It has been observed in other contexts that terror tactics tend to provoke public revulsion that can be self-defeating as long as the response is not disproportionately brutal and indiscriminate.  This self-defeating tendency was certainly present in Garissa. People described how the university attack was a wake-up call for everybody. When local Muslims saw the innocent victims of the attack they came together to reject al-Shabaab. As one local woman put it: “it pained us to see that children had been killed. This is not in Islam teachings”’. 
In the immediate aftermath, there was a wave of shared grief and horror that ran across communities. Many Muslims donated blood and sheltered victims. Under the leadership of local sheikhs, they also organised dialogue forums that enabled them to come together as a community.
In this case therefore, Shabaab’s strategy backfired. Local Muslims hadn’t seen the attacks on local hotels, churches, police and the military as their problem. But when the university attack occurred, it affirmed the realisation that many had made when the KDF burnt Garissa town’s market. Local leaders and sheikhs liaised with the government and campaigned throughout the county calling for an end to the violence. They told communities that “killing innocent people is not allowed by the Koran”, and on the BBC and local stations they condemned the killings.
When a local sheikh was denounced by Shabaab commander Gamadhere for siding with the Kenyan authorities, instead of being intimidated, he lashed back: “Why don’t you develop where you are in Somalia?” The story of what happened next illustrates well the role of public trust and unity in protecting Garissa – as well as the policing challenges to be overcome.
Shabaab soon began to threaten the sheikh’s life. He was told by phone to expect visitors and to open the door when they came – but when he asked for police help, he was told they had no transport to come to his aid. It was the local chief who mobilised a crowd of people to chase the assassins away. When the police arrived the next morning, an angry crowd chased them away, saying: ‘”Did you come to pick up a corpse?”
The Sheikh’s story highlights an aspect of Garissa’s resilience: solidarity around rejection of violence. At the same time, it reveals the holes in local security provision that need to be fixed. The university college attack exposed these holes for the world to see. Some locals recall posters warning of a possible attack , and the Technical College in Garissa was even closed as a precaution on the day the university was attacked. Indeed, a high-profile warning had been made in advance about an attack on a Kenyan university,  but despite this only two security guards were on duty.  In the weeks after the attack, Kenya’s Minister of Interior admitted: “There was intelligence that this place was going to be attacked”.  In an indication of how fraught relations between national and county authorities were undermining security, he went on to point the finger at the then Garissa County Commissioner, who he said “received this information and did not act on it”. 
It is widely believed among locals that “corruption made the security agencies allow elements of al-Shabaab into Garissa town”, and that police were given money not to prevent the university attack.  Although Saferworld couldn’t confirm this, the security force response to the unfolding atrocity was undeniably slow and shambolic.  The killings dragged on for 15 hours, and the Daily Nation described in damning terms how officers of the Recce Company of the General Service Unit had taken 11 hours to reach the scene from Nairobi after the alarm was first sounded.  With the world watching, these failures could not be concealed or brushed aside.
Popular revulsion at the attacks and recognition of the shortcomings of the security response triggered a remarkable and successful change in approach. Security provision was restructured. Within a month of the university attack,  acting on the recommendation of north-eastern leaders,  the president appointed a new regional coordinator for the north-east region, with authority over the county commissioners, county cecurity and intelligence committees, local chiefs and responsibility for security and peace initiatives in the region.
As an Ogaden Somali who had had a role in ending the ‘Shifta’ War, Mohamud Saleh commanded respect not only on paper but because of who he was and his track record. As a member of the County Assembly put it: “The coming of the new regional coordination was a serious relief to the terror attacks”. 
Many people spoke about his positive role in healing relations between the government, security agencies and communities in a non-partisan way. According to a county government official:
“Saleh opened the communication between the community and security. He has given out his cell phone number to the public. If any suspect is seen in the community, then he is called directly. That’s how he has been able to pre-empt many of the attacks in Garissa town and its environs. 
Others credited him with unblocking obstacles in command structures and restoring relations between national and county government structures. 
