%), corruption in government (90.38%), nation building beyond ethnic differences (89.43%), securing Kenya’s borders (89.42%), internal security (88.35%) and corruption in local government (86.67%), addressing corruption in security forces still received 86.66%. Considering that family and friends perceived local government to be more corrupt, 85.05% placed assistance to address corruption at local government level second and assistance to address corruption in national government third (81.65%). After 80.33% of the detained sample who called for assistance to address corruption in security forces, 78.51% asked for assistance to address corruption at local government level and 76.23% to deal with corruption at national government levels.

52 ENCA. Harassment haunts Somalis in Kenya a year after Westgate attack. 19 September 2014. Available at https://www.enca.com/harassment-haunts-somalis-kenya-year-after-westgate-attack (accessed on 10 October 2014).

64 65

FCA_KenyaAnalysis.indd 64 24.1.2020 12.07

300

250

200 64.7 76.23 60.78 76.24

150 49.5 83.62 87.39 82.76 83.48 100 53.91

Family (n=111) Detained (n=119) Criminal Justice (n=102)

After determining who respondents perceived to be corrupt, the two non-government samples were asked if they ever had to pay a bribe, of which 58.87% of the detained sample (n=124) and 57.94% of family members and friends answered in the affirmative. Starting with the 86 respondents who indicated that they had to pay a bribe, the majority of respndents (57.83% ) had to pay a bribe to a police officer and even more than five times. Despite being fifth on the list of being perceived to be corrupt, 44.57% had to bribe a Home Affairs official, 43.21% had to bribe a national government official, while 25.92% had to bribe a local government official and only 24.7% had to pay a bribe to a traffic police officer (see Figure 69 for more information).

Figure 69: Actual corruption - family and friends sample

1.23 100% 3.61 3.7 5.06 2.47 3.7 8.64 4.94 7.41 1.27 1.23 90% 16.87 7.23 8.64 7.41 11.39 80% 17.28 9.88 15.66 70% 33.73 25.93 60% which 30.85%25.3 had to pay a bribe five times and more. In reinforcing the sentiment that religious and50% ethnic minorities are being further marginalised, ethnic Somali Kenyans95.06 complained96.3 in the 40% 82.28 aftermath of the Westgate attack regarding74.07 their treatment75.31 at the hand of the police: we suffered30% at home in Somalia55.42 at the56.79 hands of al-Shabaab, while in Kenya, the security forces became like20% another42.17 ‘al-Shabaab’ in the way they treated us. Police fully understand that most of us have nothing10% to do with al-Shabaab but we are the cash cow, like a bank ATM. Once you are arrested, money0% is your solution and your innocence is nothing.48 Police officer Home Affairs National Local Traffic police Education Military officer Religious Government Government officer Department leader Similar to family and friends, despite being fifth on the list of being perceived to be corrupt 48.91% had to bribe a Home Affairs officialNo 1-2 and3-4 42.7%5> had to bribe a national government official. However, according to the detained sample, 36.67% had to pay a bribe to a traffic

policeFigure officer 68. Actual while corruption 27.27% - family had and to friends bribe sample a local government official (see Figure 70 for more Altinformation).hough the actors implicated by family members and friends are very similar to the detained sample, the numbers are different, especially with reference to the police. Of the 97 respondentsFigure 70: Actual who corruption had - detainedto pay sample a bribe, the majority (64.9%) had to bribe a police officer of

1.11 100% 3.37 2.27 5.68 1.14 10.87 7.87 2.25 5.56 90% 17.78 7.95 3.26 8.99 11.24 30.85 62 80% 5.56 13.64

70% 34.78 25.84 13.33 7.45 60%

50% 26.6 93.33 96.59 40% 83.15 72.73 63.33 30% 57.3 51.09 20% 35.11 10%

0% Police officer Home Affairs National Traffic police Local Education Religious Military officer Government officer Government Department leader

No 1-2 3-4 5>

Figure 69. Actual corruption - detained sample Although some actors – most notably traffic police officers – are being perceived to be more corrupt than they actually are, others were expected to do better, but expectations were misplaced with reference to, for example, Home Affairs. At the same time, it is important to also take note of the percentages of respondents who were not required to pay a bribe to officials representing the mentioned institutions.

After determining some of the realities surrounding corruption in Kenya, especially the overestimation of perceptions associated with corruption, it did not come as a surprise that government’s attention was called for when respondents were asked where government should divert more resources in an attempt to address security challenges. Both respondents part of family members and friends (n=110) and the detained sample (n=128) placed a call for resources in the security forces (specifically the police) on the top of the list with 85.18% of family and friends and 80.33% of the detained sample placing this need at ‘major’ and ‘severe’. Although criminal justice respondents placed resources addressing corruption in security forces at seventh place after the training of security personnel (95.19%), corruption in government (90.38%), nation building beyond ethnic differences (89.43%), securing Kenya’s 64 65 borders (89.42%), internal security (88.35%) and corruption in local government (86.67%), addressing corruption in security forces still received 86.66%. Considering that family and friends perceived local government to be more corrupt, 85.05% placed assistance to address

FCA_KenyaAnalysis.indd 65 63 24.1.2020 12.07 9. Interaction with security forces

Before assessing the police’s legitimacy and perceptions related to the police and security forces, it is important to frame the type and extent (number of times) of respondents’ interaction with the police. Both samples recalled being stopped and searched (39.55% amongst detained sample and 38.2% of family members and friends) and asked for identity documents (46.67% family members and friends and 43.94% amongst detained sample) more than five times. Both samples also recalled other intrusive experiences: amongst family members and friends, 46 respondents (73.91%) were detained at least once; of 61 respondents, close to 64% were arrested at least once and 24.59% were arrested twice; of 68 respondents who answered the question, 63.24% were called at least once by the police followed by 11.76% for four and five times. Of 75 respondents, 60% indicated that their place of residence was searched at least once while 21.33% indicated that their homes were searched twice and of 58 respondents (close to 60% of respondents) indicated that their vehicles were searched, followed by 17.24% who recalled five occasions and 13.79% referred to their vehicles being searched twice.

Of the 127 respondents who went through the criminal justice system (detained sample), 59% were arrested, 32% were questioned, 8% indicated that a family member was arrested, while 1% indicated that he was in a rehabilitation centre and the other one percent indicated that he received amnesty. Of the 103 respondents that were part of the family and friends sample, 44% identified being a family member of an arrested person, 23% were questioned, 19% were arrested themselves, 9% indicated that they were a family member of a person who had been killed while 3% was related to a person who had disappeared and similar to the detained sample, 1% indicated that he was related to a person in a rehabilitation centre and the other one percent indicated that the family member received amnesty.

As presented in Figure 72, the majority of the detained sample (64.23%) as well as the person referred to by family members and friends (detained 2) sample (62%) were arrested for belonging to a terrorist organisation; 20% of the detained 2 sample were implicated in the execution of a terrorist attack and 10.27% of the detained sample were implicated for the same type of offence. Section 24 of the Prevention of Terrorism Act deals with membership of terrorist groups and states that: A person who is a member of or professes to be a member of a terrorist group commits an offence and is liable, on conviction, to imprisonment for a term not exceeding thirty years. Although easy to charge a person for being a member, it is traditionally a very difficult offence to prove in court.

Of the family member or friends sample (see Figure 74), 34.23% indicated that they were arrested after the arrest of a sibling (31%), partner (25%), a parent (22%) or another family member (14%) (see Figure 75). Although there might have been valid reasons for arrests, there is also a concern that these arrests were made for investigation purposes or to facilitate interviewing that has the risk of becoming an interrogation. If it was the case, this practice contradicts two basic elements when making an arrest, namely that there always has to be a lawful basis for arrest and detention (alleged offense) and that there must be a clear link between the suspect and the offence. Arrest should rather be the result of an investigation in which evidence support the need to make an arrest. In case a police officer has reasonable suspicion (meaning that officers have an objectively reasonable basis for suspecting criminal activity before detaining someone) a person can be stopped and searched. It is also advised that before conducting a search, officers must reasonably suspect that the person is, for example, armed and dangerous. Police officers can however ask a person to stop and answer questions without having reasonable suspicion. The importance of prior information and intelligence before making an arrest was also emphasised by a family member who was interviewed: In fighting violent extremism,

66 67

FCA_KenyaAnalysis.indd 66 24.1.2020 12.07

corruption at local government level second and assistance to address corruption in national government third (81.65%). After 80.33% of the detained sample who called for assistance to address corruption in security forces, 78.51% asked for assistance to address corruption at local government level and 76.23% to deal with corruption at national government levels. 9. Interaction with security forces

Figure 71: Number of times family members and friends interacted with the police

2.17

Detained (n=46) 73.91 13.04 6.52 4.35 1.64 Arrested (n=61) 63.93 24.59 4.92 4.92

Questioned (n=86) 47.67 16.28 18.6 2.33 15.12

House searched (n=75) 60 21.33 5.33 4 9.33

Vehicle searched (n=58) 56.9 13.79 5.17 6.9 17.24

Searched in person (n=70) 47.14 21.43 14.29 4.29 12.86

Receiving phone calls from police (n=68) 63.24 4.41 8.82 11.76 11.76

Asked for identity papers (n=90) 17.78 13.33 8.89 13.33 46.67

Stopped routine check (n=89) 25.84 16.85 12.36 6.74 38.2

Visited in custody (n=80) 27.5 10 12.5 17.5 32.5

Visited by a police officer (n=84) 47.62 16.67 9.52 5.95 20.24

Ask for assistance (n=94) 25.53 24.74 15.96 7.45 26.6

Enter police station (n=104) 43.27 14.42 6.73 5.77 29.81

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

1 2 3 4 5

Figure 70. Number of times family members and friends interacted with the police Figure 72: Number of times detained respondents interact with the police

Detained (n=85) 69 15 4 1 11 Arrested (n=107) 70 17 3 3 7 64 Questioned (n=112) 47 16 8 1 28

House searched (n=74) 57 24 7 12

Vehicle searched (n=61) 44 15 10 8 23

Searched in person (n=83) 39 17 14 4 26

Receiving phone calls from police (n=42) 38 17 7 2 36

Asked for identity papers (n=94) 11 13 9 5 62

Stopped routine check (n=100) 17 13 9 8 53

Visited in custody (n=82) 20 20 12 7 41

Visited by a police officer (n=59) 39 20 14 3 24

Ask for assistance (n=72) 39 20 15 8 18

Enter police station (n=70) 27 20 10 10 33

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

1 2 3 4 5

Figure 71. Number of times detained respondents interact with the police Of the 127 respondents who went through the criminal justice system (detained sample), 59% were arrested, 32% were questioned, 8% indicated that a family member was arrested, while 66 1% indicated that he was in a rehabilitation centre and the other one percent indicated that he 67 received amnesty. Of the 103 respondents that were part of the family and friends sample, 44% identified being a family member of an arrested person, 23% were questioned, 19% were arrested themselves, 9% indicated that they were a family member of a person who had been killed while 3% was related to a person who had disappeared and similar to the detained

FCA_KenyaAnalysis.inddsample, 67 1% indicated that he was related to a person in a rehabilitation centre and the other 24.1.2020 12.07 one percent indicated that the family member received amnesty.

As presented in Figure 72, the majority of the detained sample (64.23%) as well as the person referred to by family members and friends (detained 2) sample (62%) were arrested for belonging to a terrorist organisation; 20% of the detained 2 sample were implicated in the execution of a terrorist attack and 10.27% of the detained sample were implicated for the same type of offence. Section 24 of the Prevention of Terrorism Act deals with membership of terrorist groups and states that: A person who is a member of or professes to be a member of a terrorist group commits an offence and is liable, on conviction, to imprisonment for a term not exceeding thirty years. Although easy to charge a person for being a member, it is traditionally a very difficult offence to prove in court.

