Vanessa Humphries is the Project Manager at the Security Governance Group. She holds a Master’s of International Public Policy and was a Humanitarian Affairs Intern with the United Nations Office for the Coordination of Humanitarian Affairs, Regional Office for the Pacific. She presented this paper at the Academic Council of the United Nations System Annual Meeting in June 2012.


The author would like to thank Dr. Ken Jackson (Wilfrid Laurier University) for his mentorship and support in constructing a methodology.


In the past decade the humanitarian system has had to respond to natural disasters and complex emergencies of increasing severity.  In 2005, as an attempt to increase coordination amongst humanitarian actors and improve coherence in humanitarian response, the United Nations implemented a coordination mechanism called the Cluster Approach. The aim of this paper is to present common challenges of the Cluster Approach raised since its implementation and to provide lessons learned, based on the findings of a meta-analysis of 18 existing case studies, evaluations, and literature.  The paper assesses progress the Cluster Approach has made toward meeting its intended goals, exposing different stakeholder perspectives and aggregating findings from various clusters and country contexts.

Screen shot 2013-04-29 at 12.11.13 PM

The research reveals that, overall, the Cluster Approach has increased the effectiveness of humanitarian action, suggesting that it is a worthwhile mechanism to pursue. However, there are many challenges associated with the approach. First, there are large gaps in predictable leadership.  This is primarily due to the high turnover rates of cluster coordinators, lack of impartiality of cluster lead agencies, and insufficient training and experience of cluster coordinators. Second, there are barriers to inclusive partnership in the Cluster Approach.  Cluster coordination is not only labour intensive, requiring a significant amount of time and resources for effective participation, but it has largely failed to create a sense of NGO ownership and involvement.  Third, the Cluster Approach does not have sufficient mechanisms in place to enhance accountability to affected populations.

The key findings of the meta-analysis highlight that there are many opportunities for improvement within the Cluster Approach, but that the structure of the coordination mechanism is a positive shift in humanitarian relief efforts.


In the past decade the humanitarian relief system has responded to over a thousand natural disasters and complex emergencies in the world, affecting hundreds of millions of people (Gilmann 2010: 21). Extreme weather and climate events have increased in both frequency and intensity, placing populations and assets at great risk (IPCC 2012). On top of the growing severity of natural disasters, there are increasing numbers of internally displaced persons, refugees, and asylum seekers due to war or internal conflicts (Gilmann 2010: 21).  In 2011 alone, natural disasters and political upheaval threatened tens of millions of lives in the developing and developed worlds (OCHA 2012). In response to this growing need, the humanitarian system has evolved into an industry, with a plethora of organizations, all with different missions, mandates, and agendas.  With the increase of humanitarian actors, the relief system has met a series of challenges, including the need to both increase resources toward humanitarian ends and to improve operational effectiveness and efficiency (Gilmann 2010: 21).  Despite efforts to confront these challenges, much criticism has been leveled at the humanitarian system for failing to meet the basic requirements of affected populations in a timely manner, with the quality of response varying greatly from crisis to crisis (Stumpenhorst, et al. 2011: 587; Adlinofi et al. 2005: 8).

The Cluster Approach was implemented by the United Nations to address some of these concerns and to improve the coordination of humanitarian relief and actors.  Coordinating relief efforts entails minimizing the duplication of humanitarian services, whether by filling gaps or preventing overlap, and ensuring various organizations are synchronized to work together to achieve a common objective, thereby enabling a more coherent, effective, and efficient response (Gillmann 2010: 326; James 2008: 351-2). Although the need for coordination in relief efforts is not disputed, there are generally two schools of thought on how coordination is best executed in humanitarian relief.  The first group is driven by governmental and inter-governmental bodies, and places an emphasis on a centralized, unified, hierarchical structure, which is assumed to be more effective and efficient (James 1008: 351).  The second group, preferred by NGOs, is based on a loose centralized approach to coordination.  This group tends to regard the centralization as a means of control over actors and focuses on how a diversity of efforts and approaches can ensure success: if one fails, all do not fail (James 2008: 351).  Proponents of this approach believe that NGOs have to protect themselves, to preserve their humanitarian principles, by ensuring that they are being coordinated with, rather than  for (Currion & Hedlund 2010a: 4). With a hierarchical structure featuring accountable lead agencies and encouraging equal-footed partnerships and collaboration, the Cluster Approach can be understood as an attempt to find a ‘middle-ground’ between the two schools. [1] Although the Cluster Approach is not the only means of coordinating humanitarian assistance, examining the strengths and weaknesses of alternative coordination architectures is beyond the scope of this paper. [2]

The aim of this paper is to present common challenges to the Cluster Approach and to provide lessons learned since its implementation in 2005. It is based on the findings of a meta-analysis of 18 existing evaluations that assess whether the Cluster Approach is meeting its intended goals.  The paper begins by offering background on humanitarian reforms, humanitarian relief coordination, and the Cluster Approach.  Next, the paper provides an overview of the methodology used to assess the strengths and weaknesses of the Cluster Approach.  Third, the paper explains the key findings and lessons learned from this study. While the findings highlight that there are many opportunities for improving the existing structure of relief coordination, the paper concludes that the Cluster Approach provides the appropriate structure for coordination of actors and, overall, is a positive shift in humanitarian relief.


