Download Chapter

Chapter 2.1 Themes and issues



Humanitarian coordination can be defined as ‘the systematic use of policy instruments to deliver humanitarian assistance in a cohesive and effective manner. Such instruments include strategic planning, gathering data and managing information, mobilising resources and ensuring accountability, orchestrating a functional division of labour, negotiating and maintaining a serviceable framework with host political authorities and providing leadership’.+L. Minear et al., UN Coordination of the International Humanitarian Response to the Gulf Crisis 1990–1992 (Providence, RI: Thomas J. Watson Institute for International Studies, 1992) (, cited in N. Reindorp and P. Wiles, Humanitarian Coordination: Lessons from Recent Field Experience, 2001 (, p. ii.

The subject of humanitarian coordination has been widely written about and studied. This section focuses on urban issues associated with coordination. It discusses coordination with city authorities and steps for engaging with them in planning processes. It reviews coordination with emergent (voluntary) groups and gangs, as two examples of particularly urban phenomena. It reviews cluster coordination in urban areas.

2.1.1 Levels of coordination

Recent research+P. Knox Clarke and L. Campbell, Exploring Coordination in Humanitarian Clusters (London: ALNAP, 2015) ( into organisations engaged in emergencies identifies three levels of coordination: communication, where information and knowledge is shared between organisations; alignment, where organisations ‘may adjust their activities to create a more effective response on the basis of the activities of other organisations’, for example ensuring they work in different neighbourhoods; and collaboration, where organisations may have common goals and share activities.

In urban areas there are of course a wide range and diversity of stakeholders (see Section 1.5 on actors), many with competing interests and different degrees of power and legitimacy. While humanitarian coordination is often with national government structures, local government bodies are essential stakeholders. Coordinating with local government may not always be possible, particularly in situations of conflict or immediately after a disaster, but international actors should never assume that local government is not functioning. Instead, they should make working with the city authorities the default, unless this proves impossible for reasons of government capacity or lack of neutrality.

2.1.2 Engaging with city authorities

Local government and its officials are essential actors in urban humanitarian response. Yet the humanitarian community has repeatedly been found wanting in how it engages with city authorities. One study of the response of 13 INGOs following Typhoon Haiyan in 2013 found that, across the track of the typhoon, local government was largely bypassed by aid agencies, which worked directly at the community level.+D. Sanderson and Z. Delica Willison, Philippines Typhoon Haiyan Response Review, 2014 (, p. 15. Following the 2010 Haiti earthquake, ‘local authorities … complained that three months after the earthquake they felt “like strangers in [their] own city”’.+F. Gru?newald et al., Real-time Evaluation of Humanitarian Action Supported by DG ECHO in Haiti, 2009–2011 (Plaisian: Groupe URD, November 2010–April 2011) ( This is a particular problem in urban areas, where meaningful action often relies on complex systems of governance. A wide-ranging study of cities and crises in 2016+IMPACT and UCLG, Consultations on Humanitarian Responses in Urban Areas: Perspectives from Cities in Crisis, World Humanitarian Summit, 2016 ( found the following:

  • Overwhelmed and under-resourced municipalities that also lacked expertise to deal with ‘the exponential needs resulting from crisis’.
  • A failure by international agencies ‘to understand local dynamics and to engage with local stakeholders’.
  • An approach that was sector-based rather than settlement-based.
  • A ‘lack of city-level multi-stakeholder coordination mechanisms’, contributing to ‘the divergence between international and local actors’.

The report found that these problems led to a sub-optimal response (for example bad targeting and duplication), undermined local systems and fuelled community tensions (for example in not consulting with all actors, leading to perceptions of bias). Failing to engage with municipal authorities can also mean that short-term emergency interventions in areas such as shelter and settlement, WASH and public health fail to articulate with – and even set back – longer-term development planning.

Research by ALNAP emphasises that working with government structures is a necessity, not a choice: ‘in many situations there are possibilities to work closely with line ministries or other parts of government, even where the government is engaged in internal conflicts. Where even this is not possible, coordination models should be designed to align with government structures to the degree possible, to allow for government ownership at a later date’.+Knox Clarke and Campbell, Exploring Coordination in Humanitarian Clusters, p. 7.

Humanitarian organisations’ roles may therefore need to shift – ‘depending on the capacity of the local authorities, the humanitarians’ role may be more about facilitation and enabling than direct service provision’.+B. Mountfield, Using the Sphere Standards in Urban Settings (Geneva: The Sphere Project, 2016) (, p. 11. However, engagement with authorities can often be sensitive, particularly in conflict situations. Where authorities are partisan or may be the aggressors, engagement may not be possible, as is evident in the example from Mogadishu in Box 2.1.