Figure 73: Offences for being arrested

Committing a terrorist act 10.57 20 Undocumented individuals 1 2.44 Illegal documentations 1.634 Incitement to violence 0.812 Recruiting others into an illegal organisation 5.696 Financing a terrorist organisation 2 8.13 Belonging to an illegal organisation 6264.23 Bomb-making materials 23.25 Weapons-related offence 6.57 Murder 1 4.88 Drug trafficking 3.25

Indecent assault 1 Burglary 1 Property damage 0.814 Minor assaults 5 6.5 Road traffic offences 1 2.44 0 10 20 30 40 50 60 70

Detained (n=132) Detained 2 (n=100)

Figure 72. Offences for being arrested Of the family member or friends sample (see Figure 74), 34.23% indicated that they were arrested after the arrest of a sibling (31%), partner (25%), a parent (22%) or another family member (14%) (see80 Figure 75). Although there might have been valid reasons for arrests, there 68.66 is also a concern 70that these arrests 65.77were made for investigation purposes or to facilitate interviewing that has the risk of becoming an interrogation. If it was the case, this practice contradicts two basic60 elements when making an arrest, namely that there always has to be a lawful basis for 50arrest and detention (alleged offense) and that there must be a clear link between the suspect40 and the 34.23offence. Arrest should rather be the result of an investigation in 31.34 which evidence support30 the need to make an arrest. In case a police officer has reasonable suspicion (meaning that officers have an objectively reasonable basis for suspecting criminal 20 activity before detaining someone) a person can be stopped and searched. It is also advised that before conducting10 a search, officers must reasonably suspect that the person is, for example, armed and0 dangerous. Police officers can however ask a person to stop and answer questions without having reasonableFamily (n=111) suspicion. The importanceDetained (n=134) of prior information and intelligence before making an arrest was alsoYes emphasisedNo by a family member who was interviewed: In fighting violent extremism, government should invest so much [more] i n intelligence Figure 73. Arrest followed by another arrest and Figureinvestigations 75: Relationship to toensure the person that arrested people first are not victimized on the basis of mere allegations but from well investigated evidence.

Figure 74: Arrest followed by another arrest government Familyshould invest14 so much [more]9 in intelligence and investigations to ensure that people are not victimizedBusiness partner on the5 basis of mere allegations but from well investigated evidence.

Considering theSon arrest1 of9 parents and other family members, ages at the time of arrest were higher for the family sample while the detained sample included younger individuals with the largest com- Grandparents 2 ponent (13.39%) being between 22 and 23 years of age. Friend 5 30

Partner 25 14

Sibling 31 30

68 Parents 22 3 69 0 10 20 30 40 50 60 70

Family (n=73) Detained (n=58) 66

FCA_KenyaAnalysis.inddConsidering 68 the arrest of parents and other family members, ages at the time of arrest were 24.1.2020 12.07 higher for the family sample while the detained sample included younger individuals with the largest component (13.39%) being between 22 and 23 years of age.

Figure 76: Age arrested for the first time

80 68.66 70 65.77

40 34.23 31.34 30

0 Family (n=111) Detained (n=134)

Yes No

Figure 75: Relationship to the person arrested first

Family 14 9

Business partner 5

Son 1 9

Grandparents 2

Friend 5 30

Partner 25 14

Sibling 31 30

Parents 22 3

0 10 20 30 40 50 60 70

Family (n=73) Detained (n=58)

Figure 74. Relationship to the person arrested first Considering the arrest of parents and other family members, ages at the time of arrest were higher for the family sample while the detained sample included younger individuals with the largest component (13.39%) being between 22 and 23 years of age.

FigureOlder 76: Agethan 50arrested years for the first time 2.36 15.04 45 - 49 years 1.77 4.72 40 - 44 years 3.94 14.16 38 - 39 years 2.65 36 - 37 years 2.65 8.66 34 - 35 years 5.516.19 32 - 33 years 5.31 6.3 30 - 31 years 4.72 15.04 28 - 29 years 7.09 7.96 26 - 27 years 2.65 12.6 24 - 25 years 4.42 10.24 22 - 23 years 7.08 13.39 20 - 21 years 5.31 8.66 18 - 19 years 6.19 7.09 16 - 17 years 2.653.15 14 - 15 years 0.881.57 0 2 4 6 8 10 12 14 16 Detained (n=127) Family (n-113) 67

Figure 75. Age arrested for the first time 9.1 Police Legitimacy

Police legitimacy measured in terms of the public’s acceptance to obey law enforcement is grounded on the police’s interactions with the public and the fairness of police procedures.49 Police legitimacy can thus be studied from the viewpoint of citizens, looking at their perceptions of fairness in policing and the impact on citizens’ willingness to cooperate with the police. Furthermore, from an organizational point of view, in how the police as an organization deal with the demands placed on them. This study in focusing on the public's perceptions about the lawfulness and legitimacy of law enforcement, provides important criteria for judging the police. 68 Lawfulness means that police comply with constitutional, statutory and professional69 norms, whereas legitimacy is linked to the public's belief about the police and the willingness of the public to recognize their authority. Trust can be described as the ‘belief that someone or something is reliable, good, honest and effective.’50 High levels of trust promote healthy interactions, whereas low levels of trust undermine the possibility of a constructive relationship.

FCA_KenyaAnalysis.inddTrust in 69 law enforcement is essential for the belief in the legitimacy of law enforcement or a 24.1.2020 12.07 feeling of obligation to obey the law and adhere to decisions made by legal authorities.

In measuring trust all three samples were asked to rate their trust in law enforcement in relation to other institutions (government and community) of which both detained sample (n=91) and family and friends (n=84) expressed the least trust in the police with 5.88% and 7.6% respectively. In contrast, criminal justice representatives (n=80) rated their trust in the police at 55.7%. Despite this positive reflection, trust in the police was lower than in the military (62.5%), intelligence agencies (62.5%) and prison authorities (61.25%).

On a practical level, respondents were asked if they would come forward with information regarding the suspected involvement of a friend and/or family member in an illegal organisation of which 51.61% of respondents representing family members and friends and 61.15% of respondents representing individuals who went through the criminal justice system answered in the affirmative. In a follow-up question, respondents representing both samples were asked to whom they would report it. As explained in Figure 77 both samples expressed most trust in community leaders followed by the police and religious authorities.

68 9.1 Police Legitimacy

Police legitimacy measured in terms of the public’s acceptance to obey law enforcement is grounded on the police’s interactions with the public and the fairness of police procedures.53 Police legitimacy can thus be studied from the viewpoint of citizens, looking at their perceptions of fairness in policing and the impact on citizens’ willingness to cooperate with the police. Furthermore, from an organizational point of view, is in how the police as an organization deal with the demands placed on them. This study focusing on the public’s perceptions about the lawfulness and legitimacy of law enforcement provides important criteria for judging the police. Lawfulness means that police comply with constitutional, statutory and professional norms, whereas legitimacy is linked to the public’s belief about the police and the willingness of the public to recognize their authority. Trust can be described as the ‘belief that someone or something is reliable, good, honest and effective.’54 High levels of trust promote healthy interactions, whereas low levels of trust undermine the possibility of a constructive relationship. Trust in law enforcement is essential for the belief in the legitimacy of law enforcement or a feeling of obligation to obey the law and adhere to decisions made by legal authorities.

On a practical level, respondents were asked if they would come forward with information regarding the suspected involvement of a friend and/or family member in an illegal organisation of which 51.61% of respondents representing family members and friends and 61.15% of respondents representing individuals who went through the criminal justice system answered in the affirmative. In a follow-up question, respondents representing both samples were asked to whom they would report it. As explained in Figure 77 both samples expressed most trust in community leaders followed by the police and religious authorities.

However, when respondents were asked to rate possible reasons for not coming forward with information, fear of being arrested and harassed by security forces rated very high within both samples. The majority (52.43%) of the family and friends sample expressed fear for their lives as well as that of their family. In light of this finding, it is important to note that people who believe that police are performing their duties with professionalism and integrity are more likely to obey laws and support the system by coming forward with information and act as witnesses.55 As one respondent part of the detained sample explained: “If you report to police, they will start questioning you on things you are not aware of. For instance, what type of gun did they have? Where are they from? Questions you cannot answer.” This concern was echoed by others: “When we offer information to the police, they will try to fix you as being part of the illegal activity you report to them.” “Being a victim of security agencies the reason behind was that before I was suspected to be member of terror group I have reported an incident that happened near my home village. Immediately thereafter they start investigating me and relating me to that terror incident which I have reported to them. Therefore, I fear if I do again same will happen.”

53 Jannie Noppe, Antoinette Verhage, Anjuli Van Damme, (2017) “Police legitimacy: an introduction” Policing: An International Journal, Vol. 40 Issue: 3, pp.474-479 54 Michael Friedman. What Happens When We Don’t Trust Law Enforcement? The Importance of Law Enforce- ment’s Role in Our Society’s Well-Being. Psychology Today, 9 September 2014. Available at https://www. psychologytoday.com/us/blog/brick-brick/201409/what-happens-when-we-dont-trust-law-enforcement-0 (accessed on 28 December 2018). 55 Horowitz, Jake. “Making every encounter count: Building trust and confidence in the police.” National Insti- tute of Justice Journal 256, no. 1 (2007): 10

70 71

FCA_KenyaAnalysis.indd 70 24.1.2020 12.07

Figure 77: Who will be trusted with this information

Representatives of another country 2

NGO 17.5 18

Humanitarian assistance organisation 5 10

Human Rights Commission 27.5 14

Other government agency in your country 2.5 9

Police 56.25 44

Teacher 2.5 9

Religious authority 30 41

Community leader 72.5 59

Politician 20 10

0 20 40 60 80 100 120 140

Family (n=80) Detained (n=100)

Figure 76. Who will be trusted with this information However, when respondents were asked to rate possible reasons for not coming forward with information, fear of being arrested and harassment by security forces rated very high within both samples. The majority (52.43%) of the family and friends sample expressed fear for their livesIn as another well as example that of police their officers family. went In lightfurther of in this treating finding the personit is important coming forward to note as that a suspect. people who believeAlthough that notthe a police common are occurrence, performing a mother their asduties part ofwith the professionalismfamily sample in her and explanation integrity why are more likelyshe to would obey not laws come and forward, support provided the thesystem following by comingdisturbing forward testimony: with I volunteered information to tell and the act as police that my son joined al-Shabaab but what we went through in the hands of police and ATPU only witnesses.51 As one respondent part of the detained sample explained: “If you report to police, they Godwill knows.start questioning We were all youvictimised. on things Several you male are police not aware [officers] of. cameFor instance, [to] search what my house type andof gun did told my girl to remove her clothes and remain naked in order to be searched and [they] later raped they her.have? Recognising Where thisare serious they from?allegation Questions and the factyou thatcannot it was answer. not supported” This concernby legal action,was echoed the by otherreaders: “When is reminded we offer that information similar to other to thetestimonies, police, theyit is reflective will try toof fixhow you respondents as being experienced part of the illegal activityand youperceived report the to actions them.” of “Being police aofficers. victim of security agencies the reason behind was that before I was suspected to be member of terror group I have reported an incident that happened near my homeIn village.the absence Immediately of trust in thethereaf police,ter the they latter start would investigating have to resort me to otherand relatingtactics – forme example to that terror incidentarresting which witnesses I have reportedor even worse, to them. abuse Therefore, – to initiate I interactionfear if I do with again the public.same willAs one happen.” detained respondent explains: “I was arrested together with my elder brother, in as much as (he) was not the In anothertarget, whatexample I witnessed police my officersbrother going went through further was in very treating traumatizing. the person Firstly, thecoming reason forward of our as a arrest was not explained and secondly, my brother was slapped at the back of his neck almost going suspect.unconscious, Although assaulted not a like common an animal. occurrence, The beating abecame mother severe as wheneverpart of hethe tried family to ask sample why we in her explanationwere arrested. why Thatshe timewould we werenot come all handcuffed. forward, Eventually provided we thewere following released without disturbing even goingtestimony: I volunteeredthrough theto tellcourt, the I feelpolice bitter that and my ever son since joined I learnt al-Shabaab not to respect but whator rely we on wentany police through for help.” in the hands of policeAnother and testimony: ATPU only“I am Godfed up knows. with police, We I weredon’t trustall victimised. them because Several of how mtheyale treated police me [officers] during came [to] [a]search raid. Theymy house brutally and punished told me,my almostgirl to killing remove me and her imagine clothes we and were remain just suspects, naked what in orderif we to be searchedwere trueand criminals. [they] later Even rapedthe guilty her. ones Recognising are not supposed this serious to be treated allegation this way.” and the fact that it was not supported by legal action, the reader is reminded that similar to other testimonies, it is reflectiveThe unfortunate of how respondents circumstance eofxperienced these practices and is perceived the fact that the it willactions enforce of policeexisting officers.negative perceptions of the public. Especially ‘targeted’ communities will foster a negative attitude towards the police leading to a cycle of not wanting to assist law enforcement. In the worst cases, these In theperceptions absence ofwill trust lead into thea perception police, the that latter the police would become have the to ‘enemy’resort toor ‘them’,other tacticsbut on the – for other example arrestingside in witnesses the mind of or law even enforcement worse, abuse officials – the to public,initiate especially interaction these with communities the public. will Asequally one bedetained respondentcategorized explain as ‘them’s: “I and was by arrested association together the ‘enemy’. with my elder brother, in as much as (he) was not the target, what I witnessed my brother going through was very traumatizing. Firstly, the reason of our arrest was not explained and secondly, my brother was slapped at the back of his neck almost going unconscious, assaulted like an animal. The beating became severe whenever he tried to ask 70 71

