Analyses of security provision in Africa's cities reflect Western ideals and experience. Although it is increasingly accepted that the divide between formal/informal policing is blurred rather than binary, researchers continue to assume that regime-oriented internal security is separate from community safety. Fieldwork in Mogadishu shows that this is not the case: the boundaries between counter-terrorism and community safety are porous, as are those between security and development, with the balance determined by residents' need for physical security. This paper focuses on Mogadishu's most successful scheme for increasing the community intelligence needed for effective counter-terrorism (CT) while facilitating the social mobilisation and welfare that the city's high-level security plan depends on: the Neighbourhood Watch (NHW) scheme found in Wabeeri district. NHW shows that the boundaries between CT and everyday soft security are porous; the two may be procedurally and financially separate but they are not analytically exclusive. Western value-based divisions between security and development do not necessarily transfer to the Somali environment, but the notion of a continuum based on everyone's need for physical security may. The best way to conceptualise this as an urban process remains challenging.

This paper approaches public authority in Rwanda from below, mapping the actors involved in the prevention of street crime. It progresses through a series of subsections - on house guards, community policing committee members, police community liaison officers, judicial police officers, district security officers, private security guards, RDF military patrols and abanyerondo night patrolmen - looking both at the individual duties associated with these positions and their points of contact in the broader policing system. Each subsection amalgamates information from a range of interviews and informal discussions that are drawn from twenty months of fieldwork between 2013 and 2015.

Empirical material on the various roles in the Rwandan policing network are illustrated through the thick description of four selected criminal incidents that occurred in 2014 and 2015. These descriptions demonstrate the importance of Rwandan policing systems below the cell tier of government administration. They show the diversity of actors working at this level and the range of channels available for information to pass to state authorities. Providing this degree of detail on Rwanda's policing groups, their work and their interactions has not been done before in a systematic manner. It allows not only for the identification of central themes in Rwandan policing, but also for the comparison of nuanced mechanisms of local crime prevention between different urban centres in the country.

Many ethnic groups are politically organised through institutions of traditional governance. These indigenous political systems are based on customary law and organise communities through internal policing and customary judicial proceedings. States deal differently with the resulting parallelism of state and ethnic institutions: whereas some do not regulate the relationship to traditional governance institutions, some recognize the institutions, grant them autonomy rights, or integrate them into the state's institutional infrastructure. The effect of the state's relationship to traditional governance on violence is not yet sufficiently investigated. While state-level research in Africa has shown that unregulated jurisdictional overlap increases the risk of communal violence (Eck 2014), there is also group-level evidence that traditionally organised ethnic groups have a decreased risk of being involved in communal violence. In this paper, I apply new data at the group level, containing geo-referenced information on the existence and capacities of traditional governance institutions, and at the state-level, coding the regulation of the relationship between these institutions and states. I argue that unregulated parallel governance structures are a potential source of violent conflict by producing opportunity structures favourable to violent mobilisation through overlapping jurisdictions. The regulation, however, reduces the likelihood of violence due to an advanced justice and security infrastructure. To test this claim, I empirically examine the relationship by applying multilevel analysis. Thereby I am able to account for variation and significance of traditional governance between states, which offers a more fine-grained understanding of the effect of existing traditional governance on violence.

The relationship between police and non-statutory security actors in the provision of everyday security is an under-researched aspect of policing in Luanda, Angola. Little research has proved that bringing together different agents or organisations with different logics, mechanisms, norms, values, and motivation will lead to more inclusive public safety and security mechanisms. Therefore, this paper focuses on the relationship between the police and non-statutory security actors in the city of Luanda, with the objective to determine in which circumstances non-state security actors complement, accommodate, compete, substitute or undermine state security providers. Specifically, this project examines the objectives of the identified non-state actors, their resources and modus operandi, and the nature of the cooperation with the police. The output of the interactions between the police and the identified non-state actors and the residents' experience will be highlighted. I argue that in less homogenous urban residential areas, in the context of limited statehood and fragility, non-statutory security actors have a less constructive relationship with the police and other statutory security providers.