Screen shot 2013-04-29 at 12.18.33 PM

Given the widespread perception that the humanitarian response during the Darfur crisis was “woefully inadequate,” the United Nations Emergency Relief Coordinator (ERC) and the Under-Secretary General for Humanitarian Affairs  launched the Inter-agency Real-Time Evaluation of the Humanitarian Response to the Darfur Crisis[3] to identify what future lessons could be learned from the 2004 Darfur crisis (Broughton & Maguire 2006: 1).  Overall, the study found that the crisis in Darfur was delayed and inadequate, primarily due to the inability of agencies to mobilize capacity and resources (Gillmann 2010: 63).   One of the key findings of the report was its recognition that a directive approach to coordination was not feasible: “Agencies cannot be forced to work within the parameters of a common plan—ultimately sector leads must persuade the majority of the value of a cohesive approach” (Broughton & Maguire 2006: 63).   Additionally, the report recognized the value of a stronger field presence, which would have allowed the United Nations to better “shape the response” (Broughton & Maguire 2006: 5).  There were also significant gaps found in coordination, information sharing, interagency coordination, strategic leadership, and cross-sectoral issues such as livelihoods, gender, and protection.  This study reinforced the importance and urgency of improving relief response in the international humanitarian system.

In response to the heightened public awareness of disasters following the 2004/5 Indian Ocean tsunami and the frustrations surrounding the inadequate response to Darfur, the ERC commissioned the Humanitarian Response Review (HRR), with the goal to improve the effectiveness and timeliness of humanitarian response to emergencies (HRP 2009: 9; Adinolfi et al. 2005: 9).  The HRR assessed humanitarian response capacities of the UN, NGOs, Red Cross/Crescent Movement, and other key humanitarian actors and recommended ways to mitigate shortcomings in capacity and response (Gillman 2010: 62; Maxwell & Parker 2011: 26).

Some of the findings from the HRR were then translated into a series of humanitarian reforms, implemented by the United Nations in 2005.  The reforms are based on four pillars: (1) Improve humanitarian leadership (through the Humanitarian Coordinators system); (2) Better coordination of humanitarian action (through the Cluster Approach); (3) Promote faster, more predictable and equitable funding (through improved humanitarian financing, such as the Central Emergency Response Fund (CERF)); and (4) More effective partnerships among all humanitarian actors (through the Principles of Partnership implemented in 2007) (HRP 2009: 9).

The focus of this paper is on the ‘second’ pillar of humanitarian reform, improving the coordination of humanitarian actors through encouraging a coherent response to emergencies.  In December 2005, based on the findings from the HRR, the Inter-Agency Standing Committee (IASC) established the Cluster Approach as a mechanism to improve the efficiency and effectiveness of coordination and thereby save lives and reduce suffering. The Cluster Approach was established to address gaps in humanitarian response and to improve the predictability, accountability, and effectiveness of relief efforts through a more coordinated humanitarian relief response (IASC 2006:1).  Its aim is to enhance partnerships and complementarity among United Nations agencies, the Red Cross movement, international organizations and NGOs at both global and country levels  (James 2008: 359; IASC 2006: 1). The concept of a ‘cluster’ implies a group of organizations coalescing around a common sector of humanitarian relief and ‘approach’ implies that the organizations work together in a spirit of partnership and inclusivity (Shepherd-Barron 2009: 6).

The Cluster Approach assigns a lead agency to organize coordination, strengthen preparedness, and to act as the provider of last resort for each major response area.  Clusters were introduced for nine response and two service areas, with four crosscutting issues (see Figure 1). Lead agencies are usually large UN agencies, such as the High Commissioner for Refugees or the World Food Program, capable of facilitating the coordination of all humanitarian actors, including local and national NGOs.  Clusters are permanently established at the global level and on an ad-hoc bases during times of emergencies at the country level.

Screen shot 2013-04-29 at 12.20.57 PM


At the global level, the aim is to strengthen preparedness and technical capacity of response through designating global cluster leads and ensuring predictable leadership and accountability for each sector of activity (IASC 2006: 13). Global clusters convene to develop surge capacities, technical guidance, training, stocks, tools and operational support and have two main goals: (1) to provide systematic, timely and predictable support to country-level clusters and (2) to ensure predictable leadership and accountability (Maxwell & Parker 2012: 26; Steets et al. 2010: 24).

Clusters are established at the country-level during times of emergencies.  The aim is to ensure a more coherent and effective response through mobilizing humanitarian actors to respond in a strategic way across all sectors, with each sector having a designated lead agency (IASC 2006: 13).  Country level clusters most often mirror the lead agency arrangements at the global level and meet regularly to share information, create cluster strategies and work plans, contribute to the preparation of funding appeals, share information, or organize joint activities (Steets et al. 2010: 24-25).

Currently, the Cluster Approach is implemented in 27 of the 29 countries[4] that are experiencing humanitarian emergencies (One Response 2012). Implementation has required significant investments in time, money, and energy on behalf of all member organizations, with cluster lead organizations often contributing additional funds from their own budgets.  From 2006 to 2008, over 57 million USD was raised through global appeals to implement the Cluster Approach, in addition to the annual coordination costs in each country at around several million dollars per annum (Steets et al. 2010: 8).  As the main instrument of the Office for the Coordination of Humanitarian Affairs (OCHA), the Cluster Approach is the prevailing strategy for humanitarian emergency response (Atalay & Labonte 2011: 89).


In order to assess the overall strengths and weaknesses of the Cluster Approach, this study is based on  a meta-analysis of existing case studies and literature that evaluate the Cluster Approach.  A meta-analysis was used in order to summarize and interpret key findings from recent reports, case studies, and evaluations.  The aim is to present common challenges and useful lessons learned from the Cluster Approach since its implementation and to determine if the Cluster Approach has been successful in terms of its intended goals.

The meta-analysis analyzes a total of 18 evaluations, which were selected to include all major evaluations completed to date,[5] case studies in natural disasters and complex emergencies, response and service clusters, and a variety of stakeholder perspectives. The evaluations assessed are based on a variety of country case studies, cluster-specific assessment reports, global and regional interviews, surveys, and existing literature pertaining to the effectiveness of the Cluster Approach (see Appendix 1 for the full list of evaluations surveyed).