FCA_KenyaAnalysis.indd 71 24.1.2020 12.07

why we were arrested. That time we were all handcuffed. Eventually we were released without even going through the court, I feel bitter and ever since I learnt not to respect or rely on any police for help.” Another testimony: “I am fed up with police, I don't trust them because of how they treated me during [a] raid. They brutally punished me, almost killing me and imagine we were just suspects, what if we were true criminals. Even the guilty ones are not supposed to be treated this way.”

The unfortunate circumstance of these practices is the fact that it will enforce existing negative perceptions of the public. Especially ‘targeted’ communities will foster a negative attitude towards the police leading to a cycle of not wanting to assist law enforcement. In the worst cases these perceptions will lead to a perception that the police become the ‘enemy’ or ‘them’, but on the other side in the mind of law enforcement officials the public, especially these communities will equally be categorized as ‘them’ and by association the ‘enemy’.

Figure 78: Prevent family members and friends to come forward with information

Fear for broader ethnic community harassed by security forces 22.12 18.27 24.04 13.46 22.12

Fear for broader religious community harassed by security 14.71 19.61 19.61 19.61 26.47 forces

Fear of no action being taken 19.42 8.74 20.39 16.5 34.95

Fear of being arrested 13.08 6.54 14.02 20.56 45.79

Fear of being victimised by the community 10.68 8.74 11.65 21.36 47.57

Fear of investigations 10.583.85 15.38 32.69 37.5

Did not trust the security forces with information 6.8 6.8 15.53 26.21 44.66

Fear for life or broader family 8.745.83 11.65 21.36 52.43

Fear family being harassed by security forces 8.654.819.62 29.81 47.12

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

1-20% 30-40% 50-60 70-80% 90-100%

Figure 77. Prevent family members and friends to come forward with information Both samples expressed relative limited fear of their religious and ethnic communities being harassed by security forces.

Figure 79: Prevent respondents previously detained from coming forward with information Fear for broader ethnic community harassed by security forces 30.63 11.71 14.41 13.51 17.12 12.61

Fear for broader religious community harassed by security 24.32 13.51 14.41 13.51 21.62 12.61 forces

Fear of no action being taken 21.3 11.11 12.96 15.74 14.81 24.07

Fear of being arrested 20.66 10.74 8.26 14.05 18.18 28.1

Fear of being victimised by the community 16.36 4.558.18 13.64 20.91 36.36

Fear of investigations 15.18 5.368.04 13.39 26.79 31.25

4.31 Did not trust the security forces with information 9.48 7.76 18.97 25 34.48 3… Fear for life or broader family 10.17 5.08 15.25 27.12 38.98 70 2… Fear family being harassed by security forces 16.07 2.6811.61 29.46 37.5

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60 70-80% 90-100%

Figure 78. Prevent respondents previously detained from coming forward with information 9.1.1 Personal interactions Personal interactions have the strongest impact on perceptions as explained throughout this section. People form opinions of the police based on their own interactions with them or those of close family and friends or from experiences with the broader community, especially if the community is not diverse.52 In other words, when people form opinions of the police based on their interactions, they tend to focus on the process more than the outcome. Impressions of police encounters72 are influenced by the demeanour as well as the actions of the officer. People pay 73 close attention to the “neutrality of decision making, respectful and polite interpersonal treatment, and opportunities for input into decisions.”53 For example as one detained respondent explained after coming forward with information: “During interrogation [that should have been an interview], police ask questions with a harsh tone.”

FCA_KenyaAnalysis.indd Considering 72 the fact that people form opinions of the police based on their interactions, in a 24.1.2020 12.07 series of questions assessing respondents’ interaction with security forces, respondents were asked if they felt that the treatment received from the police was ‘fair’. Respondents of both samples were of the opinion that they did not receive fair treatment from security forces.

Figure 80: Were you treated fairly by security forces

71 9.1.1 Personal interactions Personal interactions have the strongest impact on perceptions as explained throughout this section. People form opinions of the police based on their own interactions with them or those of close family and friends or from experiences with the broader community, especially if the community is not diverse.56 In other words, when people form opinions of the police based on their interactions, they tend to focus on the process more than the outcome. Impressions of police encounters are influenced by the demeanour as well as the actions of the officer. People pay close attention to the “neutrality of decision making, respectful and polite interpersonal treatment, and opportunities for input into decisions.”57 For example as one detained respondent explained after coming forward with information: “During interrogation [that should have been an interview], police ask questions with a harsh tone.”

Considering the fact that people form opinions of the police based on their interactions, in a series of questions assessing respondents’ interaction with security forces, respondents were asked if they felt that the treatment received from the police was ‘fair’. Respondents of both samples were of the opinion that they did not receive fair treatment from security forces.

Research also shows that an officer’s demeanour58 and actions are crucial to perceptions of police legitimacy. Should officers communicate well, listen and treat citizens with respect, citizens would respond in kind.59 This view was also echoed by Weitzer (2005), but unfortunately not experienced by respondents in vulnerable communities.60

100% 90% 80% 70% 73.33 60% 87.05 50% 40% 30% 20% 26.67 10% 12.95 0% Detained (n=139) Family (n=120)

Yes No

Figure 79. Were you treated fairly by security forces Research also shows that an officer's demeanour54 and actions are crucial to perceptions of police legitimacy. Should officers communicate well, listen and treat citizens with respect, citizens would respond56 Schuck, in kind. Amie M.,55 andThis Dennis view P. wasRosenbaum. also “Globalechoed and by neighborhood Weitzer attitudes (2005) toward, but the unfortunately police: Dif- not experienced ferentiationby respondents by race, ethnicity in vulnerable and type of communitiescontact.” Journal.56 of quantitative criminology 21, no. 4 (2005): 391-418. 57 Horowitz, Jake. “Making every encounter count: Building trust and confidence in the police.” National Insti- Whereas memberstute of Justice of theJournal community 256, no. 1 (2007): preferred 8-11. to call on the assistance of family when in danger, criminal58 Rosenbaum, justice Dennis actor P., sAmie express M. Schuck, a Sandraparticular K. Costello, high Darnell preference F. Hawkins, toand call Marianne the K.police Ring. “At when- in danger. Workingtitudes towardfor andthe police: being The effectsin contact of direct andwith vicarious police experience.” officers Police as quarterlycriminal 8, no.justice 3 (2005): actors , presented a 343-365.positive image amongst criminal justice respondents that increases trust in the 59 Horowitz, Jake. “Making every encounter count: Building trust and confidence in the police.” National Insti- institution. In tuteturn of it Justice enhances Journal 256,the no. expectation 1 (2007): 8-11. of receiving assistance. Referred to as a police subculture,60 every Weitzer, police Ronald, agencyand Steven around A. Tuch. “Determinants the world of demonst public satisfactionrates awith subculture the police.” Policethat quarterlyconsists of a set of beliefs,8, attitudesno. 3 (2005): and 279-297. behaviour that bind members of law enforcement. Working under 72 the same circumstances, being confronted with similar challenges and threats police officers73 tend to view members of the public as untrustworthy and potentially hostile. At the same time fellow officers look at their colleagues (but also in some cases to members of the broader criminal justice system) for support and unity, forming a ‘brotherhood’. As a result, ‘us’ (the police) versus ‘them’ (the public) is being created. Although there is a positive in the form of comradeship being FCA_KenyaAnalysis.inddformed between 73 officers, there is also a negative side to a police subculture in that it can lead 24.1.2020 12.07 to some values and behaviour that deviate from what can be considered as acceptable in the broader society.57

Figure 81: Call the police when in danger

Detained (n=121) 16.95 11.86 13.56 25.42 19.49 12.71

Family (n=109) 19.27 8.26 18.35 17.43 12.84 23.85

0.99 Criminal Justice (n=101) 3.96 7.92 42.57 42.57 1.98 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

100% 90% 80% 70% 73.33 60% 87.05 50% 40% Whereas members of30% the community preferred to call on the assistance of family when in danger, criminal justice actors express20% a particular high preference to call the police when in danger. Working for and being in contact with police officers as criminal justice actors,26.67 presented a positive image amongst 10% criminal justice respondents that increases12.95 trust in the institution. In turn it enhances the expectation of 0% receiving assistance. Referred toDetained as a (n=139)police subculture, everyFamily police(n=120) agency around the world demonstrates a subculture that consists of a set of beliefs, attitudes and behaviour that bind members of law enforcement. Working under the same circumstances,Yes No being confronted with similar challenges and

threats police officers tend to view members of the public as untrustworthy and potentially hostile. At the same time fellow officers look at their colleagues (but also in some cases to members of the broader Research also shows that an officer's demeanour54 and actions are crucial to perceptions of policecriminal legitimacy. justice system) Should for officers support communicate and unity, forming well, listena ‘brotherhood’. and treat Ascitizens a result, with ‘us’ respect, (the police) citizens versus ‘them’ (the public) is being created. Although there is a positive in the form of comradeship being would respond in kind.55 This view was also echoed by Weitzer (2005), but unfortunately not formed between officers, there is also a negative side to a police subculture in that it can lead to some 56 experienced by respondents in vulnerable communities. 61 values and behaviour that deviate from what can be considered as acceptable in the broader society. Whereas members of the community preferred to call on the assistance of family when in danger,When criminalasking respondents justice actor to rates express their experiences a particular with high potential preference actors theyto call might the have police come when across in danger.during Workingtheir experiences for and with being the criminal in contact justice withsystem, police both samplesofficers rated as criminal the police justice as ‘poor’. actors Most, presentednotably respondentsa positive imagepart of amongstthe detained criminal sample justice rated respondents the police (80%), that followedincreases by trust intelligence in the institutionagencies. In (37.5%)turn it andenhances prosecutors the expectation (33.33%) as of‘poor’. receiving assistance. Referred to as a police subculture, every police agency around the world demonstrates a subculture that consists of a set Familyof beliefs, and friendsattitudes rated and the behaviour police (57.14%) that bind followed members by intelligence of law enforcement. agencies (34.69%) Working and underprison the officialssame circumstances, (22.77%) as ‘poor’. being confronted with similar challenges and threats police officers tend to view members of the public as untrustworthy and potentially hostile. At the same time fellow officersConsidering look at the their circumstances colleagues around (but also above in interaction,some cases it tois tomembers be expected of the that broader respondents criminal may harbour ill feelings towards those part of the judicial process. However, to what extent criminal justice justice system) for support and unity, forming a ‘brotherhood’. As a result, ‘us’ (the police) versus actors played a part in the radicalisation of people will be assessed below. ‘them’ (the public) is being created. Although there is a positive in the form of comradeship being formed between officers, there is also a negative side to a police subculture in that it can lead to some values and behaviour that deviate from what can be considered as acceptable in the broader society.57

Figure 81: Call the police when in danger

Detained (n=121) 16.95 11.86 13.56 25.42 19.49 12.71

Family (n=109) 19.27 8.26 18.35 17.43 12.84 23.85

0.99 Criminal Justice (n=101) 3.96 7.92 42.57 42.57 1.98 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 80. Call the police when in danger

61 Paoline III, Eugene A. “Shedding light on police culture: An examination of officers’ occupational attitudes.” Police quarterly 7, no. 2 (2004): 205-236.

