Issue 46 - Article 13

Hard lessons for humanitarian financing from Pakistan

March 25, 2010
Benedict Dempsey, Save the Children UK

The humanitarian funding system for the Pakistan emergency in the summer of 2009 did not work well. In the absence of an in-country pooled fund, many international donors gave their funding directly to UN agencies, to be distributed through a malfunctioning Cluster system. The decision to use UN agencies as proxy pooled funds sacrificed speed, effectiveness and transparency. NGOs with the capacity to deliver aid on the ground experienced long delays in receiving funds, with a serious knock-on effect on their ability to help affected people. The system also placed intolerable pressure on the Clusters themselves, making it difficult for them to act as objective coordinating bodies. The UK’s Department for International Development (DFID), one of the biggest donors in Pakistan, chose to fund NGOs bilaterally when faced with delays, immediately increasing the speed of disbursal. This kind of flexibility should be applied in future emergencies. The Pakistan emergency also starkly illustrated the need to maintain the Clusters’ intended coordination role, and not to expect them to act as direct conduits for funding. 


Funding through UN agencies

In early May 2009, escalating conflict between the Pakistani military and insurgents in North-West Frontier Province caused over two million people to leave their homes. Humanitarian agencies mobilised to help them. No Emergency Response Fund (ERF) or other pooled fund was operating in Pakistan, so donors were faced with a choice. Either they could fund implementing agencies, including NGOs, bilaterally, or they could give the bulk of their funding to UN agencies, to be disbursed onwards to implementing partners. Many donors, including DFID, chose the second option, giving funding mainly to UN agencies. In DFID’s case, £12 million was earmarked for the UN, with £8m going to the Red Cross/Red Crescent.[2]

 The stated aim of funding in this way was to ensure coordination. Funding UN agencies would allow monies to be allocated through the Cluster system, maximising coordination between agencies. In the absence of an in-country pooled fund, it also cut down on donors’ transaction costs. However, with an estimated 80% of delivery capacity residing with NGOs, agencies like Save the Children voiced concern that funding through the UN would cause funding delays. Unfortunately, that is exactly what happened.



As Tables 1 and 2 show, donor funding through the UN had a big effect on which projects received money. Although the appeal as a whole was underfunded, UN agencies in the first two months received more than 40% of the money they requested, and nearly half of UN projects received some funding. In contrast, only 13% of non-UN projects received any funding, amounting to less than 5% of the overall funding they required in the first two months of the response.


If the system had been functioning properly, the money received by UN agencies would quickly have made its way to NGOs, as they had the delivery capacity required. However, there is a lack of transparency in funding to UN agencies, so it is hard to know where funds were ultimately spent and whether they were delivered to implementing agencies. This is a problem not just in Pakistan but globally. UN agencies are not required by donors to specify exactly how and when they have passed funding onwards to implementing partners such as NGOs. As a result, beyond broad sectoral allocations it is often unclear which specific projects are being funded. Even if UN funds are passed on to NGO partners for implementation, the process adds an extra layer of bureaucracy, cost and delay. Using UN agencies as disbursal bodies, therefore, not only slowed the transfer of funds to NGOs, but also made it difficult to determine whether funds were actually reaching the intended beneficiaries.

Malfunctioning Clusters

The other reason for the breakdown of the funding system in Pakistan was that, generally, the Clusters did not function as they should have. The Cluster approach was originally devised in an attempt to ensure that humanitarian partners operate in an inclusive, consultative and transparent manner, with leadership and responsibilities established at a sectoral level. When functioning properly, Clusters play a crucial role in identifying needs during a humanitarian emergency, enabling donors to fund priority projects. Coordination and prioritisation provided by properly functioning Clusters is extremely beneficial to humanitarian response.

For Clusters to work effectively, national NGOs, international NGOs and UN agencies must operate on an equal and collaborative basis. In Pakistan, however, some Clusters were poorly run and coordination was weak. In most cases, the person designated by the UN to lead the Cluster was not dedicated to it full-time, having also to perform a pre-existing job for his or her UN agency. NGOs also often felt that some Cluster coordinators regarded Clusters as an extension of their parent agency. This resulted in some Cluster coordinators seeking funding for their own agencies before attempting to find an international NGO to implement a project. The way the Clusters functioned generally also meant that national organisations in Pakistan had limited or no access to the Clusters, and therefore were unable to access funding.

In this context, the deliberate channelling of funds through UN agencies only made matters worse. Many of the concerns with the way the Clusters were operating were confirmed by an Inter-Cluster Diagnostic Mission[1] to Pakistan, which took place in July. The mission returned with a number of conclusions, including:

  • Disbursal of funds through Clusters resulted in a perceived lack of transparency in decision-making, delays in disbursement of funds, lack of objectivity in resource allocation and conflicts of interest between coordinators and their respective agencies.
  • Cluster coordinators faced a conflict of interest from so-called ‘double-hatting’ (having to manage agency responsibilities and expectations while simultaneously managing the Cluster).
  • Each Cluster requires dedicated full-time coordinators to meet the onerous workload of coordinating a Cluster and to help reduce the inter-agency rivalries that inevitably arise in a situation of conflicting loyalties.
  • Direct financial disbursement through a Cluster undermines the objective ethos of a coordinating body. Funds should be disbursed directly from donors to partners or through a common pooled funding mechanism.

