Improving NGO coordination: lessons from the Bam earthquake

In a recent lessons-learned meeting on the Bam earthquake in Iran, a polite and respectful colleague from the Iranian Ministry of Health related his frustration at international NGO coordination in the early days of the emergency. He said that, at the same time as he was desperately trying to set up field hospitals and bury the dead, representatives from over 100 international NGOs had individually requested meetings with him. He appreciated their help, he said, but some organisations wanted to ask him about the siting of rural clinics when he was still trying to arrange emergency medical evacuations. Was there no way, he asked, that these agencies could organise themselves better in the early days of a disaster?

This article tries to answer his question. It reflects first on the nature of international NGO coordination and cooperation in the Bam response, then asks if there are ways that inter-agency coordination could be improved – without losing sight of agencies’ individual identity and independence. Finally, it suggests areas for further discussion and research.

Background

The earthquake hit in the early hours of the morning on 26 December 2003. Its scale was massive: some 26,000 people out of a population of about 90,000 were killed, and about 80% of the housing stock and government infrastructure was destroyed. Key government officials, health staff and national response personnel were among the dead.

The Iranian government took the unprecedented step of requesting international assistance and – in a brave move for a country largely closed to outsiders for 30 years – instituted a no-visa, open skies policy for relief supplies and personnel to assist the affected population. In the first two weeks after the earthquake, over 120 tonnes of relief supplies and 1,900 international personnel arrived in Bam, complementing the 4,500 tonnes of supplies and 8,600 personnel already deployed by the Iranian government and the Red Crescent. Many professionals in the relief sector considered the initial response exemplary, and it is true that performance on key indicators was impressive. There were no significant disease outbreaks; according to the Iranian Red Crescent, the first food and water distributions began within 12 hours, and over 20,000 tents were distributed within the first three days.

Initial coordination mechanisms

Over 200 international organisations arrived in Bam in the first two weeks after the earthquake. To assist in the coordination effort, the Iranian government and the UN Office for the Coordination of Humanitarian Assistance (OCHA) set up a camp for international organisations on a former sports field. OCHA also established sectoral coordination meetings and strengthened links with the Iranian Red Crescent, the lead agency for the response. Organisations helped each other, and personal relations were good. Oxfam borrowed tents from Save the Children, and Médecins Sans Frontières provided supplies to other organisations. There was a strong sense that everyone simply wanted to get things done to help the people of Bam.

There were also more formal coordination mechanisms. A group of five agencies banded together to initiate a joint latrine procurement tender using agreed standards and specifications. The tender failed in the end, but it was proof that organisations were willing to cooperate and to agree standards and approaches in a coordinated matter. A system initiated by the Iranian Red Crescent, dividing the city into 12 zones each to be assisted by a different provincial branch, was rapidly picked up by international NGOs, who agreed to accept assigned zones and to focus their assessment and intervention work in these areas.

Coordination problems

Perhaps it was inevitable, given the scale of the disaster, that there were difficulties in coordination among the agencies working in Bam. Some agencies came ill-equipped – some even arrived without sleeping bags or enough food and water. Small agencies were assigned huge zones to work in, and record systems looked at the amount of work being done, not at whether needs were being covered. Most international NGOs conducted individual assessments, the results of which were generally not shared between agencies. Several agencies said that they did not share their reports because they were not finished, or because they felt uncomfortable sharing information that had been gathered quickly. Others said simply that they were not asked to share their assessments, or that they did not have time to do so. Project designs varied significantly between agencies, and some were not accepted locally: women would not use latrines with plastic sheet superstructures because their shadow could be seen through them. Sometimes it took weeks for this learning to be communicated between agencies. Suppliers played agencies off against each other to get higher prices, and there was general confusion about what was being done and distributed where.

Sorting through the chaos

Was this situation inevitable? How could we have communicated better in the early days of the response to reduce the duplication of effort? Could international NGOs have been better coordinated as a group before the emergency?

The traditional answer to these questions usually has two parts. First, coordination is the job of OCHA. If international NGOs set up their own coordination system, this would just add to the mess and potentially increase duplication. Second, individual NGOs have their own missions, mandates and systems, and it is exactly this independence that can make them so effective. Push inter-agency coordination too far and gaps in coverage will appear, and essential flexibility will be lost. These are valid points, but is there really no way to do better?

Old hands say that things are far better than they were 20 years ago, mainly due to three factors: the broad acceptance by most (though not all) international NGOs of international standards such as Sphere; the development of OCHA and the UN Disaster Assessment and Coordination (DAC) teams; and the emergence of inter-agency coordination bodies such as the International Council of Voluntary Agencies (ICVA) in Geneva, the Washington-based VOCA and the broader-based Inter-Agency Standing Committee (IASC). Applying this argument to the early days of the Bam earthquake, however, suggests that there is still room for improvement.