74 75

FCA_KenyaAnalysis.indd 74 24.1.2020 12.07

When asking respondents to rate their experiences with potential actors they might have come across during their experiences with the criminal justice system, both samples rated the police as ‘poor’. Most notably respondents part of the detained sample rated the police (80%), followed Whenby intelligence asking respondents agencies (3 to7. 5%)rate andtheir prosecutors experiences (33.33%) with potential as ‘poor’. actors they might have come across during their experiences with the criminal justice system, both samples rated the police as ‘poor’.Figure 82 Most: Personal notably experience respondents of individuals partpreviously of thedetained detained sample rated the police (80%), followed by intelligence agencies (37.5%) and prosecutors (33.33%) as ‘poor’. 2.68 FigureForeign 82 humanitarian: Personal experience agencies of individuals previously detained 91.96 4.46 0.88 Foreign intelligence 88.5 5.31 4.42 0.88 2.68 Foreign humanitarianCommunity membersagencies 41.38 5.1791.96 18.97 25 9.484.46 0.88 ForeignReligious intelligence leaders 45.76 88.55.93 18.64 22.88 5.31 4.426.78 0.88 HumanCommunity Rights organisations members 41.3850 5.17 3.3918.97 21.19 25 20.34 9.485.08 ReligiousPrison leaders staff 45.7652.99 5.93 18.6418.8 22.8818.8 8.556.78 0.85 Human RightsPrison organisations Authorities 51.7250 3.39 20.6921.19 19.8320.34 7.765.08 PrisonCourts staff 47.4652.99 22.03 18.8 22.8818.8 7.968.555.31 0.85 PrisonDefence Authorities council 51.7261.06 20.6914.16 11.519.83 7.96 7.765.31 ProsecutorsCourts 47.4652.99 22.03 33.33 22.88 7.9611.11 5.311.710.85 IntelligenceDefence agency council 35.83 61.06 37.5 14.16 11.523.33 7.96 5.311.67 1.67 ProsecutorsMilitary 52.99 80 33.33 14.1711.11 1.715 0.850.83 IntelligenceFigure 81 .agency CallPolice the2.4 police when35.83 in danger 80 37.5 23.33 16 1.670.81.670.8 Military 0% 10% 20% 30% 40%80 50% 60% 70% 80% 14.1790% 5100%0.83 Police 2.4 No Poor Relative 80Good Excellent 16 0.8 0.8

0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Family and friends rated the policeNo (57.14%)Poor Relative followedGood by Excellentintelligence agencies (34.69%) and prison officials (22.77%) as ‘poor’. Figure 81. Personal experience of individuals previously detained FamilyFigure 83 :and Personal friends experience rated of familythe police members (57.14%) and friends followed by intelligence agencies (34.69%) and prison officials (22.77%) as ‘poor’.

FigureForeign 83 humanitarian: Personal experience agencies of family members and friends 80.85 8.51 7.45 2.13 1.06

Foreign intelligence 79.17 10.42 7.29 3.12

Foreign humanitarianCommunity membersagencies 38.3 80.858.51 21.28 23.4 8.51 7.458.512.13 1.06

ForeignReligious intelligence leaders 34.38 6.2579.17 23.96 28.1210.42 7.297.293.12

HumanCommunity Rights organisations members 38.341.24 8.514.12 20.6221.28 24.7423.4 9.288.51 2.04 ReligiousPrison leaders staff 34.38 67.356.25 23.96 19.3928.12 10.27.29 1.02 0.99 Human RightsPrison organisations Authorities 41.24 69.31 4.12 20.62 24.7422.77 5.949.28 0.99 2.04 PrisonCourts staff 65.3567.35 17.8219.39 10.8910.2 4.95 1.020.99 0.99 PrisonDefence Authorities council 69.3169.39 13.2722.77 11.22 5.945.1 1.02 0.99 ProsecutorsCourts 65.3569.39 17.8217.35 10.8911.224.952.040.99 2.04 IntelligenceDefence agency council 46.94 69.39 34.69 13.27 11.2211.22 5.15.1 1.02 1.01 ProsecutorsMilitary 69.3977.78 17.35 17.17 11.223.032.041.01

Intelligence agencyPolice 12.24 46.94 57.14 34.69 22.45 11.22 8.165.1 2.04 1.01 Military 0% 10% 20% 30% 77.7840% 50% 60% 70% 80% 17.1790% 3.03100%1.01

Police 12.24 No Poor Relative57.14Good Excellent 22.45 8.16 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% Figure 82. Personal experience of family members and friends No Poor Relative Good Excellent

74 75

FCA_KenyaAnalysis.indd 75 24.1.2020 12.07 9.1.2 Procedural justice

Procedural justice is the notion that the judicial process is fair, free of bias and that people are represented equally (treated politely and respectfully), have the opportunity to be heard and are judged by a neutral system.62

The following elements will influence police legitimacy:

• Ability to conduct investigations and other basic police functions

• Abuse of force

• Profiling based on ethnicity and/or religion

• Accountability of managers and supervisors

• Equal treatment of all members of the public

• Citizen accessibility to the police

• Corruption – as presented in a separate section above, corruption in the police is cause for concern, impacting on the legitimacy of the police, but equally seriously hampering the functions of the police.

Hough made a distinction between normative legitimacy that exists when authorities meet certain objective criteria, such as the absence of corruption and empirical legitimacy, that is based on the perceptions of civilians.63 Although corruption in the police was discussed under a separate heading as part of corruption throughout government institutions, abuse of force – actual and perceived – requires specific attention.

As mentioned above, the procedural conduct of officers, in this case following the criminal procedural code, will also impact on police legitimacy. According to respondents, while the twenty-four hour rule was followed in the minority of cases, the majority of detained and the sample of family and friends, recalled periods stretching from days to three months or even never being arraigned. In a few tragic cases, families reported that the person ‘disappeared’ from police custody or was found dead.

62 Tyler, Tom R., and Jeffrey Fagan. “Legitimacy and cooperation: Why do people help the police fight crime in their communities.” Ohio St. J. Crim. L. 6 (2008): 231. 63 Hough, M. (2010), “Policing, new public management and legitimacy”, in Brookes, S. and Grint, K. (Eds), The New Public Leadership Challenge: The Rhetoric and Reality of Public Reform, Palgrave Macmillan, Bas- ingstoke, pp. 70-84.

76 77

FCA_KenyaAnalysis.indd 76 24.1.2020 12.07 Detention order

According to Article 49 of the Constitution under ‘Rights of arrested persons’ An arrested person has the right—

(f) to be brought before a court as soon as reasonably possible, but not later than––

(i) twenty-four hours after being arrested; or

if the twenty-four hours ends outside ordinary court hours, or on a day that is not an ordinary court day, the end of the next court day.

Article 49 of the Constitution is reiterated in the Prevention of Terrorism Act under Section 32 ‘Right to be released’

(1) A person arrested under section 24 (referred to as the suspect) shall not be held for more than twenty-four hours after his arrest unless —

(a) the suspect is produced before a Court and the Court has ordered that the suspect be remanded in custody; or

(b) it is not reasonably practicable, having regard to the distance from the place where the suspect is held to the nearest Court, the non-availability of a judge or magistrate, or force majeure to produce the suspect before a Court before the expiry of twenty four hours after the arrest of the suspect.

(2) A police officer holding a suspect under subsection (1) may release that suspect at any time before the expiry of twenty four hours on condition that the suspect appears before the Court or such other place as may be specified, in writing, by the police officer and may, for this purpose, require the suspect to execute a bond of a reasonable sum on the suspect’s own recognizance.

9.1.3 Abuse of force and radicalisation How police officers treat citizens, especially when using force, is critical in assessing and understanding police legitimacy or lack thereof. Determining the way respondents experience the actions of the police, all three samples were asked to rate a list of possible reasons why individuals will join an illegal organisation. Anger towards the police was the most prominent reason both the detained sample (66.96%) and family and friends (75.97%) identified as reasons for joining illegal organisations. As a follow-up question, respondents were asked if they thought there had to be a catalyst that finally motivated a person to join the organisation. The majority of both the detained sample and family and friends answered in the affirmative (see Figure 84).

According to family members and friends (n=80) the killing of a family member or friend (55%) was the most prominent, followed by government action (51.25%) and the arrest of a family member or friend (45%) as the most influential in going over into action. This sentiment was echoed by the detained sample (n=87), but government was rated first (54.02%), followed by the killing of a family member or friend (51.72%) and the arrest of a family member or friend (39.08%). The following testimony from a father was echoed by many: The security forces are not good people. They arrested my boy (but) they did not find any evidence they released him. But later on come and took him and made him disappear, I can’t tell whether he is dead or alive. I don’t trust them at all.

76 77

FCA_KenyaAnalysis.indd 77 24.1.2020 12.07

such other place as may be specified, in writing, by the police officer and may, for this purpose, require the suspect to execute a bond of a reasonable sum on the suspect’s own recognizance.

9.1.2.1 Abuse of force and radicalisation How police officers treat citizens, especially when using force is critical in assessing and understanding police legitimacy or lack thereof. Determining the way respondents experience the actions of the police, all three samples were asked to rate a list of possible reasons why individuals will join an illegal organisation. Anger towards the police was the most prominent reason both the detained sample (66.96%) and family and friends (75.97%) identified as reasons for joining illegal organisations. As a follow-up question, respondents were asked if they thought there had to be a catalyst that finally motivated a person to join the organisation. The majority of both the detained sample and family and friends answered in the affirmative (see Figure 84).

Figure 84: Presence of a catalyst to initiate joining of an illegal organisation

100%

90% 21.9 24.14 80% 70% 60% 50%

40% 78.1 75.86 30% 20% 10% 0% Family (n=105) Detained (n=116)