Asking Clusters to manage funding poisoned coordination mechanisms, increased the perception of bias in funding allocation, exacerbated inter-agency rivalry and further delayed disbursement.

The DFID example

DFID was a major donor to the Pakistan emergency response, pledging funds early and in large amounts. However, the decision was made to divide these funds only between the Red Cross/Red Crescent and UN agencies. Save the Children met DFID on 11 May and maintained contact throughout May and June, expressing concern at the decision to channel most funds through UN agencies, and at the slow pace of disbursement. Throughout this period, DFID put pressure on UN agencies in an effort to hasten funding disbursal. Eventually, though, DFID decided a new approach was needed, and by early June it had indicated that it would begin to fund NGOs bilaterally, provided the proposed activities were part of the OCHA Flash Appeal. This involved an earmarking process by which DFID funds were allocated in Cluster meetings, but it still required NGOs to submit proposals through UN agencies as Cluster leads.

On 16 and 17 June, Save the Children was earmarked $500,000 of DFID funding in the Early Recovery Cluster and $400,000 in the Health Cluster. Save the Children duly submitted proposals to the Cluster lead agencies for submission to DFID. Over a week later, on 26 June, this funding had still not come through, so Save the Children contacted DFID directly. At this point, DFID recommended that Save the Children rewrite both proposals into a DFID format, combining them into one and submitting them directly to DFID, thereby circumventing Cluster lead agencies entirely. Having done this, final confirmation of $900,000 in funding for health and early recovery projects was received on 13 July.

Although the funding had been delayed, DFID’s positive decisions when delays occurred increased the speed of disbursal, while maintaining the coordination function of the Clusters.



The delays in funding had real consequences. Save the Children had to wait until 13 July, over two months after the initial escalation of the crisis, for formal approval of DFID funding. Ironically, Save the Children’s projects, which were designed to assist IDPs in Pakistan, received funding approval on the same day that the government of Pakistan decreed that IDPs should return home. Bilateral funding from the US Office of Foreign Disaster Assistance (OFDA), in contrast, was received over a month earlier.

Projects that Save the Children was unable to implement in June and July because of the delays in DFID funding included ten mobile health clinics and ten mobile nutrition clinics. Each clinic would have reached 120 people per day, suggesting that, for several weeks, 2,400 people per day did not receive these important health and nutrition services. Once received, the funds were used for vital health and early recovery projects, but operations could not be scaled up quickly enough when needs were greatest.

It is also worth noting that the heaviest burden of assisting IDPs in Pakistan fell on host communities, who made an enormous effort to support displaced people. Without such assistance, independent of the international emergency response, the consequences for those affected would have been even more serious. The support provided by host communities in Pakistan cannot be relied upon in other countries or other, future emergencies, because other communities may not have the same level of resources as was available in Pakistan.



The importance of coordination is not in doubt and the Cluster system must be supported. However, the Pakistan experience offers some key recommendations for how the system can work better.

First, country-based pooled funds should be used wherever possible, to allow Clusters to carry out their coordination function without the burden of acting as funding mechanisms. Second, donors need to maintain flexibility in the way they provide funding. If there is no established pooled fund, donors should fund NGOs bilaterally as necessary. They should not automatically entrust all their funds to Cluster lead agencies that are unsuited to the task of onward funding disbursal. This would enhance, not detract from, the coordination function of Clusters. When working properly, Clusters are extremely positive mechanisms for identifying areas of need and prioritising projects. But Clusters should remain coordination mechanisms. They should identify priorities to go into the UN Flash Appeal, but they should not be expected to act as direct conduits for funding. Third, the role of Cluster coordinator should be separate from the operational activities of Cluster agencies. Only then might Clusters be viewed as impartial mechanisms for earmarking funds. It is unreasonable to expect an individual to coordinate a Cluster and simultaneously represent his or her own agency. Until this changes, Clusters will generally continue to function poorly. Finally, donors should require greater transparency in the onward disbursal of funds. If funds are allocated through Clusters, lead agencies should indicate which Cluster member (including itself) has received funding, what that funding is for and when it was disbursed. Cluster coordinators should be able to publish details of onward disbursal of funding within three days of disbursement. Unless donors insist on full transparency it will remain difficult to be sure the funds are reaching beneficiaries as intended.

It is to be hoped that the flexible use of pooled funds and bilateral funding, together with the effective implementation of Clusters, can avoid some of the problems faced in Pakistan. Providing life-saving aid to crisis-affected people depends on it.


Benedict Dempsey is Humanitarian Advocacy Officer, Save the Children UK. His email address is


[1] Inter-Cluster Diagnostic Mission to Pakistan, Islamabad and Peshawar, Pakistan, 13–17 July 2009.

[2] DFID’s stated funding for this emergency was £22m, but this includes £2m allocated in late 2008, some time before the emergency in question.


Comments are available for logged in members only.