It is no surprise that the word “accountability” is so widely misunderstood by the aid world when one realizes, as any supporter of Silvio Berlusconi surely does, that there is no such word in Italian. Or French. Or, for that matter, that other lingua franca of the aid business, Spanish. The nearest one gets to it is something similar to “responsibilité” … which of course, means something entirely different (see Section on ‘Consensus Management’).

Accountability is all about the responsible use of power (HAP, 2009). It comes in many shapes and forms:
  • upward accountability to donors and governments
  • lateral accountability to peers within the Cluster and other Clusters
  • downward or forward accountability to beneficiaries.

According to the Global WASH Cluster, “Accountability is the process through which an organization makes a commitment to and balances the needs of stakeholders in its decision-making and activities, and delivers against this commitment … that allows the organization to give account to, take account of, and be held to account by its stakeholders.”

Language is a judgment system … an encyclopedia of ignorance.” (Edward de Bono)

Challenging power imbalances is at the heart of accountability, and some would argue that an emergency is not the place for seeking informed consent; that the fragile psychological state of a population necessitates strong leadership and top down decision making. However, it is important to recognize that improving downward accountability is a process that starts with the capacity to listen and respond to those affected.

To assess the extent to which these principles and that of predictable leadership have been fulfilled by Cluster Lead Agencies, evaluators might verify and measure the following indicators (from draft IOM Handbook on Clusters):

  • CLA’s are carrying out their duties according to their specific IASC-mandated Terms of Reference (ToRs);
  • The CLA gives evidence of its commitment to cluster activities and not only to its own operational obligations;
  • Cross-cutting issues are integrated into cluster work plans;
  • Whether cluster leads have acted as providers of last resort;
  • That information shared with cluster members, other relevant clusters, the Humanitarian Coordinator and the Country Team is given with high levels of quality and technical detail;
  • Funding proposals and appeals (e.g. CERF, ERF, CHAP, CAP and Flash) reflect collective cluster priorities and equitably include projects and programmes from those individual agencies (with particular attention to NGOs and local partners).

To evaluate the effective application of the principles of partnership and the extent of cluster cohesiveness among all partners, different indicators are used, such as:

  • Collectively shared vision of all members is reflected in joint strategies, work plans, assessment and appeals;
  • Non-UN agencies participate in cluster strategies and decision making (e.g. NGOs and donors);
  • Linkages and cooperation with other relevant clusters are established.

Suggested indicators for evaluating accountability at the country level are the following:

  • Consistent application of globally agreed standards and tools by all partners;
  • Inclusion of requirements for the monitoring of performance in joint plans;
  • Consideration of the priority requirement of accountability to beneficiaries (e.g. in meeting the needs of affected populations) in cluster strategies;
  • Common understanding of the TORs and identification of roles, responsibilities and capacity of each stakeholder.

The Humanitarian Coordinator (who may or may not also be the Resident Coordinator) is accountable for the following. Normally, most of these functions would be delegated to the OCHA Head of Office who would, in turn, delegate those that are Cluster specific to the dedicated Inter-Cluster Coordinator (who becomes the de-facto Deputy HC). If at any time, you think he or she is failing to discharge these responsibilities, you should raise the issue with the head of the respective Cluster Lead Agency. Should nothing happen, inform the OCHA regional representative



  • Ensuring common service provision
  • Enabling inter-sectoral cooperation
  • Facilitating integration of cross-cutting issues
  • Provision of contextual socio-economic, political, inter-sectoral, and hazard risk analyses
  • Facilitation of joint needs assessments
  • Coherence of information management (including establishment of denominator baselines)
  • Consistent geo-physical mapping (including administrative boundaries)
  • Advising on joint survey design and coordinating the logistics of implementation
  • Overseeing the monitoring of performance management
  • Advocating on operational issues of common interest to all Clusters
  • Providing advice on internal and external governance structures
  • Mobilizing financial resources (using CERF, Flash Appeal, CAP, CHAP)
  • Joint contingency planning
  • Inter-Cluster strategy formulation
  • Policy development
  • Liaising with military and private security actors
  • Reporting
  • Identifying key partners and analysing stakeholder capacities
  • Coherent use of IT applications and web platforms for information-sharing
  • Ensuring Cluster Lead Agencies commit to ‘provider of last resort’ conditions
  • Clusters disengage/merge/phase down only once stated humanitarian objectives have been met and appropriate handover mechanisms are in place


Humanitarian organizations work in a complex web of relationships where accountability is invariably ‘upward’ to the donor, rather than ‘sideways’ to peer agencies (never government), or ‘downward’ to those affected by disaster. Because of these complexities, information is necessarily presented selectively in different ways to different audiences.