One factor was that he transcended clan divisions and cut through institutional corruption - he would be equally tough about poor performance regardless of whom he dealt with. “Before it was a case of bribery and taking the money from suspects, but after the regional coordinator was appointed, suspects were arrested”. 
Another factor was the signal sent by his clampdown on indiscriminate arrests of young people. If Saleh’s capital was trust, he invested it in community-driven security approaches. Kenya has a community policing system called ‘Nyumba Kumi’ – meaning ‘ten households’ – under which clusters of households are supposed to report suspicious behaviour. In many parts of Kenya, the public has good reason not to trust the Nyumba Kumi model. However, in Garissa, trust in Saleh enabled people to trust in Nyumba Kumi.
Rather than work primarily through highly militarised police units or the army, Saleh’s model was to work through the Administration Police (AP), who were closer to communities. He introduced Nyumba Kumi structures in local communities for informing on suspicious developments. At the same time, he set up a direct line for the public to reach his office, to ensure nothing important got stuck at the wrong level because of corruption or inefficiency. When action was needed he could call upon either a Quick Response Unit in the AP or a Rapid Response Unit in the Kenya Police. With this setup, he could obtain quick, reliable information, then act on reports and challenge inaction at lower levels.
Under the new system, we were told, if an attack occurred, rather than rushing there and rounding up, beating and torturing people, authorities call a barasa (public forum) to identify underlying issues.
The new system includes better training and surveillance equipment,  and is clearly not ‘soft’ in its pursuit of would-be attackers: it includes the profiling of mosques and the investigation of young males who have disappeared. It appears that Somalis and Muslims in general are still subject to more checks, and security checks on other groups are not suitably thorough. The violence has left wounds that will never heal: “We don’t know where our sons went”, cried one tearful group of women.
Nonetheless, the new system’s primary success has been replacing a forceful approach with one that is based on trust and accountability that builds relationships with local communities to gain intelligence.  With this system in place, the police told us that “close collaboration of the community and security agencies has greatly improved.”  Importantly, we also heard this from nearly everyone we spoke to, including local youth.
Attacks in Garissa have indeed dropped dramatically, even though some still occur. Shabaab cells have been shut down, caches of guns and grenades have been discovered,  and a planned attack on a KDF base was foiled. Progress on security has a ripple effect and as one NGO worker said: “with confidence in security growing, there are less KDF roadblocks and presence”. 
Although there are other ways of interpreting the new developments in Garissa – for example, the killing of Mohamed Mohamud Ali (‘Gamadhere’) in mid-2016  appears also to have diminished Shabaab momentum in Garissa – almost all of those we spoke to attributed improved security in Garissa to Nyumba Kumi and the approach taken under Saleh. This was echoed in a young people’s focus group discussion:
“Nyumba kumi initiative is very effective in Garissa and has drastically reduced criminal and violent attacks, and improved relations between communities and security agencies”. 
The apparent success of the changed approach to security provision in Garissa is important because it illustrates what is possible when there is a break from the ‘hard security’ model for reacting to terror attacks. In many places, governments are supporting ‘Countering Violent Extremism’ (CVE) initiatives to dissuade people from siding with violent groups, but these efforts often sit alongside –and are rendered ineffective by – implacably repressive methods for attacking the problem. No one should idealise the Garissa story or claim the problem is lastingly solved, but the story shows what is possible when authorities make a meaningful attempt not to fight dirty, and to build momentum instead by regaining public trust.