Yes No

Figure 83. Presence of a catalyst to initiate joining of an illegal organisation According to family members and friends (n=80) the killing of a family member or friend (55%) was the most prominent, followed by government action (51.25%) and the arrest of a family member or friend (45%) as the most influential in going over into action. This sentiment was When comparing the perception of criminal justice actors, an interesting trend emerges: This sample echoed by the detained sample (n=87), but government was rated first (54.02%), followed by rated anger towards the police in the fifth place (59.4%) as the most prominent reason, after finan- the killing of a family member or friend (51.72%) and the arrest of a family member or friend cial incentives, religious ideology, political ideology and anger towards government. However, when (39.08%).criminal justice The followingrespondents testimony (n=85) were from asked a father to shed was light echoed onto potentialby many: catalysts, The security the majority forces are not(60%) good identified people. the They killing arrested of a family my boy member (but) or they friend, did governmentnot find any action evidence (56.47%) they andreleased the arrest him. But laterof a on family come member and took or friendhim and (44.71%). made himTherefore, disappear, although I can't criminal tell whether justice heactors is dead are reluctantor alive. Ito don't trustrecognise them at anger all. towards the police as a reason for joining an illegal organisation, it is prepared to acknowledge the way the police respond to the threat as catalyst. When comparing the perception of criminal justice actors, an interesting trend emerges: This sampleAs a possible rated consequence, anger towards family the members police in and the friends fifth place (n=133) (59.4%) listed the a polices the mostand military prominent (36.09%) reason, afterfirst whenfinancial asked incentives, to identify religious ‘them’. Respondents ideology, politicalpart of the ideology detained and sample anger (n=136) towards placed government. the po- However,lice and military when incriminal the fourth justice position respondents (22.8%) after (n=85) another were ethnic asked group to in shedKenya, light government onto potential and catalysts,another religion the majority in Kenya. (60%) identified the killing of a family member or friend, government action (56.47%) and the arrest of a family member or friend (44.71%). Therefore, although criminalIn this ‘us’ justice versus actors ‘them’ are mind-set, reluctant the topolice recognise can easily anger start towards to interpret the thepolice legal as framework a reason –for espe joining- ancially illegal human organisation, rights – as working it is prepared against ‘them’,to acknowledge developed tothe protect way the criminals police or respond to tie the to police’s the threat ashands catalyst. behind their backs. As a result it can become ‘acceptable’ for officers to see themselves as being above the law to ‘allow’ them to work around the criminal justice system, seeing that it is being perceived as protecting criminals rather than actively facilitating the prosecution and conviction of criminals. In other words, rather than investigate, gather evidence and follow criminal procedure to secure a conviction, officers resort to taking justice into their own hands through eliminating (execut- 75 ing) a possible suspect, instead of acting within the boundaries of the law. In becoming the ‘judge, jury and executioner’ by resorting to extrajudicial killings, these police officers become part of the problem and criminals themselves. According to a report by Amnesty International, Kenya was ranked as the country with the most cases of police shootings and killing of civilians in 2017. The report indicated that by October 2016, a total of 122 out of 177 cases of extrajudicial killings in Africa had been reported in Kenya. This placed Kenya ahead of fourteen other African nations namely Botswana, Burundi, Cameroon, Chad, Côte d’Ivoire, Gambia, Mauritania, Nigeria, Somalia, South Sudan, Sudan, Tanzania, Togo and Zambia where cases of police harassment and deaths had also been recorded.64 In addition to extrajudicial killings, Kenyan security forces had also been implicated in the disappearances of terrorism-related suspects especially after the state was unable to offer sufficient evidence

64 Silas Apollo. Amnesty report says Kenya tops Africa in extrajudicial killings’, Daily Nation, 23 February 2017. Available at https://www.nation.co.ke/news/Kenya-top-Africa-police-shootings/1056-3824890-1183k27/ index.html (accessed on 24 February 2018).

78 79

FCA_KenyaAnalysis.indd 78 24.1.2020 12.07 in legal proceedings. As a result the perception under communities most affected is that security agencies rather eliminate terrorism-related suspects than going through the effort of building a successful case. In one of the most prominent cases, lawyer Willie Kimani, his client Josphat Mwenda and their driver Joseph Muiruri were abducted and subsequently executed in July 2016.65

Against this background, it is not surprising that during focus group discussions the sentiment was raised that there existed an ‘unofficial policy’ to rather eliminate suspects associated with terrorism offences. This perception was echoed by respondents. For example, according to one respondent: Automatically terror suspects will be murdered without any detailed investigations which is wrong according to me. That’s why I may hesitate to report on terror suspects. Not coming forward with information is being perceived by the police as being part of ‘them’, but even more important to take note of, it is a criminal offence according to Section 10 of the Prevention of Terrorism Act: A person who harbours or conceals, or prevents, hinders or interferes with the arrest of a person knowing, or having reason to believe that such person –

(a) has committed or intends to commit a terrorist act; or

(b) is a member of a terrorist group,

commits an offence and is liable, on conviction, to imprisonment for a term not exceeding thirty years.

However, considering existing personal experience on how those who previously came forward with information were treated, the public finds itself in a catch-22 situation.

“Moto tatu” – the triple threat

In one focus group discussion, participants referred to “moto tatu” (triple effect) that vulnerable communities are confronted with: At one end, Al-Shabaab presents a threat to the community in the sense that they do not discriminate when executing attacks that lead to the community paying the price. In another manifestation of being a target, 29 respondents (41%) part of family members and friends sample indicated that either they or another family member were harassed at the hands of the illegal organisation following the arrest of a family member or friend. 16% were abducted or disappeared, 11% were approached by the illegal organisation and threatened and 13% were killed. It is however interesting to note that in 19% of the cases members of the illegal organisation approached the family and offered support.

On the other end, the security agents are not any better since they too target communities. To illustrate this concern, of the 49 respondents part of the family sample, 39% indicated that they or another family member were harassed by security forces following the arrest of a family member or friend, 25% were assaulted, 23% disappeared and 13% were killed. Being referred to as ‘collective responsibility’ as one respondent explained: The police use the aspect of collective responsibility. They arrest everyone including the one forwarding the information. Knowing the treatment that will follow for coming forward with information, can one really blame people for not coming forward?

This all occur in areas that are underdeveloped, without any sign of development, as the third onslaught.

Consequently, the community bears the brunt of this ‘triple effect’.

65 Irene Ikomu. Extra-Judicial Killings and Enforced Disappearances in Kenya: Balancing the Respect for Hu- man Rights and Maintaining National Security. Heinrich-Böll-Stiftung,10 April 2018. Available at https:// ke.boell.org/2018/04/10/extra-judicial-killings-and-enforced-disappearances-kenya-balancing-respect- human-rights (accessed on 18 September 2018).

78 79

FCA_KenyaAnalysis.indd 79 24.1.2020 12.07 Another contributing factor for this division relates to the expectations the public has towards police officers as defenders, protectors and the explicit responsibility to enforce the law within the framework of human rights. Consequently, the police are seen as an extension of the government implying that the public will scrutinise the actions of individual officers. This kind of scrutiny often leads to criticism, and criticism from the public and the media feeds into the idea of ‘us’ versus ‘them’.

To the more extreme, as a consequence of extrajudicial killings, forceful disappearances and kidnappings by security agents the only way for some citizens to push back out of revenge is to join terror groups. In addition to the effect it has on the public, officers need to be reminded that they are not above the law, especially with the introduction of the International Convention for the Protection of All Persons from Enforced Disappearance (ICPPED) that Kenya signed on 6 February 2007 (not yet ratified).

Falling Short on Human Rights Standards

While Kenya signed the International Convention for the Protection of All Persons from Enforced Disappearance (ICPPED), it has not ratified the Convention, and consequently the existing legal framework falls short of international human right standards, and is thus unable to comprehensively deal with the problem. The Convention requires that state parties must enact specific laws establishing the crime of enforced disappearance. They must investigate complaints and reports of enforced disappearance and bring those responsible to justice. Other obligations are of a preventive nature, such as the obligation to detain persons only in officially approved and monitored institutions in which all prisoners are reg- istered, the absolute right to Habeas corpus (a legal action, through which a prisoner can be released from unlawful detention, that is, detention lacking sufficient cause or evidence), the interdiction of concealment of the whereabouts of arrested persons which are in this way placed outside the protection of the law, as well as the right to receive information on prisoners. The Convention recognises the right of victims and their families to know the truth regarding the circumstances and fate of the disappeared person. It also treats the unlawful abduction of children whose parents were victims of enforced disappearance as well as the faking of these childrens’ identities and their adoption.66

9.1.4 Conduct investigations and collect evidence

Respondents were asked if the prosecution was able to present physical and forensic evidence in court. Of the 55 detained respondents who answered the question, 60% indicated that no physical evidence and no (79%) forensic evidence were presented in court; whereas of the 26 respondents – part of the family sample – 85% indicated that no forensic or physical evidence were presented in court. For any case to be successfully prosecuted, the police must have sufficient evidence to establish that the suspect committed the crime he/she is being charged with. In the absence of strong enough evidence, the case will be dismissed. In an interview with Daily Nation, Mr Raphael Daudi Musau, a director at Hawk Eye Technologies Ltd, a private investigations firm in Nairobi, explained that even after the nine- month training, the police lacked hard skills to carry out investigations. Consequently, shortcuts, laxity and impatience when information is not readily available often lead to the failure of an investigation.67

66 Human Rights Information Platform. International Convention for the Protection of All Persons from En- forced Disappearance. 21 May 2014. Available at https://www.humanrights.ch/en/standards/un-treaties/ disappearance/ (accessed on 10 August 2018). 67 Millicent Mwololo. With evidence, there is no room for false steps. Daily Nation, 20 September 2016. Avail- able at https://www.nation.co.ke/lifestyle/dn2/Why-cases-collapse-in-court--/957860-3388322-4jylso/ index.html (accessed on 10 January 2019).

80 81

FCA_KenyaAnalysis.indd 80 24.1.2020 12.07 For example, according to one detained respondent who explained why he would not come forward with information: The security agencies are not well trained to handle any person who report information hence the reporter will be the first victim hence fear for my life. According to another: The police will always arrest the person (who) reported the information without proper investigation.

Consequently, 73.64% of families (n=110) and 62.6% of the detained sample (n=128) felt strongly (support of between 70-100%) that the Kenyan government should divert more resources to the training of security personnel. According to criminal justice respondents (n=106) this ‘need’ was listed first by receiving 95.19% support. In a later question, respondents were asked whether there was a need for more security personnel and whether security personnel should be retrained to deal with the security challenges Kenya is being confronted with. Of the 107 respondents part of the sample representing family and friends, 61.68% strongly supported the proposal that security officials should be retrained, while 56.07% strongly supported the idea that more security officials should be recruited. Amongst 122 respondents being part of the detained sample the retraining of security personnel was listed third on the list of priorities by receiving 59.66% of overwhelming support, while only 40% supported (listed as the least important) the proposal of more security officials. Respondents part of the criminal justice sample (n=104) called for more security officials (71.57%) and the proposal for security officials to be retrained received 69.91% support.

Splitting up recruitment, training and other forms of assistance, respondents further were asked to shed light on what the police need to address in order to enhance its effectiveness in addressing security challenges in Kenya; training on how to better conduct investigations was one of the first priorities amongst all three samples. Criminal justice respondents rated the training second (88.46%) after the need to receive better equipment (89.52%). Second (74.36%) to the list detained respondents noted the need for police officers to receive training on human rights, but this need was placed third (83.34%) by the sample representing family and friends after calling for training for the police to better serve and protect the community (85.32%). It is important to note that 85.44% of respondents part of the criminal justice sample recognised the priority for the police to better serve and protect the community to be followed by initiatives to enhance the training of police officers (83.81%), human rights training (74.76%) and lastly to change the criteria recruiting new police officers. Although the changing of the training criteria for police officers received more support from the other two samples, it was also placed last.

Training is not only required to be more effective, it is also a matter of life and death as one family member explained: You know once you are murdered, and it’s established later on that you were mistaken with the true suspect that doesn’t help, it in fact makes the bitter feeling severe to those left behind. Police should be trained well on conducting proper investigations.