The incentive to innovate is distorted by this upward accountability which over-emphasizes linear, quantitative, time-limited, and incremental approaches to performance assessment. This also negatively affects the ability of NGOs to advocate or be effective catalysts for social change, thus weakening downward accountability.

It is our responsibility to demonstrate to stakeholders, foremost of whom are disaster-affected people, that humanitarian assistance complies with agreed standards.” (Sphere 2000)

Measurement of what NGOs and international aid organizations are trying to achieve with humanitarian action is almost always a matter of judgment and interpretation.

Agencies’ survival has been almost completely de-linked from performance because there is little in the way of ‘checks and balances’ that objectively measures the impact of what they do. There are inherent reasons why measuring impact is difficult in humanitarian operations, including:

  • difficulty in establishing causality
  • challenging operational environment
  • lack of shared objectives
  • curative rather than preventive approach to risk reduction
  • valid planning data is not available in time
  • agencies work under pressure, with limited human and financial resources

Structured interactions between relief organisations and their donors favour short-term and easily measurable activities at the expense longer-term processes that address underlying causes of vulnerability. Agencies gain funding by providing information on ‘successful’ projects, thereby conferring a positive reputation on themselves and, by inference, on their donors. This mutually reinforcing or ‘symbiotic’ dependency means there is no incentive to promote high quality work in the field; in fact, there is every incentive to over-promise what can be done.

Unfortunately, this situation is further entrenched by traditional monitoring and evaluation mechanisms where consultants tend to write only that which they know their client will tolerate so as not to jeopardize their future employability. Relating narrowly-focused evaluations to broader processes of change, and engagement with affected populations is compromised as a result. This is a critical systemic weakness in humanitarian aid delivery.

Results-based monitoring systems can be similarly counter-productive. The field view is that they take a long time to carry out and divert already limited management, logistics, human, and financial resources from on-going operations.

Results of (logical framework) analysis, ‘lessons learned’ studies, and so-called ‘real-time’ evaluations rarely affect meaningful or timely change on the ground as they tend to be done late and not provide information that is practically useful in aid agency work at community level. Partly, this is because the people doing the evaluations have limited experience of how coordination is actually carried out in the field, and limited knowledge of the context in which that particular disaster is taking place.

Lack of institutional memory, exacerbated by low staff retention and rapid rotation, means that many ‘lessons learned’ from one crisis fail to be applied to the next – they are lessons “continually un-learned”, in other words.

Linking “quick and dirty” lessons learning in real-time to skills upgrading through on-the-job ‘mentoring’  is probably the most efficient and effective way to transfer skills.




Most Global Cluster Leads (GCLs) have not yet articulated a coherent position and stance in terms of their own cluster leadership role and accountabilities. This, together with a growing perception that CLAs not able to sufficiently handle the management of its Clusters at global level, has increasingly led to frustration from external stakeholders. Meanwhile, imprecision around the Cluster role and selective interpretation by Country Offices as to how ‘leadership’ is applied so that Clusters are empowered and enabled is also progressively more at odds with both internal staff expectations and the expectations of partners.

Since 2005, the interpretation of Humanitarian Reform has evolved as indicated in evaluations, donor reviews, cluster field experiences and other discussion forums. This is contributing to ever widening interpretations which vary greatly among different stakeholding partners. This in turn has resulted in wide differences of opinion and expectation from donors, NGOs, governments and others regarding the Cluster responsibilities of CLAs.

More specifically, there is a lack of progress on clarifying accountabilities and designating these amongst all levels (HQ/RO/CO), and integrating these into work-plans, job descriptions and performance evaluations. Job descriptions frequently make no specific reference to their Cluster accountabilities.

There is an absence of any formal communication to Cluster partners on commitments and deliverables that they can expect and rely on from their respective CLA, either at global or at country level. Likewise, no ‘compact’ outlining the ethics, principles or code of conduct on the part of non-governmental organisations exist that clarifies what a CLA should expect from them.