In order to measure the relative strengths and weaknesses of the Cluster Approach through literature that evaluates the reforms from varying logics and methodology, the evaluations were standardized according to the four main intended outcomes of the Cluster Approach, as defined by IASC.  They are:

1. Overall effectiveness at improving humanitarian response,

2. Creating predicable leadership,

3. Enhancing partnership between humanitarian actors, and

4. Increasing the accountability of relief efforts (IASC 2006: 2).

To find commonalities and irregularities throughout the evaluations, sub-categories were designated as dimensions of analysis.  Each category was given a rating on a scale from one to five, with one indicating a weak ranking and five indicating the strongest ranking.

Category Assessed Dimensions of Analysis
To improve overall effectiveness in humanitarian response. Has the Cluster Approach improved overall humanitarian action? Are there positive benefits for humanitarian relief and actors? Does the Cluster Approach have the potential to increase the effectiveness of humanitarian action?
To create predictable leadership. Does the Cluster Approach improve coordination of humanitarian actors (both inter-cluster, intra-cluster)? Is there information sharing across cluster members? Do cluster lead agencies carry out tasked responsibilities? Do cluster lead agencies act as providers of last resort?
To enhance partnership between humanitarian actors. Are there strong partnerships between UN agencies, with international, national, and local NGOs, and with government departments (when applicable)? Are the partnerships based on the Principles of Partnership (2007)?
To increase the accountability of relief efforts. Is there coherence in the application of global standards of aid? Are appropriate monitoring and evaluation of cluster plans carried out? Has the Cluster Approach included downward accountability to affected populations?

The findings from the meta-analysis are not necessarily transferable to all situations, as different clusters, countries, contexts, or types of emergencies present case-specific challenges.  In addition, as noted in Steets et al. (2010), it is difficult to attribute changes in coverage to the introduction of the Cluster Approach (54).  This is due to the inability to construct alternative, counterfactual scenarios for the cases examined, no possibility for control groups, and unclear causal links (Steets et al. 2010: 54).  In addition, constraints of this research are similar to any desk study, as interviews were not done first hand and the inability to offer first hand knowledge of the field.  However, this approach is useful in order to have an aggregate analysis of the Cluster Approach to date, reflecting general trends in its successes or failures, and exposing different stakeholder perspectives in the humanitarian field.


The graphs featured below provide a snapshot of the findings from the meta-analysis. The vertical axes indicate the one to five rating scale and the horizontal axes indicate the number of evaluations assessed. Overall, effectiveness reflects that the majority of evaluations find the Cluster Approach to be an effective instrument in coordinating humanitarian relief, as indicated by its strong ratings (in the range of three to five).  However, as reflected in the other categories by ratings heavily weighted below a score of three, the Cluster Approach continues to face challenges in areas of predictable leadership, partnership, and accountability. Despite these challenges, various stakeholders find the cluster approach, in terms of its overall effectiveness, to be a useful coordination mechanism to improve relief efforts.

Screen shot 2013-04-29 at 1.19.31 PM   Screen shot 2013-04-29 at 1.20.56 PM

Screen shot 2013-04-29 at 1.23.00 PM   Screen shot 2013-04-29 at 1.23.37 PM

1. Has the Cluster Approach improved overall effectiveness of humanitarian response?

Category Assessed Dimensions of Analysis
To improve overall effectiveness in humanitarian response. Has the Cluster Approach improved overall humanitarian action? Are there positive benefits for humanitarian relief and actors? Does the Cluster Approach have the potential to increase the effectiveness of humanitarian action?

The vast majority of evaluations assessed found that the Cluster Approach is an effective tool and has the appropriate structure for improving the coordination of humanitarian relief. This reflects the commonly accepted perspective that the Cluster Approach has brought a positive change and is worth continuing to invest in as a means for humanitarian relief coordination.

The below are recurring examples of how the Cluster Approach has improved the overall effectiveness in humanitarian response:

  • Although no formal cost-benefit analysis of the Cluster Approach has been done, the evaluations generally found that the costs of coordination are justified and outweighed by their benefits (Steets et al. 2010: 10; Stoddard et al. 2007: 1; Featherstone 2010: 19).
  • There is optimism about the future of the Cluster Approach as, since its implementation, it has been seen as improving and growing based on lessons learned over the past six years (Charles et al. 2010: 163).
  • Clusters are most effective in sectors with long-standing relationships, prior to cluster implementation, such as in Kenya (IRIN 2008b) and the WASH cluster in the Democratic Republic of the Congo, Ethiopia, and Zimbabwe (HRP 2009: 25).
  • In general, overall response was found to be successful in key areas such as food distribution, nutrition, health, and water and sanitation (Polastro 2012: 18; Steets et al. 2010: 8).
  • It was also seen to improve coverage of humanitarian needs, address gaps and avoid duplications of assistance, and increase learning from crisis to crisis (Steets et al. 2010: 50)

However, although overall effectiveness of humanitarian relief has improved, there are a few recurring challenges:

First, the Cluster Approach’s existing coordination mechanisms reached their limits in situations of large-scale, complex crises.  This was particularly apparent during the Haitian earthquake in 2010 where there was a plethora of international, national and local humanitarian actors, all with varying levels of skills, experience, and capacity, during which the cluster’s coordination mechanism was stretched beyond its capacity (Stumpenhorst, Stumpenhorst & Razum 2011; Maxwell & Parker 2012).

Second, there is discrepancy on whether the Cluster Approach has improved overall effectiveness from the perceptions of local and national NGOs.  For example, the Humanitarian Reform Project (HRP) (2009) evaluation states that NGOs are still unsure whether the clusters are effective or efficient (23).  In contrast, the International Council of Voluntary Agencies (ICVA) evaluation states that most respondents express their support of the cluster system (Currion & Hedland 2010: 5).  In general, it seems that although local and national NGOs express concerns about its implementation, there seems to be optimism about future inclusive approaches to the Cluster Approach.  The ICVA evaluation explained that, while NGOs were initially reluctant to support its implementation, with increasing partnership, they are becoming more receptive to it (Currion & Hedland 2010a: 5).