300

250

88.46 200 85.44 74.76 89.52 83.81 150 44.23 79.31 73.27 74.36 100 67.83 61.4 52.14

50 86.24 85.32 83.34 69.44 64.81 55.96 0 Train police to better Train police to serve Provide training on Provide better Enhance the training Change the criteria conduct and protect the human rights equipment to police of police officers to recruit police investigations community officers officers

Family (n=111) Detained (n=119) Criminal Justice (n=105)

In FigureFigure 84. 85 Enhancing respondents the abilities identified of the police the typeto be ableof training to address they security hope challenges to see within the police to enhance its abilities in addressing security challenges in Kenya. In yet another set of questions 80 police officers, part of the criminal justice sample, were asked to rate the training they81 had received and whether they felt confident in dealing with the different facets of tasks associated with countering terrorism. With reference to Table 4, although 64 respondents part of security forces did not receive specialised training (financing of terrorism, cyber investigations, training in improvised explosive devices (IEDs) and hostage situations), respondents expressed confidence in training in community policing and associated interactions with the community, FCA_KenyaAnalysis.inddhuman 81 rights, the appropriate use of force etc. 24.1.2020 12.07

Table 10: Level of confidence in training

(n=64) Not provided Limited Only theory Moderate Excellent Community policing 5,08 3,39 11,86 25,42 54,24 Human Rights 1,67 1,67 16,67 28,33 51,67 Appropriate use of force 8,33 5 3,33 31,67 51,67 Treatment of youth 5,17 3,45 5,17 37,93 48,28 Building trust with community 5 3,33 15 30 46,67

Treatment of women 3,33 3,33 8,33 38,33 46,67 Proportional use of force 5,17 8,62 41,38 44,83 Community engagement 11,48 3,28 9,84 32,79 42,62 Dealing with large crowds 11,26 8,06 9,68 29,03 41,94 Intelligence gathering 5,08 11,87 11,86 35,59 35,59 Crime scene management 18,64 10,17 8,47 35,59 27,21

Active shooter incidents 25,42 16,95 8,47 28,81 20,34 Advanced investigation 21,67 21,67 13,33 23,33 20 techniques

79 In Figure 85 respondents identified the type of training they hope to see within the police to enhance its abilities in addressing security challenges in Kenya. In yet another set of questions police officers, part of the criminal justice sample, were asked to rate the training they had received and whether they felt confident in dealing with the different facets of tasks associated with countering terrorism. With reference to Table 4, although 64 respondents part of security forces did not receive specialised Terrorist tactics in the country 22,22 19,05 7,94 33,33 17,46 training (financing of terrorism, cyber investigations, training in improvised explosive devices (IEDs) and hostage situations), respondents expressed confidence in training in community policing and associatedHostage survival interactions with the 26,98community, human19,05 rights, the appropriate12,7 use of23,81 force etc. 17,46 Convoy planning 38,1 11,11 11,11 23,81 15,87 Table 10. Level of confidence in training IED counter-measures 41,27 23,81 9,52 11,11 14,29 Differentiate (n=64) between civilians 26,98Not provided Limited19,05 Only17,46 theory Moderate25,4 Excellent11,11 andCommunity terrorist policings in crowds 5,08 3,39 11,86 25,42 54,24 Human Rights 1,67 1,67 16,67 28,33 51,67 Appropriate use of force 8,33 5 3,33 31,67 51,67 AftermathTreatment of of youth a roadside IED 46,675,17 3,4518,33 5,1715 37,93 11,67 48,28 8,33 Building trust with community 5 3,33 15 30 46,67 Treatment of women 3,33 3,33 8,33 38,33 46,67 Cyber crime investigations 58,33 13,33 6,67 13,33 8,33 Proportional use of force 5,17 8,62 41,38 44,83 Community engagement 11,48 3,28 9,84 32,79 42,62 FinancingDealing with largeof crowds terrorism 46,6711,26 8,0621,67 9,6815 29,03 10 41,94 6,67 investigationsIntelligence gathering 5,08 11,87 11,86 35,59 35,59 Crime scene management 18,64 10,17 8,47 35,59 27,21

WithActive Figure shooter incidents86 and Table 4 in mind25,42 the immediate16,95 question8,47 that come28,81s to the 20,34fore is whether securityAdvanced officials investigation see techniques themselves as21,67 ‘able’ to 21,67deal with vulnerable13,33 communities23,33 and20 wheter the trainingTerrorist tacticsthat inofficers the country received really22,22 was sufficient19,05 to 7,94deal with 33,33practical challenges17,46 when deployedHostage survival. Considering the overwhelming26,98 experience19,05 of12,7 family members23,81 and friends17,46 of those whoConvoy went planning through the criminal justice38,1 and the11,11 experiences11,11 of respondents23,81 who themselves15,87 went throughIED counter-measures the criminal justice system41,27 (detained 23,81sample), the9,52 concern is 11,11that security14,29 officers might wellDifferentiate be overconfident between civilians in both their26,98 training 19,05and the abilities17,46 they 25,4had mastered11,11 since. Being detachedand terrorists from in crowds the public, officers’ hope to protect and serve can have severe consequences in broadeningAftermath of a the roadside divide IED between 46,67‘us’, security18,33 officials, and15 ‘them’ the11,67 public, most8,33 notably the vulnerableCyber crime investigationscommunities. 58,33 13,33 6,67 13,33 8,33 Financing of terrorism investigations 46,67 21,67 15 10 6,67 Figure 86: Ability of security officials to work with vulnerable communities

100% 90% 31.25 26.56 80% 34.38 70% 56.45 60% 26.56 50% 31.25 32.81 40% 17.74 30% 21.88 25 42.19 20% 19.35 10% 12.5 6.45 10.94 0% 4.69 Deal with vulnerable Deal with the families of Deal with the victims of Deal with terrorists communities suspected terrorists terrorism

Not Applicable No In most cases Yes

Figure 85. Ability of security officials to work with vulnerable communities Despite placing changes to the criteria in recruiting police officers last, everything starts with recruiting the right person with the right temperament to join law enforcement. Worldwide the 82 83 challenge has been to attract the right person to join law enforcement as a profession and not just another job. This requirement is particularly important considering the type of pressures, demands, and expectations the community has of police officers, while at the same time keeping

FCA_KenyaAnalysis.indd 82 80 24.1.2020 12.07 With Figure 86 and Table 4 in mind, the immediate question that comes to the forefront is whether security officials see themselves as ‘able’ to deal with vulnerable communities and wheter the training that officers received really was sufficient to deal with practical challenges when deployed. Con- sidering the overwhelming experience of family members and friends of those who went through the criminal justice system and the experiences of respondents who themselves went through the system (detained sample), the concern is that security officers might well be overconfident in both their training and the abilities they had mastered since. Being detached from the public, officers’ hope to protect and serve can have severe consequences in broadening the divide between ‘us’, security officials, and ‘them’ the public, most notably the vulnerable communities.

Despite placing changes to the criteria in recruiting police officers last, everything starts with recruiting the right person with the right temperament to join law enforcement. Worldwide the challenge has been to attract the right person to join law enforcement as a profession and not just as another job. This requirement is particularly important considering the type of pressures, demands and expectations the community has for police officers, while at the same time keeping in mind that it is a dangerous career with little financial rewards (although often with good benefits such as medical aid and housing). Despite these challenges, law enforcement presents a vast array of career opportunities. Recruiting the best candidate to serve in law enforcement is important, but the focus has to be on civil service overall and the candidates’ aim should be to serve their country and the community. This internal drive to serve the community is further supported by the ability to effectively communicate (considering that most officers interact with the public at one stage or another) and the ability to think on his/her feet. This level-headedness is required when the officer needs to respond to highly stressful situations by remaining calm. The latter also touches on the need to be mentally and physically strong. Being called on to deal with the most tragic of situations and its human toll on police officers manifests in high divorce rates, alcohol abuse and suicides. When criminal justice respondents (n=101) were asked to rank the reasons why they joined civil service, the five most prominent reasons are as follows: To serve my country and community (88.89%), to protect my country and community (82%), making a difference (75,25%), always wanted to be a police officer, prosecutor etc. (54.17%) and the sense of belonging he/she experiences (46.39%).

When criminal justice respondents (n=104) were asked to rank the reasons why they wanted to remain a civil servant, the five most prominent reasons are as follows: Responsibility to humanity (94%), responsibility to Kenya (93.27%), being part of change (88.34%), to stand up for something he/she believes in (85.14%) and the belonging the person experiences being part of the organisation he/she is part of (66%). Financial benefit received 49.49% and being respected 36.36%.

9.1.5 Religious and tribal representation of law enforcement

Before discussing above profiling based on ethnicity and religion it is necessary to determine the importance respondents placed on religious and ethnic/tribal identity as well as diversity within security forces to reflect the religious and ethnic demographic profile of the geographical areas where the study was conducted.

Diverse societies are traditionally difficult to police, especially when police interact with communities different to the main biographic profile the police in that area represents. These communities consistently have less trust in police for two primary reasons: firstly, diverse communities often report having direct negative personal experiences with police officers. Secondly, there is often evidence of discrepancies in the procedures officers refer to when dealing with communities different than the majority. Although a minority can be a majority in a particular geographical area, the challenge comes down to whether the police deployed in that area represent the majority the public can relate to.

Racial and ethnic minority perceptions that the police lack lawfulness and legitimacy, based largely on their interactions with the public, can lead to distrust of the police. Distrust of police has serious consequences. It undermines the legitimacy of law enforcement; without legitimacy, police lose their

82 83

FCA_KenyaAnalysis.indd 83 24.1.2020 12.07

Figure 87: Did security forces in the area reflect the main religious affiliation found in the community

100% 90% 80% 70% 61.47 60% 88.98 83.74 50% 40% 30% 20% 38.53 10% 11.02 16.26 0% Figure 87: Did security forcesCriminal in the area Justice reflect (n=109) the main religiousFamilies affiliation(n=118) found in Detainedthe community (n=123)

Yes No 100% 90% It is especially interesting80% to note that 61.47% of criminal justice actors as opposed to 88.98% of family members70% and friends61.47 of individuals who went through the criminal justice system were under the impression60% that the majority did not represent the same religion as that of the 88.98 83.74 community. 50% 40% Not representing30% the demographic profile of the community can manifest in religious insensitivity that often manifests in the conduct of police officers when they interact with members of a faith 20% not their own. 38.53 10% 11.02 16.26 0% Similar to religious representation,Criminal Justice (n=109) the overwhelmingFamilies (n=118) perceptionDetained amongst (n=123) all three samples was that the police are not representative of the communities they serve. Yes No FigureFigure 88: Majority 86. Did of securitypolice officers forces represent in the area ethnic reflect groups the in themain area religious affiliation found in the community

It is especially interesting to note that 61.47% of criminal justice actors as opposed to 88.98% 100% of family members90% and friends of individuals who went through the criminal justice system were under the impression80% that the majority did not represent the same religion as that of the community. 70% 60% 79.09 83.74 90.6 Not representing 50%the demographic profile of the community can manifest in religious insensitivity that often manifest40%s in the conduct of police officers when they interact with members of a faith not their own. 30% 20% 10% 20.91 16.26 Similar to religious representation, the overwhelming perception amongst9.4 all three samples was that the police are0% not representative of the communities they serve. Criminal justice Detained Families

Figure 88: Majority of police officers represent ethnic groupsYes in theNo area

Figure 87. Majority100% of police officers represent ethnic groups in the area Despite the fact that90% police officers did not reflect the ethnic/tribal or religious composition of the geographical area80% they serve, the suggestion to recruit police officers from the region was only supported by70% 25.49% of respondents part of the criminal justice sample (n=105) and 27.82% of detained60% respondents79.09 (n=119). Respondents83.74 part of families and friends (n=111) ability and authority to function effectively. Legitimacy, in turn, has been90.6 found to influence legal werecompliance more open and to50% people’sthis suggestion willingness as to43.12% support rated the police this suggestionin fighting crime. between 70-100%. 40% Assessing whether30% the police is representative of the community, respondents were asked if the police 20% in the area is representative of their religion and ethnicity. Starting with religion, the dominant religion 10% 20.91 16.26 of the areas most affected by violent extremism is Islam. When respondents9.4 were asked if criminal jus- 0% tice actors represented theCriminal majority justice religion in theDetained area, all three samplesFamilies answered in the negative as presented in Figure 87. It is important to note that according to criminal justice respondents 38.53%82 were under the impression that security forcesYes reflectedNo the same religion as those in the area they

were deployed. Despite the fact that police officers did not reflect the ethnic/tribal or religious composition of the Itgeographical is especially interesting area they to serve,note that the 61.47% suggestion of criminal to recruit justice policeactors asofficers opposed from to 88.98%the region of fam was- onlyily supported members and by friends 25.49% of individualsof respondent who wents part through of the the criminal criminal justice systemsample were (n= under105) theand impression that the majority did not represent the same religion as that of the community. 27.82% of detained respondents (n=119). Respondents part of families and friends (n=111) were more open to this suggestion as 43.12% rated this suggestion between 70-100%. Not representing the demographic profile of the community can manifest in religious insensitivity that often manifests in the conduct of police officers when they interact with members of a faith not their own.

Similar to religious representation, the overwhelming perception amongst all three samples was that the police are not representative of the communities they serve.