Box 2.1 Relations between humanitarian actors and municipal authorities in Mogadishu

‘One of the main conceptual difficulties for aid actors has been to approach Mogadishu as a city and not as a classic humanitarian situation. This requires strategic sector-based coordination linked to administrative units rather than cluster-type sector-based coordination, and, above all, an attempt to engage with urban authorities. While some of the NGOs working in Mogadishu have tried to establish Memoranda of Understanding with the Ministry of Health, they have bypassed the municipal level and gone down to the district commissioner level, which is responsible only for law-and-order control functions rather than urban planning. The reasons why there has not been any engagement with municipal authorities include the fear of politicisation, the risk of corruption; and, more broadly, ignorance about their roles, if not reluctance to work with these urban actors.’

Source: F. Grünewald, ‘Aid in a City at War: The Case of Mogadishu, Somalia’, Disasters 36(1), July 2012 (

 2.1.3 Engaging with local authorities

From its work in urban areas,+IRC, From Response to Resilience. Working with Cities and City Plans to Address Urban Displacement: Lessons from Amman and Kampala (New York: IRC, 2018) (, p. 37. IRC identifies the following key steps for establishing urban partnerships:

  • Begin a dialogue with local municipal authorities to identify opportunities for collaboration around shared outcomes.
  • Determine whether a master plan exists, and other relevant documentation and policies.
  • Identify the interests and incentives of all stakeholders (including the authorities).
  • Ensure that all actors are operating based on the same information, including where necessary information on humanitarian approaches.
  • Ensure as far as possible effective coordination between all stakeholders, including local authorities, NGOs, faith-based organisations and the private sector.
  • Link all recovery actions to long-term developmental approaches, strategies and goals.

Another collaborative planning approach developed by ACTED and IMPACT is AGORA, which aims to provide ‘predictable capacity to localise aid action and promote efficient, inclusive and integrated local planning and service delivery in contexts of crisis’.+See AGORA uses a settlements-based approach (see Section 3.2 on area-based approaches). For further information see The ‘local generalist approach’ seeks to support city authorities by providing financial and technical assistance. For further information, see

Box 2.2 Using area-based approaches to improve coordination in Mafraq, Jordan

The Community Engagement Project, which works in Mafraq in Jordan and in a number of other countries, seeks to improve coordination between local government, NGOs and community-based organisations in efforts to build social cohesion. The vehicle for this has been through enacting an area-based approach (see Section 3.2 for further discussion). Joint activities between actors and communities have included assessments and stakeholder mapping of key capacities.

For further discussion of this initiative, and for additional examples from Bangui, Bogo, Gaziantep, Port-au-Prince and Tacloban, see IMPACT and UCLG, Consultations on Humanitarian Responses in Urban Areas: Perspectives from Cities in Crisis, World Humanitarian Summit, 2016 (

2.1.4 Engaging with emergent (voluntary) groups

While not exclusively an urban phenomenon, the scope and range of emergent groups (introduced and discussed in Section 1.5 on actors) in cities is particularly noticeable. However, aid organisations can find it difficult to meaningfully engage with these groups. According to one study: ‘the core problem is that emergency planners and plans rarely take emergent groups and spontaneous volunteering into account. They do not understand the nature and characteristics of emergence or the strong motivations behind it. Emergence is an implicit challenge to the “command and control” approach of most official disaster management and emergency response agencies, with their top-down bureaucratic systems and standard operating procedures’.+J. Twigg and I. Mosel, ‘Emergent Groups and Spontaneous Volunteers in Urban Disaster Response’, Environment and Urbanization 29(2), 2017 (, p. 452. The study also gives an example of more successful coordination: ‘in the Kobe earthquake, initial problems regarding duplication of effort were overcome by creating an umbrella group, the Nishinomiya Volunteer Network, to coordinate the work of emergent groups and collaborate with the government on distributing food and other goods, collecting information about survivors’ needs from temporary shelters, and liaising between survivors and government’.+Ibid., p. 454.