Third, some NGOs felt that the Cluster Approach impedes the humanitarian principles of independence, impartiality, and neutrality, which are central to their NGOs ability to reach those in need and to ensure the security of humanitarian workers.[6] When NGOs are perceived as pursuing political goals through politicized or regionally based assistance, it fuels a negative perception of humanitarian workers and breaks down humanitarian principles.  This can occur in two ways:

i) NGOs are often financially dependent on cluster funding or lead agencies, which reduces their independence.  This is because the organizations involved in clusters have access to humanitarian donors, and are often the organizations that support national or international NGOs in emergencies.  When participating in the clusters is no longer voluntary, but rather necessary for funding, it hampers an NGO’s independence.  The ICRC and Médicins Sans Frontières are not formally part of the Cluster Approach for fear that being too closely aligned with the UN, which may jeopardize their independence, though they actively share information about relevant activities (IRIN 2008). In addition the Central Emergency Response Fund (CERF), with a grant element of up to 500 million USD (Gillman 2010: 323) and stemming from the UN’s ‘third’ pillar on more equitable, faster, and predictable humanitarian funding, has been used in an increasingly politicized manner–serving political interests instead of prioritizing immediate needs (Derdarian et al. 2008: 37).  For example, three-quarters of CERF installments in 2006 focused on infrastructure and rehabilitation projects, which were structural, long-term, and high-visibility projects that fulfilled a security objective more than a humanitarian objective (Derdarian et al. 2008: 37).

ii) When clusters or their lead agencies have close links with integrated missions, peacekeeping operations, and governments that are parties to the conflict, it threatens humanitarian principles of impartiality (Steets et al. 2010: 64). For example, the High Commissioner for Human Rights serves as the head of human rights components in peace missions while also participating in the protection cluster.  This means that the High Commissioner can pass through information relating to protection issues to actors involved in the conflict, something that removes the neutrality of humanitarian actors (Steets et al. 2010: 64).  This has made some NGOs question the effectiveness of the Cluster Approach, especially in politicized contexts such as Chad, Somalia or Afghanistan (Steets et al. 2010: 64).

However, there is diversity within coordination structures, which allows for more flexibility to adapt to local situations (HRP 2009: 29).  This flexibility can help NGOs ensure their independence from the UN during complex emergencies, such as conflict and post-conflict situations and is one way to counter the concerns over humanitarian principles being threatened due to NGO partnerships with UN agencies. In addition, Polastro (2012) has explained that in order to avoid eroding humanitarian principles, cluster lead agencies should make sure that they only fund activities that are aligned with the “do no harm” principle, which puts a focus on securing access and creating a humanitarian space, protecting civilians, and ensuring impartiality (20).

2. Has the Cluster Approach created predictable leadership in relief efforts?

Category Assessed Dimensions of Analysis
To create predictable leadership. Does the Cluster Approach improve coordination of humanitarian actors (both inter-cluster, intra-cluster)? Is there information sharing across cluster members? Do cluster lead agencies carry out tasked responsibilities? Do cluster lead agencies act as providers of last resort?

Overall, the evaluations show that there are significant gaps in predictable leadership in humanitarian relief, with a few exceptions.  Predictable leadership involves assessing the role of cluster lead agencies and examining the effects of the introduced coordination mechanisms.

The majority of cluster lead agencies do not provide predictable leadership.  There are four recurring factors for this:

  • There are high turnover rates of cluster coordinators, which contributes to low institutional memory (Stumpenhorst, Stumpenhorst & Razum 2011: 591; Maxwell & Parker 2012: 33; Steets et al. 2010: 65);
  • Cluster leads often have responsibilities tied to their relative agency, which gives the impression that cluster coordinators tend to prioritize their own agency’s projects for funding allocations, rather than prioritizing funds on the basis of need (Lattimer & Berther 2010: 18; HRP 2009: 35);
  • Cluster coordinators are often inexperienced, with no deployment background and with little training on facilitation and coordination  (Doull 2011: 4; Stumpenhorst, Stumpenhorst & Razum 2011: 590; Steets et al. 2010: 9);
  • Many cluster lead agencies have inadequate surge capacity and are unable to deploy adequate numbers of skilled or experienced personnel for coordination roles (Lattimer & Berther 2010: 19; Maxwell & Parker 2012: 30).

In addition, there is a lack of coordination between clusters, or inter-cluster coordination.  In most country-contexts, the Office of the Coordination of Humanitarian Affairs (OCHA) is responsible for providing the infrastructure and framework for inter-cluster coordination; however, this area of response was repeatedly criticized as being inadequate.  In most cases, OCHA was successful at information sharing, but much weaker in identifying multidisciplinary issues, strengthening coordination between clusters, and following-up on identified issues for improvement.   (Homes 2010: 2; Steets et al. 2010: 35).

However, when there is strong leadership from the cluster lead agency, clusters are better managed and have a strategic, action-oriented focus (HRP 2009: 24).  This often occurs when the cluster lead agency dedicates a full-time and experienced cluster coordinator (HRP 2009: 24). The strongest leadership is found in the service clusters: emergency telecommunication and logistics.  For example, the logistics cluster lead, the World Food Program, set-up a ‘one stop shop’ for humanitarian logistics, providing up-to-the minute information on stockpiles, road conditions, customs information, and logistics capacity assessments (Atlay & Labonte 2011: 94). This brought enhanced technical coordination across the cluster by identifying which humanitarian organizations were best-equipped to provide particular services (Atlay & Labonte 2011: 86).