82 84 85

FCA_KenyaAnalysis.indd 84 24.1.2020 12.07

Despite the fact that police officers did not reflect the ethnic/tribal or religious composition of the 9geographical.1.2.3 Profiling area based they serve, on religion the suggestion and ethnicity to recruit police officers from the region was only sup- Familyported bymembers 25.49% expressedof respondents a greaterpart of the concern criminal for justice being sample singled (n=105) out and by 27.82%the police of detained based on religiousrespondents affiliation (n=119). Respondents when compared part of familiesto gender, and friends age, (n=111)nationality, were more physical open to appearance this suggestion and politicalas 43.12% affiliation rated this. Perceptionssuggestion between of being 70-100%. singled out correspond with unjustified disparities in the rates of stops, searches and arrests. 9.1.6 Profiling based on religion and ethnicity Figure 89: Singled out based on religious affiliation Family members expressed a greater concern for being singled out by the police based on religious affiliation when compared to gender, age, nationality, physical appearance and political affiliation. Perceptions of being singled out correspond with unjustified disparities in the rates of stops, searches and arrests.

Detained (n=126) 30.95 4.76 9.52 15.08 23.81 15.87 Comparing being singled out based on religion versus ethnicity, respondents expressed a greater concern 9towards.1.2.3 Profilingbeing singled based out onon religionthe basis andof religion ethnicity than ethnicity. In the aftermath of 9/11, the global Mus- Familylim community members complained expressed of increased a greater Islamophobia concern forthat being can broadly singled be describedout by the as, police ‘indiscriminate based on religiousnegative attitudesaffiliation or emotionswhen compared directed at to Islam gender, or Muslims’. age, 68nationality, On a local level, physical Islamophobia appearance resulted and politicalin Familypoor (n=117) public-policeaffiliation. Perceptions relations29.91 and of outrage being1.71 9.4 singledagainst 14.53outcounterterrorism correspond18.8 withmeasures unjustified that25.64 were disparities framed asin the ratesevidence of stops, of anti-Muslim searches discrimination. and arrests. Within security agencies, especially the police, the term ‘institutional Islamophobia’ was introduced to describe poor relations the state has with the Muslim community, 69 Figuretargeted 89: Singled police 0%out stop based and10% on search religious20% powers, affiliation30% and the40% subsequent50% distrust60% found70% within80% Muslim 90%communities.100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

ComparingDetained (n=126) being singled30.95 out based4.76 on 9.52religion 15.08versus ethnicity,23.81 respondents15.87 expressed a greater concern towards being singled out on the basis of religion than ethnicity. In the aftermath of 9/11 the Muslim community across the globe complained of increased Islamophobia that can broadly be described as ‘indiscriminate negative attitudes or emotions directed at Islam or 64 Muslims’.Family (n=117) On a local level29.91 Islamophobia1.71 9.4 resulted14.53 in poor public18.8 -police relations25.64 and outrage against counterterrorism measures that were framed as evidence of anti-Muslim discrimination. Within security agencies, especially the police, the term ‘institutional Islamophobia’ was introduced to 0%describe10% poor20% relations30% the state40% has50% with the60% Muslim70% community,80% 90%targeted100% police stop and search powers, and the subsequent distrust found within Muslim communities.65 No 1-20% 30-40% 50-60% 70-80% 90-100% Figure 90: Singled out based on ethnic/tribal affiliation Figure 88. Singled out based on religious affiliation Comparing being singled out based on religion versus ethnicity, respondents expressed a greater concern towards being singled out on the basis of religion than ethnicity. In the aftermath of 9/11 the Muslim community across the globe complained of increased Islamophobia that can broadlyDetained (n=123) be described as ‘indiscriminate50.41 negative attitudes9.76 or9.76 emotions8.13 directed13.82 at 8.13Islam or Muslims’.64 On a local level Islamophobia resulted in poor public-police relations and outrage against counterterrorism measures that were framed as evidence of anti-Muslim discrimination. Within security agencies, especially the police, the term ‘institutional Islamophobia’ was introducedFamily (n=119) to describe poor relations44.54 the state has9.24 with 5.04the Muslim8.4 community,13.45 targeted19.33 police stop and search powers, and the subsequent distrust found within Muslim communities.65

Figure 90: Singled out based on ethnic/tribal affiliation 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100%

No 1-20% 30-40% 50-60% 70-80% 90-100%

Figure 89. Singled out based on ethnic/tribal affiliation 9Detained.2 Other (n=123) security agencies 50.41 9.76 9.76 8.13 13.82 8.13 68 Bleich, E. ‘What Is Islamophobia and How Much Is There? Theorizing and Measuring an Emerging Compara- In comparisontive Concept’, to American the police Behavioral, other Scientist, security 55, agencies 2011: p. 1582. had a better track record based on the analysis69 Julian ofHargreaves; questionnaires. Police Stop However, and Search duringWithin British focus Muslim group Communities: discussions Evidence (presented From the Crimelater in this section)FamilySurvey (n=119)respondents 2006–11, The British provided Journal44.54 ofcontext Criminology, to their Volume personal9.24 58, Issue5.04 6,interactions 58.4 October13.45 2018, that Pages were 1281–130219.33 not always positive. Starting with questionnaire analysis, 20.51% of the family and friends sample (n=84) 84 85 0% 10% 20% 30% 40% 50% 60% 70% 80% 90% 100% 83 No 1-20% 30-40% 50-60% 70-80% 90-100%

FCA_KenyaAnalysis.indd 9 85.2 Other security agencies 24.1.2020 12.07

83 9.2 Other security agencies

In comparison to the police, other security agencies had a better track record based on the analysis of questionnaires. However, during focus group discussions (presented later in this section) respondents provided context to their personal interactions that were not always positive. Starting with questionnaire analysis, 20.51% of the family and friends sample (n=84) rated their trust in the military between 70-100%, in contrast with the 7.6% the police received. 14.67% expressed the same level of trust in intelligence agencies and 12.99% in prison authorities. Similarly, 20.73% of the detained sample (n=91) rated their trust in the military between 70-100%, followed by prison authorities (15.39%) and intelligence agencies (13.09%). A mere 5.88% expressed the same level of trust in the police. In comparison, 62.5% of criminal justice respondents (n=80) had between 70-100% trust in both the military and intelligence agencies, followed by 61.25% in prison officials and 55.7% in the police.

According to one participant, three of his grandchildren disappeared three years before at the alleged hands of the Kenyan Defence Force (KDF). The young people were on their way to see their father who was sick. They were then arrested by KDF personnel who were manning a roadblock. When the reports of their arrests reached their family they decided to go and search for the children. They were referred to almost all the police stations in the area, but it was in vain. There was no one willing to inform them of the childrens whereabouts. A month later a man called the grandfather informing him that he knew where his children were and asked if he would want to see him. The grandfather quickly organised to meet him and indeed he introduced himself as a KDF member that had been sent by his commanding officer to come and get Kshs 200000 if he wanted his children. After further consultation, they settled on Kshs 100000 which was to be paid in instalments. The grandfather paid instalments of up to Kshs 85000 within two months; during that time,he was assured that his children were safe and sound. The KDF man requested him to give him clothes for the children and Kshs 3000 to buy them shoes, which he did. He was then requested to wait for his children by the roadside at midnight where a vehicle would drop them. The participant waited a week for the children and since then (three years past) had not seen his grandchildren or the alleged KDF officer. According to the family and friend sample, corruption in the military however is the exception to the rule as 95.06% never had to pay a bribe to a military officer. Yet, 1.23% had to pay a bribe between one and two times, 2.47% between three and four times and 1.23% five times and more. Amongst the detained sample, 96.59% never had to pay a bribe, but 1.14% had to pay a bribe between one and two times and 2.27% between three and four times.

In another focus group discussion, participants made reference to an al-Shabaab attack in July 2017 targeting KDF vehicles at Handaki East that resulted in the injury and death of two soldiers. In retali- ation, KDF allegedly started beating community members and destroying the only water hole that was their source of water. Community members alleged that they had shared information with the KDF prior to the attack and they failed to take heed of their warnings. One community member was killed as a result of the operation and scores injured.

Although the mandate of any military is to protect the country against foreign enemies, respondents were asked to rate their willingness to call on a number of actors (including family members and friends) when in physical danger - the military was included in this list. Recognizing that it is not traditionally the mandate of the military, all three samples listed the military last. It is however important to note that 29.7% of criminal justice respondents (n=101) rated the likeness of calling the military between 70-100%. In contrast, 5.56% of respondents part of the family and friends (n=109) and 5.22% of the detained (n=121) samples expressed the same level of eagerness to call on the military when in danger. When the detained and family and friends samples were asked to indicate how poor their interactions had been with the different criminal justice actors, including the military, of the 31 respondents part of the family and friend sample and 71% respondents of the 24 respondents who interacted with the military rated this interaction as ‘poor’. It is important to note that whereas detained respondents considered the police as worst, respondents part of the family and friend sample identified the military as worse (see Table 11).

86 87

FCA_KenyaAnalysis.indd 86 24.1.2020 12.07 Considering the section on police legitimacy, the vast majority (82%) of the 122 respondents being part of the detained sample rated their interactions with the police as ‘poor’, in comparison with 65% of the 86 respondents, part of the family and friends sample. Both samples placed intelligence agencies fourth as 58% of the detained and 65% of family and friends samples rated their interaction with intelligence agencies as ‘poor’. Considering the exposure of detained respondents to legal proceedings 77% of the 55 respondents who interacted with the prosecution authority rated it as ‘poor’. Families and friends due to their negative interaction with prison authorities placed prisons second (74%) after the military. In contrast with the family and friends sample, the majority of detained respondents rated prison authorities, courts and prison staff between ‘relative’ and ‘good’.

Table 11. Negative interactions with criminal justice actors

Position Detained Family 1 Police (n=122) Military (n=22) 82% 77% 2 Military (n=24) Prison Authorities (n=31) 71% 74% 3 Prosecutors (n=55) Police (n=86) 71% 65% 4 Intelligence agency (n=77) Intelligence agency (n=52) 58% 65% 5 Prison Authorities (n=56) Prison staff (n=32) 43% 60% 6 Courts (n=62) Prosecutors (n=30) 42% 56% 7 Prison staff (n=55) Courts (n=35) 40% 52% 8 Defence council (n=44) Defence council (n=30) 36% 43%

86 87

FCA_KenyaAnalysis.indd 87 24.1.2020 12.07 10. Recommendations

At the end of the study, recognising that the study confirmed earlier results while also identifying a few surprises (for example the disconnect between criminal justice actors and the other two samples), the value of any study should be to present solutions instead of only identify and validate known challenges. Finding a lasting solution will however require the involvement of both government and civil society, with the emphasis on the former. Reading Kenya’s counter violent extremist strategy one observation is an unequal balance between what the government of Kenya expect others to do, while there is limited introspection on the role government and its security agencies play in the radicalisation and recruitment process. It is however important to note that at the time of writing, the NCTC the was in the final stages of ‘refreshing’ the strategy.

Being aware of counter strategies is one step in the right direction in reaching out to those who are vulnerable. Secondly, what programs are being presented and by whom directly impact on success. Lastly, what vulnerable communities want to see being implemented and why existing counter and preventative strategies were not successful completes the equation.

10.1 Government of Kenya 1. While the reasons behind the radicalisation and recruitment into violent extremist organisations were highlighted in quite a few independent studies, the first suggestion to the Government of Kenya is to consider these findings in revising its strategy to prevent and counter radicalisation into violent extremist organisations. This suggestion is in line with the ‘Theory of Change’ as presented in Kenya’s CVE Strategy: “our assumption of what it will take to attain the strategic end state – is that evidence-based counter- and de-radicalisation efforts that are collaborative, accountable, sensitive to the risks and mandates of national and human security, and focused on engaging and empowering the public will sharply reduce and eventually end radicalisation and recruitment into violent extremist groups.”As well as the eigth priority under ‘research’: “Ensure that Kenya’s CVE actors have the benefit of a dynamic, action-ready and research- informed understanding of the evolution of violent extremist ideologies, organisational models, and radicalisation methodologies.”

2. Research findings can equally be useful approaching vulnerable communities in addressing identified concerns and challenges to enhance the legitimacy of government and associated agencies.