2.1.5 Engaging with gangs

In urban areas, gangs or criminal elements will control some neighbourhoods, raising questions about how or whether humanitarian organisations should engage with them. ALNAP observes that ‘Any work carried out by humanitarian players in a city neighbourhood, or in an area within a prison, that is controlled by a gang will be subject to discussion or authorisation by the gang, whether one is aware of it or not’. One programming consideration here is how far to engage with gang leaders.+D. Sanderson and P. Knox Clarke, Responding to Urban Disasters: Learning from Previous Relief and Recovery Operations (London: ALNAP, 2012) (, p. 19. ALNAP’s study cites research on the ICRC’s experience,+E. Ferris, Urban Disasters, Conflict and Violence: Implications for Humanitarian Work (Washington DC: Brookings Institution, 2012) ( and notes ‘the need to gain acceptance from gang leaders, given that many gangs are hierarchically organised; also that gang leaders need to perceive some benefit from the organisation’s operation’.+Sanderson and Knox Clarke, Responding to Urban Disasters, p. 19. See also Section 1.2.2 on urban violence.

2.1.6 The cluster approach and urban response

The cluster approach, introduced into humanitarian action as part of the 2005 Humanitarian Reform Agenda, is the internationally-accepted norm for organising aid support according to sectors, or clusters. The Inter-Agency Standing Committee (IASC)+IASC, Reference Module for Cluster Coordination at the Country Level, 2012, cited in Knox Clarke and Campbell, Exploring Coordination in Humanitarian Clusters, p. 20. describes key cluster functions as supporting service delivery, informing strategic decision-making, planning and strategy development, advocacy, monitoring and reporting on the implementation of the cluster strategy, contingency planning, preparedness and capacity-building, and integrating early recovery from the outset of the humanitarian response. ALNAP’s research into organisations engaged in emergencies (referred to above) found that ‘overwhelmingly, Cluster activities fall at the “alignment” level’. This is especially true where competing interests, stresses and pressures call on different organisations to make careful calculations on where and how to align their efforts. Other factors include workflows, policies and institutional cultures; the capacity and willingness of local government and local civil society to be involved and/or lead coordination; and what level of coordination is being attempted. That said, coordination in disasters and crises is an essential element of any successful intervention. While the cluster system overall has improved coordination and information-sharing,+See for example V. Humphries, Improving Humanitarian Coordination: Common Challenges and Lessons Learned from the Cluster Approach (Medford, MA: Tufts University, 2013) ( See also Knox Clarke and Campbell, Exploring Coordination in Humanitarian Clusters. it has been problematic in urban areas. As the IRC observes: ‘The traditional cluster system does not lend itself to the complexity of needs, services and systems across an urban landscape with humanitarian agencies struggling to deal with the complexity, density and built environment of towns and cities or [un]able to take full advantage of the potential a city has to offer’.+IRC, Humanitarian Crises in Urban Areas: Are Area-Based Approaches to Programming and Coordination the Way Forward? (New York: IRC, 2015) (, p. 5. The IASC has challenged the very validity of the cluster approach in urban recovery operations: ‘the current cluster system is structured around sectors of expertise and sectorial coordination, while in a context of urban crises there might be a need to identify and respond holistically to multi-sectorial needs in a given territory, requiring stronger inter-cluster linkages and coordination at city-level’.+IASC, Guidance Note for Improving Coordination and Responses to Urban Crises in the Humanitarian Programme Cycle through the IASC and its Cluster System, IASC Working Document, 2016, p. 1. For clusters to work well in urban areas, the following should be noted:

  • Leadership by local government is important, to avoid creating parallel structures which remove ownership from existing governance structures.
  • Cross-sectoral coordination is vital: single-sector interventions that do not recognise urban complexity risk causing harm (a systems perspective on how cities work, described in Section 1.1, is a useful starting-point here).
  • Meeting locations and the frequency of meetings in cities are important. Traffic jams in congested areas can waste time and cluster members might not bother to attend meetings. In Port-au-Prince after the 2010 earthquake, chronic traffic jams and the location of cluster meetings on the outskirts of the city meant that several hours a day could be spent in a vehicle.+C. Clermont et al., Urban Disasters: Lessons from Haiti. Study of Member Agencies’ Responses to the Earthquake in Port au Prince, Haiti, January 2010 (London: DEC, 2011) (
  • Many NGOs may find it difficult to attend a large number of meetings, especially if an organisation is involved in more than one sector.
  • Language is an important consideration – default English in a non-English- speaking location alienates local bodies and discourages engagement.
  • Many local stakeholders may well not attend cluster meetings, for instance private sector organisations or gangs. Other ways of engaging them are therefore needed.

Concerning cross-sector coordination, the IASC+IASC, Guidance Note, p. 3. has discussed adopting a city-level intercluster working group (led by city authorities wherever possible) that ‘would support stronger coordination among sectors and with local actors’. An important opportunity for cross-sectoral coordination lies in conducting multi-sectoral assessments. This is discussed further in Section 3.6 on assessments.