There has also been more predictable leadership in technical aspects of coordination such as adopting international standards of assistance (known as Sphere standards) to meet local contexts.  For example, the international standard of how suggested liters of water per person per day was infeasible to implement in Chad. In response, aid agencies agreed on a new and sufficient standard to ensure that they were consistent in aid deliveries across the region (IRIN 2008a).  This was also true in Uganda, Haiti, and Myanmar, where clusters implemented global minimum standards of relief to meet local contexts and resources, such as agreeing on a common composition of food parcels or standardizing agricultural starter kits (Steets et al. 2010: 51).

3. Has the Cluster Approach enhanced partnership between humanitarian actors?

Categories Assessed Dimensions of Analysis
To enhance partnership between humanitarian actors. Are there strong partnerships between UN agencies, with international, national, and local NGOs, and with government departments (when applicable)? Are the partnerships based on the Principles of Partnership (2007)?

The Cluster Approach is meant to enhance partnership between UN agencies, with international, national, and local NGOs, and with government departments (James 2008: 359; IASC 2006: 1). According to the IASC, successful partnerships depend on “all humanitarian actors working as equal partners in all aspects of humanitarian response: from assessment, analysis and planning, to implementation, resource mobilization and evaluation” (IASC 2007: 8).  In addition, based off the Principles of Partnership endorsed by the Global Humanitarian Platform in July 2007, partnerships should be consultative and transparent, sharing power with all cluster members.  The Principles of Partnership outline goals and methods to: enhance equality and mutual respect; promote transparency and dialogue on an equal footing; encourage a result-oriented approach that entails reality-based and action-oriented results; advance responsibility and ethical obligations; and ensures complementarity of local and international capacities (GHP 2007).  Partnership was rated very weak amongst the evaluations, with the exception of a few outliers.  Specifically the Cluster Approach has been weak in fostering partnerships between UN agencies, national NGOs and local NGOs.

Clusters often fail to incorporate the perspectives of national or local NGO actors.  NGOs often feel their role in the clusters is passive, where they take direction from the cluster lead agency rather than using the clusters as a forum to influence others and/or build capacity to support the overall humanitarian response effort (Maxwell & Parker 2012: 36; HRP 2009: 27).  For example, in September 2009 in the Eastern Democratic Republic of the Congo, 49 representatives of Congolese NGOs, civil society organizations and the Red Cross/Crescent wrote an open letter to UN agencies criticizing them for “failing to coordinate with or support civil society organizations” (Street 2011: 43).  The letter noted that, with few exceptions, “internationals were brought in to staff virtually all posts with the exception of menial jobs”, and that UN agencies failed to ensure capacity building for Congolese organizations (Street 2011: 44).  NGO perceptions of the Cluster Approach are often suspicious, criticizing the approach as being UN-centric, treating NGOs as subordinates, and feeling that it was introduced without adequate consultation with non-UN actors (HRP 2009: 23; IRIN 2008a; IRIN 2008b; Steets et al. 2010: 43). Cluster lead responsibilities have also steered some cluster leads toward an authoritative approach to coordination.  For example, an NGO aid worker in Chad explained the Cluster Approach as “nothing more than a way for the UN to control us” (IRIN 2008a). Another example is after the earthquake in Haiti, when national actors found themselves sidelined by international actors.  Haitian NGO’s were frequently relegated to the role of implementing partners, excluding them from clusters and coordination mechanisms and diminishing their ability to contribute to conceptual issues (Street 2011: 44).
There are also unresolved differences among organizations in certain approaches to intervention.  For example, in the Democratic Republic of the Congo and Chad, there is conflict between the status-based approach to humanitarian assistance used by UN High Commissioner for Refugees for IDPs and the needs-based approach used by most other humanitarian actors (Steets et al. 2010: 51). Such conflicts do not allow for common strategies to be developed or adapted in local circumstances, and to date, there has been no mechanism to resolve these types of conflicts.

Another barrier to effective NGO partnership is the way that cluster meetings are run:

First, clusters demand a great amount of staff time for participants: not only in attending meetings, but also to interpret and deal with the information flow provided by the clusters.  With limited capacity, personnel, and time, national and local NGOs find it difficult to fully engage with the cluster, and cite ‘coordination overload’ as a large problem.  In one case study from Kenya, NGO staff complained that too many meetings were held, with an average of 11 meetings held per week, sometimes lasting up to three hours each (IRIN 2008b).

Second, cluster meetings do not frequently operate in local languages and are often held in capital cities where UN agencies are based. This renders such meetings inaccessible to many local NGOs, especially during times of emergencies when roads may be blocked. (Stoddard et al. 2007: 16; Stumpenhorst, Stumpenhorst & Razum 2011; Maxwell & Parker 2012: 28; Homes 2010: 2).  For example, in Mozambique, Pakistan, Sudan, Afghanistan, and Zimbabwe, local languages were entirely overlooked making it difficult for local NGOs to participate in cluster meetings (Stoddard et al. 2007: 16).  However, there have been some positive examples. One such example is from the education cluster in Haiti that was operated entirely in French to meet local needs (Lattimer & Berther 2010: 8).

Third, a common complaint by NGOs was that meetings were overly focused on process issues, such as responding to requests of OCHA and finance mechanisms, reducing the time and ability of actors to concentrate on distributing aid on the ground (Stoddard et al. 2010: 34). For example, in 2008 a representative from the Red Cross in Chad complained that cluster meetings frequently forced him to question whether it would be “better to attend the meetings or actually do our jobs?” (IRIN 2008a). 

Fourth, cluster lead agencies wield significant power in inviting participants to cluster meetings (Derderian et al. 2008: 35).  This makes the inclusion of national NGOs and local NGOs unsystematic and varied across clusters and countries (Charles et al. 2010: 163).  Without properly integrating national and local actors, national concerns and capacities may be undermined. In several cases this resulted in weakened national and local ownership (Steets et al. 2010: 9; Homes 2010: 2).

Without fostering an inclusive partnership in the early days of an emergency, the cluster approach is inadequate, resulting in in gaps in coordinating mechanisms, resources, knowledge, and assets to expand coverage (Knudsen 2011: 7).