Meanwhile, there are clear weaknesses related to core capacity, misunderstanding of roles, and operational resourcing needs required in the field to actually implement Cluster responsibilities. These include:

  • Consistent mis-appreciation by country level management of the human and material resources required to enable Clusters to function at national and sub-national level. Both the Phase I and Phase II evaluations of the Cluster Approach conducted by OCHA on behalf of the IASC since the humanitarian reform process was initiated highlight the ongoing need to ensure adequate cluster coordination capacity, including for information management and provision of specialised technical assistance. The WASH Cluster coordination team in Haiti included fifteen international staff at the peak, but was never less than nine. By way of comparison, the Shelter Cluster coordination team in Haiti comprised twenty-seven international staff in Port-au-Prince alone.
  • The Cluster Coordinator role requires independence from parent agency affiliation and programmatic response. The role is also a full-time job in its own right except for the smallest of responses. The Global CLA for Emergency Shelter in natural disasters (IFRC) mobilises a coordination team of three people, plus all the office supplies required for independent functionality, as a standard ‘surge’ response to all disasters, regardless of advice from country level Delegations. WFP does the same for Type 2 disasters and above.
  • Coordination management requires a different set of skills from that of programme management.
  • While some advancement has been achieved in terms of human resources and global rosters (notably in WASH, including the Rapid Response Team), the pool of qualified cluster coordinators and support functions remains inadequate.

There are reasons why Heads of CLAs appear to be selectively applying their Cluster accountabilities. These include, but are not exclusive to:

  1. Incoming Cluster Coordinators (and the teams they lead) are frequently unknown entities. Sometimes relatively junior; sometimes from NGOs; quite often ‘freelance consultants’, some of whom have been ‘surged’ under standby partnership arrangements with donors; CLA Heads fear that they will be held to account for decisions taken by, or under the auspices, of these people which will have unknown and open-ended consequences under ‘provider of last resort’ principles.
  2. Being fearful of “upsetting normative relationships” with their (developmental) governmental and donor counterparts, and jeopardising longer-term development programmes and future income streams thereby.
  3. Cluster Coordinators appointed from within the CLA fear for their careers if they disagree with their Representative, so tend to conform to the ‘parent’ agency’s view. Paradoxically, many so-called ‘independents’ and seconded NGO staff view being a member of a Cluster team as a means to fast-track their careers by obtaining a position within a UN agency. This, too, injects skew into what Heads of CLAs are told.
  4. It is unfair to expect a CLA Head to have his or her performance assessed for operational or programmatic aspects of the Cluster carried out by so-called ‘partners’ over whom they have no control, little strategic oversight, and minimal opportunity to exercise quality assurance.

HCs are often ill-equipped and too inexperienced in humanitarian affairs to exert the ‘leadership’ required.

Both HC’s and CLA Heads don’t fully comprehend the reputational risk to themselves or their organizations for under-performing.

CLA Heads often don’t have enough direct experience of applying the Cluster Approach to know what levels of investment are required to establish and maintain national and sub-national coordination mechanisms.

Beyond vague principles of ‘leadership’,  ‘accountabilities’ have never been made explicit.

In major emergencies, there have been problems with inter-cluster coordination. While this is within the remit of OCHA’s role and responsibilities, there are implications for CLAs.



Since 2005, humanitarian actors have developed a whole range of policies, guidance notes and terms of reference through IASC mechanisms that crystallize inter-agency consensus on the responsibilities and expected results of those humanitarian actors and platforms that have system-wide roles (e.g. Humanitarian Coordinators, Humanitarian Country Teams, Clusters). In some cases (i.e. for Cluster Coordinators) such guidance is still under development, while in others (i.e. for HCT members, Cluster members, Heads of Global CLAs, Global Cluster members) it is yet to be developed. These documents encapsulate inter-agency expectations of how relevant humanitarian actors should work collaboratively together to achieve results that go beyond the capacity of any single organisation to deliver.

This major inter-agency policy development effort has not been matched by a commensurate effort to develop systems to hold relevant actors and platforms accountable for implementing these normative documents. As a consequence, implementation has to some extent been left to the goodwill of individual actors and platforms, leading at times to sub-optimal collective results.

The following gaps and weaknesses have been identified by OCHA in current accountability systems at country level:

Humanitarian Coordinator: The accountability system for the humanitarian coordination leadership function, which is centralised with the Emergency Relief Coordinator (ERC), Valerie Amos, is not aligned with that of RCs, which is decentralised and pivots around Regional Directors Teams. Such misalignment is problematic insofar as there is considerable overlap between the two functions: all RCs have humanitarian responsibilities vis-à-vis preparedness, early recovery and, increasingly, disaster risk reduction, and most HCs are also RCs.

It should also be noted that RCs and HCs are only accountable for being good process managers (i.e. leading and coordinating humanitarian action), and not for collective results.