81. Saferworld (2015), ‘Dilemmas of counter-terror, stabilisation and statebuilding’, January, p 13 (https://www.saferworld.org.uk/resources/publications/875)
82. Mixed Focus group discussion, Garissa town, January 2017
83. For example: Quartz Media (2015), ‘Kenya’s Garissa students thought a poster warning of an imminent terrorist attack was an April Fools’ prank’, 5 April (https://qz.com/376831/kenyas-garissa-students-thought-a-poster-warning-of-an-imminent-terrorist-attack-was-an-april-fools-prank/)
84. Zirulnick A (2015), ‘Kenya university attack puts security capabilities under fresh scrutiny’, The Christian Science Monitor, 2 April (http://www.csmonitor.com/World/Africa/2015/0402/Kenya-university-attack-puts-security-capabilities-under-fresh-scrutiny-video); BBC News (2015) ‘Kenya al-Shabab attack: Security questions as dead mourned’, 4 April (http://www.bbc.co.uk/news/world-africa-32177123)
85. BBC News (2015), ‘Kenya al-Shabab attack: Security questions as dead mourned’ (http://www.bbc.co.uk/news/world-africa-32177123)
86. Huffington Post (2015), ‘Kenya Minister Admits Intelligence Warned of Garissa University College Attack’ (http://www.huffingtonpost.com/2015/05/01/kenya-intelligence-garissa-attack_n_7189928.html)
87. Gordts E (2015), ‘Kenya Minister Admits Intelligence Warned of Garissa University College Attack’, Huffington Post, 1 May (http://www.huffingtonpost.com/2015/05/01/kenya-intelligence-garissa-attack_n_7189928.html)
88. For example: Anderson D, McKnight J, ‘Understanding al-Shabaab: clan, Islam and insurgency in Kenya’, p 549 (http://www.tandfonline.com/doi/full/10.1080/17531055.2015.1082254); Quartz Media (2015), ‘Kenya’s Garissa students thought a poster warning of an imminent terrorist attack was an April Fools’ prank’, 5 April (https://qz.com/376831/kenyas-garissa-students-thought-a-poster-warning-of-an-imminent-terrorist-attack-was-an-april-fools-prank/)
89. See: Anderson D, McKnight J, ‘Understanding al-Shabaab: clan, Islam and insurgency in Kenya’, p 549; Quartz Media (2015), ‘Kenya’s Garissa students thought a poster warning of an imminent terrorist attack was an April Fools’ prank’, 5 April (https://qz.com/376831/kenyas-garissa-students-thought-a-poster-warning-of-an-imminent-terrorist-attack-was-an-april-fools-prank/)
90. Mukinda F (2015), ‘Shame of slow response in 15-hour campus terror’, Daily Nation, 4 April (http://www.nation.co.ke/news/Shame-of-slow-response-in-15-hour-campus-terror/-/1056/2676432/-/13wgdjy/-/index.html)
91. Ombati C (2015), ‘Mohamud Ali Saleh appointed new North Eastern regional coordinator’, Standard Digital, 29 April (https://www.standardmedia.co.ke/article/2000160435/mohamud-ali-saleh-appointed-new-north-eastern-regional-coordinator)
92. International Crisis Group (2015), ‘Kenya’s Somali North East: Devolution and Security’, 17 November, p 13 (https://d2071andvip0wj.cloudfront.net/b114-kenya-s-somali-north-east-devolution-and-security.pdf)
93. Interview with member of County Assembly, Garissa town, January 2017
94. Interview with County government official, Garissa town, January 2017
95. Interview with deputy governor, Garissa County, Garissa town, January 2017; International Crisis Group (2015), ‘Kenya’s Somali North East: Devolution and Security’, 17 November, p 1 (https://d2071andvip0wj.cloudfront.net/b114-kenya-s-somali-north-east-devolution-and-security.pdf)
96. Interview with member of County Assembly, Garissa town, January 2017
97. International Crisis Group (2015), ‘Kenya’s Somali North East: Devolution and Security’, 17 November, p 13 (https://d2071andvip0wj.cloudfront.net/b114-kenya-s-somali-north-east-devolution-and-security.pdf)
98. Ibid p 13
99. Interview with senior police officer, Garissa town, January 2017
100. International Crisis Group (2015), ‘Kenya’s Somali North East: Devolution and Security’, 17 November, p 13 (https://d2071andvip0wj.cloudfront.net/b114-kenya-s-somali-north-east-devolution-and-security.pdf)
101. Interview with NGO worker, Garissa town, January 2017
102. Al Jazeera (2016), ‘Al-Shabab confirms death of Garissa attack leader’, 19 June (http://www.aljazeera.com/news/2016/06/al-shabaab-confirms-death-garissa-attack-leader-160619125725418.html)
103. Youth focus group discussion, Garissa town, January 2017
Header photo: Mural at Garissa University College. Photo: © Garissa University College