Despite these challenges, there are a few outliers that illustrate positive examples of partnerships with NGOs. Merlin, for example, an international health-based NGO, serves as the co-chair, alongside the World Health Organization, of the health cluster in Myanmar, the Democratic Republic of the Congo, and Somalia/Puntland (Doull 2011).  Merlin’s experience in the health cluster has been positive, finding that the time and resources committed were worthwhile and that the Cluster Approach strengthened thier relations with national authorities, UN agencies, and donors. Moreover, Merlin also reported experiencing greater influence over health policy and practice while in the capacity of co-chair.

In addition, partnerships tend to be strong when NGOs have preexisting relationships with UN agencies prior to the cluster roll out, or in cases of long-standing emergencies, where partnerships have developed over many years.  For example, the long-term humanitarian efforts in Uganda prior to the cluster launch meant that coordination bodies and efforts were already strong (Knudsen 2011: 7).  In this case, local NGOs reported greater visibility and influence over practices due to their knowledge of community approaches developed over years of experience.  Nonetheless, it should be noted that such partnership was made possible by the enduring crisis which allowed institutional knowledge to grow year over year and may not be easily transferable to sudden-onset emergencies where there may be little time to develop such relationships (Knudsen 2011: 7).

4. Has the Cluster Approach increased the accountability of relief efforts?

Category Analyzed Dimensions of Analysis
To increase the accountability of relief efforts. Is there coherence in the application of global standards of aid? Are appropriate monitoring and evaluation of cluster plans carried out? Has the Cluster Approach included downward accountability to affected populations?


One of the principle reasons that the Cluster Approach was established was the insufficient accountability and transparency in humanitarian relief efforts.  In short, accountability is about the responsible use of power and is important to ensure that programs are delivered effectively, meeting the needs of those in crises and reducing the possibility of error (HAP 2012).  In a humanitarian context, there are three main types of accountability: accountability to the donor (upward accountability), accountability to cluster members (lateral accountability), and accountability to the affected population (downward accountability).

As the humanitarian relief system is often characterized by competition for resources and visibility, accountability to the donor often trumps accountability to the affected population.  This can lead to wasted resources and a duplication of efforts as there is a lack of commitment to coordinate and plan actions based on a needs assessment (Webster and Walker 2009; Munz 2007; Cosgrave 2007, as quoted in Stumpenhorst, Stumpenhorst & Razum 2011: 588).  In addition, recent perception studies have suggested that people who receive aid cannot (or choose not to) distinguish between different humanitarian organizations, therefore the failings of one organization often count against the humanitarian relief effort more broadly (Featherstone 2011: 6).  Thus, individual agencies are often collectively judged based on the actions of the weakest member, disintegrating access to and the trust of affected populations and the community (Featherstone 2011: 6).  However, if accountability is exercised properly,  humanitarian programs will be delivered more effectively, meeting the needs of those in crises and reducing the possibility of error and corruption (HAP 2012).  For these reasons, the IASC has recognized the importance of ensuring accountability to affected populations within the Cluster Approach.  Recently (April 2011) the IASC Principals agreed to integrate accountability to affected populations into individual agencies’ statements and policies (Lewis & Lander 2011: 8).

Although the importance of accountability is continually identified, downward accountability is one of the most cited areas for improvement in the Cluster Approach.  Clusters are often criticized for overlooking the needs and demands of those whom they intend to serve which can have severe implications for strategy development (Homes 2010).  Numerous evaluations found no evidence or examples of clusters promoting participatory approaches or found that evaluations of such work did not focus on their impact on target populations (Derdarian 2008: 38; Steets et al. 2010: 59; HRP 2009: 25).  This means that there are few mechanisms in place to allow voices of crisis-affected communities to be heard in cluster strategies.

Another area that the Principals of the IASC recently reemphasized is agreeing to strengthen lateral or collective accountability in the humanitarian response, as individual commitments cannot be divided from collective commitments (Lewis & Lander 2011: 8). Lateral accountability is beneficial as individual agencies often adhere to different standards, and thus, having individual clusters enhance collective accountability will lead to greater predictability in aid efforts (HRP 2009: 33).   On a positive note, clusters are beginning to strengthen lateral accountability and to develop strategies for exchanging lessons learned and best practices (Steets et al. 2010: 65).  Clusters are also enhancing the coherence of relief efforts through lateral accountability.  As Nadig (2012) points out, the New Sphere Handbook (2011) was recently revised with contribution from clusters to help define minimum standards of relief (20). Clusters are increasingly adapting these global standards to local circumstances, agreeing on common strategies, such as cash for work rates (in Uganda, Haiti, and Myanmar), standard food parcels (Uganda), and agricultural kits (Uganda and Chad) (Steets et al. 2010: 51).

A key irregularity in accountability deals with whether cluster members are demanding more commitment and better leadership from lead agencies.  The Humanitarian Coordinator (HC) in Chad (2008) explained that the Cluster Approach, which designates a cluster lead agency, was useful because he “knows where to point the finger when things go wrong” (IRIN 2008a).  As a result, when there are problems or complaints, the HC can go to the cluster lead agency and hold them responsible for their actions (or inaction) (IRIN 2008a).  However, as Featherstone (2011) explains, in humanitarian hubs like Port-au-Prince, Islamabad, or Addis Ababa, where there are agency heads that plan the delivery of millions of dollars of aid, there is no single person or collective entity accountable for achieving humanitarian goals (7).  Thus, there is a discrepancy in the evaluations on whether or not cluster lead agencies are actually being held accountable to donors, other clusters and/or the affected population for their actions.