Humanitarian Country Team: No one holds HCTs as collective entities accountable for achieving collective results. The RC or HC leads and chairs the HCT, but does not have formal authority over it.

Cluster Lead Agency Country Representative: It is unclear whether an Agency Country Representative is held accountable by his/her supervisor or, in turn, holding accountable his/her staff members who participate in Clusters.

Although the Guidance Note on Using the Cluster Approach of November 2006 [see Annexes] states that CLAs are accountable to the HC for the performance of their Cluster-related functions, only few RCs and HCs actually hold Heads of CLAs accountable. While lack of understanding of such prerogative may explain this to some degree, the main reasons are to be found elsewhere. First, RCs and HCs prioritise their long-term relationship with Agency Country Representatives, which is essential for achieving results at the country level: and since RCs and HCs do not have any authority over them, the quality of this relationship depends to a large extent on Agency Country Representatives’ good will, which might be eroded by the RC or HC’s Cluster-related demands. Second, the career of RCs and HCs depends to a significant extent on agencies, as their annual performance appraisal is conducted by the Regional Directors Team, which is composed of Agency Regional Directors, and their future posting as RC, RC/HC or DSRSG/RC/HC is to a large extent determined by the UN Inter-Agency Advisory Panel (IAAP), which again is composed of agency representatives. In sum, RCs and HCs do not have sufficient authority over Heads of CLAs to be able meaningfully to hold them accountable.

It should also be noted that Heads of CLAs are only accountable for being good process managers (i.e. leading and coordinating the Cluster), and not for the results of the Cluster.

Cluster Coordinator: It is unclear whether a Cluster Coordinator is held accountable by the Head of CLA as his or her immediate supervisor for coordinating the Cluster efficiently and effectively, and, if so, to what standards of performance.

Cluster: No one holds Clusters accountable for achieving collective results. The Head of CLA is only accountable for being a good process manager.

Cluster Member: It is unclear whether an agency staff member who participates in a cluster is held accountable by his/her supervisor (usually the Agency Country Director) for being a collaborative and constructive member of the Cluster.

In an address to the annual CERF conference on 9th December 2009, the DFID Minister for International development outlined the following as two of DFID’s five humanitarian reform priorities:

“… coordinating response is core business, and in most emergencies a full time job. As Cluster leaders, UN agencies must look beyond mandates, develop innovative partnerships, and dedicate resources to professional coordination …

… we need to be more accountable and better at showing that what we do makes a difference.”

Shortcomings in how Global CLAs were perceived to be inadequately supporting the Cluster Approach were exposed in Haiti in an apparently ‘leaked’ letter from the then Emergency Relief Coordinator, John Holmes, to Agency Executive Directors on 13th February 2010. What follows is an extract from the full text:

Regarding coordination, I was disappointed to find that despite my calls for the Global Cluster Lead Agencies to strengthen their cluster coordination capacity on the ground, very little progress has been made in this critical area. In most of the twelve clusters established, cluster coordinators continue to struggle without the capacity required to coordinate efficiently the large number of partners involved in the operation. One month into the response, only a few clusters have fully dedicated cluster coordinators, information management focal points and technical support capacity, all of which are basic requirements for the efficient management of a large scale emergency operation. This lack of capacity has meant that several clusters have yet to establish a concise overview of needs and develop coherent response plans, strategies and gap analyses. This is beginning to show and is leading others to doubt our ability to deliver.

Among the many lessons already identified from this disaster is the need for robust cluster coordination teams with adequate seniority to take charge of cluster coordination at the outset of the response. To place one person as a cluster coordinator is simply inadequate and falls critically short of what Global Cluster Lead Agencies have committed to.

I would also like to request NGOs to look at ways of strengthening their own capacity on the ground and to consider contributing personnel to support cluster coordination efforts”.

This RASCI accountability framework is taken from the Health Cluster coordination handbook (p.34-35)













1. Coordination mechanisms and inclusion of all stakeholders within and between Clusters A R R S
2. Stakeholder Relations, including with national authorities and other local actors A R R S C I
3. Needs assessment, monitoring and analysis, including gap identification A R R S C
4. Strategy development and planning, including community based approaches, attention to priority cross-cutting issues, and filling gaps A R R S
5. Contingency planning A R R R
6. Application of standards R S R S A R
7. Training and capacity-building, including emergency preparedness A S R S C
8. Advocacy and Resource Mobilization, including reporting A R S C S C
9. Communication, including Media Relations A R S C C
10. Provider of last resort A R S S C


R = Responsible: Those who do the work to achieve the task. There can be multiple resources responsible.