3. Although the office of the presidency attained the most trust amongst all three samples, the Government of Kenya needs to implement dedicated steps to enhance the public’s trust in government departments. Vulnerable communities were the first to refer to unequal treatment when applying for identity cards, passports and other critical documentation.

4. Identified as the second priority in the national CVE Strategy, enhancing patriotism and building a sense of nationhood start with creating a feeling of belonging and acceptance. This should start with assessing vulnerable communities’ access to identity documents, passports, etc. In other words, more resources should be directed at minority and vulnerable communities instead of less.

88 89

FCA_KenyaAnalysis.indd 88 24.1.2020 12.07 10.2 Kenya Police Service

Often the most obvious advice is the most constructive. To start with, returning to the basic principles of policing is a sure way to build and strengthen police legitimacy.

Implement the nine principles of policing

In 2011 Sir Robert Peel provided nine principles of policing as a guideline to enhancing legitimacy and secure public acceptance of the police that is equally relevant to Kenya:70

1. To prevent crime and disorder is an alternative to the repression as a result of military action.

2. The power of the police to fulfil their duties and function is dependent on public approval of their existence, actions and behaviour and on the ability of the police to secure and maintain the respect of the public.

3. For the police to recognise that to secure and maintain the approval and respect of the public means to secure the willingness of the public to cooperate in upholding the law.

4. To recognise that the extent to which the police can secure cooperation from the public dimin- ishes proportionally with the use of physical force.

5. To secure and enhance public support will not be achieved by calling on public support, but to continuously demonstrate the impartial application of the law independent from political interference, social and financial status of the subject of the investigation and relationships the person may have. Police members also need to demonstrate their willingness to sacrifice themselves in protecting the public.

6. The use of physical force can only be justified as the last resort to restore order within the framework of the law. Force should be limited to what is absolutely necessary in that particular situation for the police to achieve its objective. In other words, physical force should be limited and proportional.

7. For the police to remember that the police are the public and the public the police. Members of the police are members of the public who are employed and paid to act in the interest of the existence and welfare of the community.

8. Police to strictly adhere to police-executive functions and to refrain from infringing from interfer- ing in the responsibility of the judiciary by judging and punishing those perceived to be guilty.

9. Police efficiency should be measured by the absence of crime and disorder and not the visible evidence of police action in dealing with it.

Police officers reflecting the demographics of the community

The overall sentiment is that appointing police officers who know and understand the community, requesting officers to build better relationships with neighbourhoods they serve, reducing officers’ use of aggressive tactics and increasing officer training will enhance police legitimacy. At the same time, the concern is that placing officers in the communities they originate from will facilitate favouritism and corruption. A potential solution can be to rotate officers within the station or between stations in the same geographical area on a regular basis. Furthermore, in an atmosphere of mutual responsibility, commanders and other officers should be able to come forward with information regarding abuse and the misuse of powers of colleagues to a separate command tasked with investigating these allegations.

70 Bronitt, Simon H., and Philip Stenning. “Understanding discretion in modern policing.” Criminal Law Journal, Vol. 35 No. 6 (2011). p. 331.

88 89

FCA_KenyaAnalysis.indd 89 24.1.2020 12.07 Creating a friendlier environment for the community to interact with the police

Community policing starts with the traditional interception point, the police station. Through creating a separate area where members of the community can interact with an officer(s) outside the vision of other members of the public, a ‘safe space’ will be created. These areas should however not give the impression of being an interrogation room. Additionally, members of the public should receive the same treatment irrespective of social standing, gender, religious affiliation and ethnic/tribal origins. Respect for gender roles and ethnic/tribal backgrounds will imply that officers reflecting these differences will be tasked to interact with the respective members of the community. Through this strategy people will be more inclined to ask for assistance, report suspicious activities etc. Since a common perception is that police officers do not respond to information provided, even the slightest inclination that the matter at hand will not receive the required attention should be prevented and addressed. Officers should keep a ‘brought forward’ register that should be inspected on a regular basis and when tasks were completed, the members of the public who reported the matter should receive feedback.

Enhanced training of police officers in investigation techniques, human rights and intelligence

According to Mwazighe (2012), the effective combating of violent extremism requires a delicate balance between enhancement of security on the one hand and respecting of individual rights and freedoms on the other.71 The latter is not only critical in so far as ensuring that innocent suspects are not victimised but also in creating confidence in the criminal justice system. Due process should also be seen in light of international and domestic human rights obligations imposed on the state. The key obligations to be observed include: the presumption of innocence, non-discrimination, right of access to legal counsel; prohibition of torture, inhuman and degrading treatment, right against self-incrim- ination/right to remain silent, protection from incommunicado and arbitrary detention and right to bail. The right to bail has been particularly controversial especially when viewed from a public safety perspective. In Republic v Salim Said Nassow and two others, a High Court decision in 2016, it was noted that the seriousness of the offence is not a ground per se to deny bail72 but that the potential of alleged terrorists to cause further harm influences decisions (Oluseye Oledaji Shittu v Republic, 2016).73

When respondents were asked where the Kenyan government should divert most of its resources, criminal justice respondents (n=106) requested emphasis on the training of security personnel (95.19%). Less than 73.64% of respondents part of family members and friends sample (n=110) called for training of security forces while 62.6% of respondents part of the detained sample (n=128) identified the same need. Regarding the type of training, training in how to conduct better investigations rated high amongst all three samples, followed by training in human rights and how to protect and serve the community better. Training offered to criminal justice actors in religious literacy and how to differentiate between Islam as a religion and extremist ideology will also avoid identifying religion as the real cause of domestic instability. Training in cultural/tribal and religious diversity (in itself a form of community dialogue) will send the right signals that the security agencies are serious in enhancing trust between itself and the community.

Although above initiatives form only one approach in building a relationship of trust between the public and the security agents, other initiatives should be invested in. This will bridge the gap of mistrust that has been an impediment to safety and security in especially vulnerable communities.

Public sensitization forums, information sharing and community policing will further enhance police legitimacy if conducted in a manner that is honest and not a one-sided interest to collect intelligence.

71 R v Salim Said Nassoro & 2 Others High Court at Nairobi Misc. Criminal Application No. 351 of 2016. 72 Oluseye Oledaji Shittu v Republic High Court at Nairobi Misc. Criminal Application No. 130 of 2016. 73 Bronitt, Simon H., and Philip Stenning. “Understanding discretion in modern policing.” Criminal Law Journal, Vol. 35 No. 6 (2011). p. 331.

90 91

FCA_KenyaAnalysis.indd 90 24.1.2020 12.07 Community involvement The nature of police attitudes toward citizens and vice versa will determine whether the two will be able to work together. The reality is that a negative relationship between the police and the community will result in negative attitudes that will consequently manifest in mutual ill feelings, lack of respect, disorder, and inefficient police functioning.74

The need for community involvement in countering violent extremism is recognised globally as captured in the UN Plan of Action to Prevent Violent Extremism.75 In the absence of community support and involvement, initiatives to counter violent extremism, will be hampered. Furthermore, without community involvement the early detection of radicalisation and identification of suspects will equally be severely hampered. Similarly effective long term responses require collective community action seeking to address underlying reasons for individuals to be vulnerable to violent extremism. Community involvement should take place at regional and national levels with the involvement of both community members and community leaders representing all sectors (religious leaders, youth and women organisations, etc.).

10.3 Agencies involved in counter violent extremist activities Civil society organisations and local government operational in vulnerable communities have a very important role to play, especially in preventing radicalisation into violent extremist organisations while addressing the underlying causes or conditions conducive to radicalisation and terrorism.

Criminal justice respondents regarded initiatives initiated by civil society as more important and successful with specific reference to community dialogue, youth-led initiatives, religious discussion and education and skills development. Family and detained samples on the other hand identified amnesty and prison programs as the most important. In contrast, criminal justice actors expressed the least trust in the success of amnesty and prison projects. These results are particularly important considering the emphasis placed on the initiatives. The perception of importance and success of counter initiatives directly relates to the motivations respondents identified as the reasons behind radicalisation. This is particularly evident in the religious and financial incentives criminal justice actors identified as the reasons why individuals joined violent extremist organisations. It therefore calls for an evaluation into the reasons why those most at risk expressed limited trust in community dialogue, religious discussions, sport events and education and skills development.

Online CVE Strategies

The increased reliance of online modalities for recruitment of violent extremists and perpetuation of their agenda demands robust online counter violent extremism strategies.76 This should include prevention and responsive mechanisms. This being said, there is a difference between the more face-to-face recruitment strategy into al-Shabaab and the online presence of Islamic State. Consequently, counter messaging should be developed with the actual, instead of the perceived radicalisation and recruitment strategies of each organisation in mind even in the same region within a country.

74 Radelet, L. The Police and The Community. New York, NY: Macmillan Publishing Co. (1986). p. 280. 75 UN Secretary-General Plan of Action to Prevent Violent Extremism (2016). 76 Sarah Ladbury, “Women and Extremism: the Association of Women and Girls With Jihadi Groups and Im- plications for Programming” (2015), 8 accessed March 1 2019, https://assets.publishing.service.gov.uk/me dia/57a0897fed915d622c000245/61578_Women-Extremism-Full-Report.pdf.

90 91

FCA_KenyaAnalysis.indd 91 24.1.2020 12.07 Gendering CVE

As envisaged by UN Security Council Resolutions 1325 and 2242, CVE strategies should be gendered. There is need for a gender lens in designing CVE to adequately capture approaches that address female involvement in violent extremist activities as well as their role CVE77. The involvement of women may be particularly strategic with regard to building bridges between the State and communities. Furthermore, there is need to empower women in the communities to identify and deal with family members involved in extremist activities.

77 HORN and CSCR, Mapping Dynamics and Perceptions of Violent Extremism A Study of Nature, Drivers and Perceptions of Muslim Women and Girls Towards Violent Extremism in Kenya (HORN and CSCR) 24; Sarah Ladbury, “Women and Extremism: the Association of Women and Girls With Jihadi Groups and Im- plications for Programming” (2015), 7 accessed March 1 2019, https://assets.publishing.service.gov.uk/me dia/57a0897fed915d622c000245/61578_Women-Extremism-Full-Report.pdf.

FCA_KenyaAnalysis.indd 92 24.1.2020 12.07 11. Conclusion

The main objective of this research project was to advance a deeper understanding of the ways in which violent extremist organizations use the breakdown of trust between local communities and the state and criminal justice system to harness support for their activities and to recruit individuals at risk. In doing so, interviews were conducted with respondents representing individuals who had been arrested, detained or convicted on terror charges in Kenya (referred to as the detained sample), their immediate families and friends, communities in the targeted counties as well as officials representing the Kenyan government and criminal justice actors.

At the end of the study – based on information received through making use of above questionnaires, focus group discussions and validation meetings – findings confirmed that perceptions of unjust treatment and a history of mistrust between criminal justice actors and local communities is an unfortunate reality. Starting with the reasons for joining illegal organisations, anger towards the police was identified as the main reason for joining illegal organisations such as al-Shabaab. Despite initiatives to reform the police in Kenya, in practice it is especially the youth that continue to have run-ins with the police over accusations of violent extremism and terrorism. Experiencing a history of extra-judicial killings and forced disappearances, many are still fearful of being identified as potentially involved in violent extremist activities and be a victim of these practices. Most concerning is the fear that getting onto the radar of security forces is based on perceptions. Similarly, prosecutors highlighted concerns over the quality of case dockets brought before them for procecution and challenges associated with securing successful prosecution due to limited evidence as a result of insufficient investigations and an inability to turn intelligence into evidence. Consequently, prosecutors increasingly rely on plea bargaining instead of initiating prosecutions.

Secondly, this breakdown of trust between vulnerable communities and security agencies – in particularly law enforcement – places a questionmark over the success of initiatives such as community policing that is built on a positive relationship between the police and the public.

FCA_KenyaAnalysis.indd 93 24.1.2020 12.07 Finn Church Aid P.O. Box 210 FI-00131 Helsinki Finland

FCA_KenyaAnalysis.indd 94 24.1.2020 12.07