The Cluster Approach has increased the effectiveness of humanitarian action and is a positive change in humanitarian relief coordination.  It has been repeatedly cited as having “benefits that outweigh the costs” and although it has been criticized, it is currently the most appropriate structure for relief coordination.  However, although the Cluster Approach provides an improved structure for humanitarian coordination relative to pre-2005 mechanisms and is working toward achieving the goals set out in the Humanitarian Response Review and the IASC Guidance Note, there are a number of challenges to this approach that must be addressed.  First, there are large gaps in predictable leadership.  This is primarily due to the high turnover rates of cluster coordinators, lack of impartiality of cluster lead agencies, and insufficient training and experience of cluster coordinators. Second, there are significant barriers to inclusive partnership in the Cluster Approach.  Cluster coordination is not only labour intensive, requiring a significant amount of time and resources for effective participation, but it has largely failed to create a sense of NGO ownership and involvement.  Third, the Cluster Approach does not have sufficient mechanisms in place to enhance accountability to affected populations.

Many of the challenges point to problems of directive leadership from cluster lead agencies, illustrating that the benefits of a collaborative partnership between actors in humanitarian relief, including UN agencies, international, national, and local NGOs, are not being fully realized by the Cluster Approach. This suggests that despite the goal of strengthened partnerships, the Cluster Approach remains overly concentrated on promoting a centralized decision making body.  It is important that cluster lead agencies and cluster members revisit the founding principles of a collaborative coordination effort to enhance the successes of the Cluster Approach.  These challenges can be addressed within the existing structure of the Cluster Approach.  Therefore, the key findings of the meta-analysis highlight that there are many opportunities for improvement within the Cluster Approach, emphasizing that the structure of the Cluster Approach marks, by and large, a positive shift in humanitarian relief coordination.


Appendix 1.
Comprehensive Evaluations

Currion, P. & Hedlund, K. (2010) Strength in Numbers: A Review of NGO Coordination in the Field.  Lessons Learned v1.0.  International Council of Voluntary Agencies.

Currion, P. & Hedlund, K. (2010a) Strength in Numbers: A Review of NGO Coordination in the Field.  Overview Report v1.0.  International Council of Voluntary Agencies.

Featherstone, A. (2010) Fit for the future? Strengthening the leadership pillar of humanitarian reform. NGO and the Humanitarian Reform Project. October.

Humanitarian Reform Project (HRP) (2009) Synthesis Report: Review of the engagement of NGOs with the humanitarian reform process. October.

Steets, J., Grunewald, F., Binder, A., de Geoffroy, V., Kauffmann, D., Kruger, S., et al. (2010). Cluster approach evaluation 2 synthesis report. Global Public Policy Institute and Groupe URD.

Stoddard, A., Harmer, A., Haver, K., Salomons, D., & Wheeler, V. (2007). Cluster approach evaluation final. New York: OCHA Evaluation and Studies Section

Published (Journals) Articles, & Books

Altay, N., & Labonte, M. (2011). Humanitarian logistics and the Cluster Approach: global shifts and the US perspective. In Humanitarian Logistics: Meeting the Challenge of Preparing for and Responding to Disasters. Eds. Cristopher, M., & Tatham, P. Kogan Page Limited: London: 85-101

Charles, A., Lauras, M., & Tomasini, R. (2010). Collaboration Networks involving Humanitarian Organisations — Particular Problems for a Particular Sector. International Federation for Information Processing. Advances in Information and Communication Technology. V. 336: 17-165

Daniel Maxwell & John Parker (2012). Coordination in food security crises: a stakeholder analysis of the challenges facing the global food security cluster. Food Security. Springer Netherlands. 4:25-40

Miriam Stumpenhorst, Rolf Stumpenhorst, & Oliver Razum (2011). The UN OCHA Cluster Approach: gaps between theory and practice. Journal of Public Health. 19:587-592.

Organization Publications and Articles

Derderian, K., Stobberts, E., Singh, I., Rocha, S., and Melody, D.  (2008).  UN humanitarian reforms: a view from the field. Humanitarian Exchange. No. 38. Humanitarian Policy Group. June: 36-39 

Doull, L. (2011).  NGOs and the Cluster Approach: a worthwhile investment? Health Exchange: responding to emergencies. July 1.  Available at: http://healthexchangenews.com/2011/07/01/ngos-and-the-cluster-approach/

Homes, Sir J. (former Under-Secretary-General for Humanitarian Affairs and Emergency Relief coordinator) (2010). Learning the lessons of Haiti. Humanitarian Exchange. No. 48. Humanitarian Policy Group. October: 2-3

Integrated Regional Information Networks (IRIN) (2008a) Chad: Mixed verdicts on coordination of massive relief effort.  IRIN News Africa. Ndjamena. 23 January.  Available at: http://www.irinnews.org/Report/76386/CHAD-Mixed-verdicts-on-coordination-of-massive-relief-effort

Integrated Regional Information Networks (IRIN) (2008b) Kenya: Crisis highlights cluster complications.  IRIN News Africa. Nairobi. 12 February.  Available at:  http://www.irinnews.org/Report/76698/KENYA-Crisis-highlights-cluster-complications

Knudsen, C. (UNICEF) (2011) Partnership in principle, partnership in practice. Humanitarian Exchange. No. 50 Humanitarian Policy Group April: 5-8

Lattimer, C. & Berther, A. (2010). The work of the Education Cluster in Haiti. Humanitarian Exchange. No. 48. Humanitarian Policy Group. October: 18-20

Nadig, A. (the Sphere Project) (2012).  The Sphere Project: taking stock. Humanitarian Exchange. No. 53. Humanitarian Policy Group. February: 30-32

Polastro, R. (DARA) (2012).  Humanitarian response in conflict: lessons from South Central Somalia.Humanitarian Exchange. No. 53. Humanitarian Policy Group. February: 18-21

Street, A. (2011) Humanitarian partnerships: what do they really mean? Humanitarian Exchange. No. 50 Humanitarian Policy Group April: 43-46

Works Cited (Also refer to Appendix 1)

Adlnolfi, Costanza,  David S. Bassiouni, Halvor Fossum Lauritzsen, and Howard Roy Williams (2005). Humanitarian Response Review.  United Nations Emergency Relief Coordinator & Under-Secretary-General for Humanitarian Affairs, Office for the Coordination of Humanitarian Affairs.  August.