A = Accountable: The person/people ultimately answerable for the correct and thorough completion of the task.

S = Support: Those who may help in the task.

C = Consulted: Those whose opinions are sought in a two way communication.

I = Informed: Those who are kept up to date on progress through two way communication.


The following is the text of a joint letter from Cluster Lead Agencies to their Representatives at Country Level circulated on 20 October 2009:

“Dear Colleagues,

At the Inter-Agency Standing Committee Working Group meeting in July 2009, representatives of all Global Cluster Lead Agencies discussed the implementation of the cluster approach. It was agreed that in order to strengthen coordination and accountability within the system, all Global Cluster Lead Agencies should write to their respective Country Representatives/Directors, reminding them of the crucial role they play in ensuring that their agencies fulfil their commitments as cluster leads. Specifically, it was confirmed that:

1. Representatives/Directors of Cluster Lead Agencies at the country level have a dual responsibility to represent the interests of both their own agency and the cluster(s) they lead in Humanitarian Country Team meetings (or equivalent) and in other relevant fora. Ideally, this

dual responsibility should be reflected in their job descriptions and performance appraisals.

2. While Cluster Coordinators are responsible for the day to day running of clusters in the field, it is the Country Representatives/Directors of Cluster Lead Agencies who are ultimately accountable to the Resident Coordinator / Humanitarian Coordinator (subject to agency-specific agreements defining such accountability) for carrying out their Cluster Lead Agency responsibilities (as specified in the November 2006 IASC ‘Guidance Note on Using the Cluster

Approach to Strengthen Humanitarian Response’ and Annex 1, ‘Generic Terms of Reference for Sector/Cluster Leads at the Country Level’).

3. Cluster Coordinators should act as neutral representatives of the cluster as a whole, rather than as representatives of their particular agency. One way to ensure this separation of roles is to appoint (i) a dedicated Cluster Coordinator with no agency-specific responsibilities and (ii) a separate cluster focal point for the agency who can represent the agency and participate in the work of the cluster on an equal footing with other cluster partners. These arrangements are a matter for each agency to decide on a case by case basis depending on the resources available,

the security situation and other factors.

4. Where dedicated Cluster Coordinators are appointed, their responsibility to act as a neutral representative of the cluster as a whole rather than as a representative of their own particular agency should be reflected in their job descriptions and performance appraisals. In cases where

the Cluster Lead Agency is not able to appoint a dedicated Cluster Coordinator, the individual who is given responsibility for cluster coordination at the country level will have a dual responsibility to represent both the interests of the cluster and their own agency programmes in

relevant fora. This dual responsibility should be reflected in their job descriptions and performance appraisals”.




Sources used in this article:

[1] Draft Letter from Global WASH Cluster members to the Executive Director, July 2010 (i.e before onset of floods emergency in Pakistan)

[2] Letters from DFID , Oxfam et al to UNICEF Executive Director, September 2010

[3] Haiti Earthquake WASH Cluster Lessons Learned; James Shepherd-Barron, 3rd September 2010

[4] Internal summary of exit interviews conducted by NYHQ on staff departing Haiti, Jan-Aug 2010

[5] Grunewald et al./GPPI/URD (2010) “IASC Cluster Approach Evaluation, 2nd Phase April 2010”. p.24

[6] Draft HATIS discussion paper on ‘The Cluster Approach: How does UNICEF conform?’, 11th August 2010

[7] Letter to Director PD from Global WASH Cluster members, September 2010 [See Annexes]

[8] UNICEF EMOPS/PD (2009) “Synthesis – Lessons on Implementing the Cluster Approach”

[9] Joint letter from Global Cluster Lead Agencies to their Country Office Representatives, 20th October 2009 [See Annexes]

[10] RedR-UK Cluster Training materials, 2009

[11] Interview with Head of CLA for Emergency Shelter, Yogyakarta, Indonesia, 2006

[12] Presentation by DFID to Executive Committee of UNHCR, Geneva, 7th October 2008

[13] Verbal statement made by a donor in Pakistan, 30 September 2010 (no evidence)

[14] Draft OCHA paper entitled, ‘Strengthening System-wide Accountability’; 14th September 2010

[15] IASC Guidance Note on ‘Using the Cluster Approach to Strengthen Humanitarian Response’, 24th November 2006

[16] These teams include the Regional Directors of the operational agencies of the UN system.
This is is a section from Clusterwise 2. Reproduction is encouraged. It would be nice if the author, James Shepherd-Barron, and were acknowledged when doing so.

Submit a Comment