Alter, Catherine/Hage, Jerald (1993): Organizations Working Together, Newbury Park.

Broughton, Bernard/Maguire, Sarah (2006): Inter-agency Real-Time Evaluation of the Humanitarian Response to the Darfur Crisis, New York, United Nations Emergency Relief Coordinator and Under-Secretary General for Humanitarian Affairs.

Global Humanitarian Assistance (GHA) (2012) Defining Humanitarian Aid. Available at: http://www.globalhumanitarianassistance.org/data-guides/defining-humanitarian-aid

Gillman, Nina (2010) Interagency Coordination During Disaster: Strategic Choices for the UN, NGOs, and other Humanitarian Actors in the Field. Kiel, Univ. ISBN 978-3-8329-5708-7

Global Humanitarian Platform (GHP) (2007) Principles of Partnership: A Statement of Commitment.  12 July.

Humanitarian Accountability Partnership (HAP) (2012). What is Accountability? Available at: http://www.hapinternational.org/about/what-is-accountability.aspx

Inter-Agency Standing Committee (IASC) (2006) Guidance Note on Using the Cluster Approach to Strengthen Humanitarian Response.  24 November.

Inter-Agency Standing Committee (IASC) (2012) About the Inter-Agency Standing Committee. Available at:

Intergovernmental Panel on Climate Change (IPCC) (2012).  Managing the Risks of Extreme Events and Disasters to Advance Climate Change Adaptation. A Special Report of Working Groups I and II of the Intergovernmental Panel on Climate Change [Field, C.B., V. Barros, T.F. Stocker, D. Qin, D.J. Dokken, K.L. Ebi, M.D. Mastrandrea, K.J. Mach, G.-K. Plattner, S.K. Allen, M. Tignor, and P.M. Midgley (eds.)]. Cambridge University Press, Cambridge, UK, and New York, NY, USA, 582 pp. Available at: http://www.ipcc-wg2.gov/SREX/images/uploads/SREX-All_FINAL.pdf

James, Eric. (2008) Managing Humanitarian Relief: An operational guide for NGOs.  Intermediate Technology Publications Ltd. Warwickshire.

Lewis, Gwyn and Brian Lander (2011) Only as strong as our weakest link: can the humanitarian system be collectively accountable to affected populations? Humanitarian Exchange. No. 52 Humanitarian Policy Group. October: 8-10

One Response (2012) Cluster Approach: Overview. Available at: http://oneresponse.info/Coordination/ClusterApproach/Pages/Cluster%20Approach.aspx

United Nations Office for the Coordination of Humanitarian Affairs (OCHA) (2012) Annual Report 2011. 29 May. Available at: http://unocha.org/annualreport/2011/year-in-review

Reindorp, Nicola/Wiles, Peter (2001): Humanitarian Coordination: Lessons from Recent Field Experience.

Seybold, Taylor B. (2000): Systemic network analysis of refugee relief: a business-like approach. Available at: http://reliefweb.int/report/world/systemic-network-analysis-refugee-relief-business-approach

Shepherd-Barron, James. (2009) Clusterwise: Everything you wanted to know about cluster coordination but were afraid to ask. 20 December. Available at: http://www.clustercoordination.org/files/ccohb/Clusterwise.pdf

Stephenson, Max Jr./Schnitzer, Marcy (2009): Exploring the Challenges and Prospects for Polycentricity in International Humanitarian Relief. American Behavioral Scientist. February. Vol. 52 (6): 919-932

Van Brabant, Koenraad (1999): Opening the Black Box. An outline of a framework to understand, promote and evaluate humanitarian coordination. Humanitarian Policy Group, Overseas Development Institute.  London.  Available at: http://www.odi.org.uk/work/programmes/humanitarian-policy-group/

[1] As committed to in the Principles of Partnership (GHP 2007) and set out as intended outcomes of the Cluster Approach in the IASC Guidance Note (IASC 2006).  For more information see section three in key findings.

[2] For example, see Alter and Hage (1993), and Van Brabant (1999), Seybolt (2000), Reindrop and Wiles (2001), Stephenson and Schnitzer (2009), and Gillmann (2010).

[3] Made public in 2006, but written in September 2004 and January 2005, affecting OCHA’s need for systemic reform (Gillmann 2010: 61).

[4] Clusters are formally implemented in 27 of the 29 countries with Humanitarian Coordinators.  Humanitarian Coordinators are designated in countries by the UN Emergency Relief Coordinator on behalf of the Secretary-General on the occurrence of a complex emergency or when a humanitarian situation worsens (Gillmann 2010: 330). As of August 2012, clusters were active in: Afghanistan, Burundi, Central African Republic, Chad, Colombia, Côte d’Ivoire, Democratic Republic of Congo, Dominican Republic, Ethiopia, Guinea, Haiti, Indonesia, Iraq, Kenya, Liberia, Myanmar, Nepal, Niger, occupied Palestinian territories, Pakistan, Philippines, Somalia, Sri Lanka, Timor-Leste, Uganda, Yemen, and Zimbabwe (Relief Web 2012).

[5] When the cluster Approach was implemented in 2005, the IASC called for two formal phases of external evaluations to be done on the Cluster Approach: Stoddard et al. 2007 and Steets et al. 2010.  At the time of writing, these evaluations constitute the most robust evaluations of the Cluster Approach.

[6] This was particularly noted in Derderian et al. (2008: 34-39), HRP (2009: 29), Polatro (2012: 20), Steets et al.  (2010: 9, 42, 64), and IRIN (2008b).

Comments are closed.