The application of international standards

Almost all agencies operating in Bam publicly professed to be applying key international standards such as Sphere in their work. Other technical standards were also used, and there is no doubt that standardising the fittings of relief water items, for example, easily proved its worth. Yet Sphere was seldom actively used as a coordination and planning tool. None of the weekly UN progress reports systematically compared actual data to Sphere. And while Sphere was used in reviewing assessment data, it was of only limited benefit in speeding up the design process for latrines and other materials.

OCHA and UNDAC

Despite years of emergency interventions, there remains confusion regarding the role of OCHA. OCHA is clear that its coordinating role within a humanitarian response does not extend to leading it. Yet many NGO representatives in Bam suggested that they would have appreciated OCHA playing a more active part. One senior staffer told me that one reason why agencies were so willing to embrace the zonal system was that there was an urgent need for organisations to be clearly directed – ‘they really wanted to be told where to go so that they could get down to work’. Government officials also wanted OCHA to assume this role, although this clearly was not part of its mandate.

UNDAC meanwhile remains an essentially UN-centric tool. The UNDAC system is ‘designed to assist the United Nations in meeting international needs for early and qualified information during the first phase of a sudden-onset disaster and in the coordination of incoming international relief at national level and/or at the site of the emergency. It also aims at strengthening national and regional disaster response capacity’. In Bam, an assessment team was deployed within 72 hours. This included representatives from a range of UN bodies, plus at times donor representatives. Despite the fact that UNDAC is meant to act as ‘part of a joint effort to enhance system-wide coordination’, there is currently no provision for international NGOs to be represented on the UNDAC team. One UNDAC team member told me that they would have welcomed such participation, but were unsure who would be best suited for it, and how to get good-quality staff at such short notice. More discussion on this is needed.

Current coordination mechanisms

This brings us to the third improvement in the relief sector in the last two decades: the reinvigoration of NGO networks such as ICVA and Interaction. Despite the fact that most organisations arriving in Bam were a member of one of these coordinating bodies, evidence of this membership was not obvious. In the first month of the response, not one UN meeting or inter-agency assessment report appears to have mentioned these organisations. Although clearly it is not the aim of these bodies to actively coordinate response on the ground, are there ways that they could facilitate more communication between members prior to a crisis, and even in the early days of an emergency? By building a culture that reminds organisations of their inter-agency commitments, and making sure that common documents such as the humanitarian charter or other standards are well disseminated, could improve communication and potentially the quality of interventions in the field. It could also encourage partner organisations to get in touch with each other and potentially undertake joint assessments or pool information and infrastructure.

Conclusions and recommendations

The experience of Bam provides important lessons for improving coordination among international NGOs in rapid-onset emergencies. Some key questions should guide this debate.

Can Sphere be used to reinvigorate the concept of a set of agreed designs for relief items? There is no doubt that, in Bam, if agencies had been able to show vendors three or four pre-agreed ‘Sphere-compliant’ latrine designs, items could have been more quickly and more effectively deployed. The concept of standardisation has fallen from favour recently, with most agencies stressing the need for locally appropriate solutions, rather than bringing in too many items from abroad. However, given the wide acceptance the Sphere standards now enjoy, perhaps a next step could be to adapt these into a set of designs that could be reviewed and considered for local production in an emergency. Like Sphere, these would not have the status of rules, but would rather be suggested guidelines. These would need to be adapted to individual circumstances, but they could at least jump-start a process or provide a base for concrete discussion.

Could NGO bodies like ICVA, VOCA or Interaction be more active in putting together joint assessment teams in rapid-onset disasters? Is there scope for agreements between international NGOs on this prior to the emergency, so that when disaster strikes the inter-agency mobilisation system is clear? This could improve the chances for the rapid deployment of a good-quality team with a range of skills and attributes. It could also reduce duplication in assessments and add credence to the results generated.

Alternatively, or additionally, could international NGOs be represented on UNDAC teams? While the modalities of this would need to be worked out, given that many organisations already have emergency rosters it should be possible to work out a system where NGOs could propose participants at short notice. Using the IASC or other existing coordination mechanisms could be a good place to start.

Could existing coordination bodies play an expanded role in promoting information exchange between individual agencies and between agencies and external bodies, such as the UN, donors and local governments? For example, could Interaction put members considering responding to an emergency in touch with agencies already there, to help them make an informed decision on whether to intervene? While clearly it would not make sense for these bodies to become operational in the field, they could perhaps designate focal point agencies in emergencies who could ensure that members were in touch with each other. These interventions would have to be well-designed to complement the existing UN humanitarian information systems, but they could provide a much-needed, less public coordination forum. They could also ensure that the NGO perspective is clearly articulated to donors, the government and other actors, and not eclipsed by UN or Red Cross appeals and media messages.

The example of international NGO actions in Bam suggests both a need for improvement, and a willingness to achieve this. Clearly, each of these ideas requires much further thought and discussion. However, if international NGOs were able to make progress, we would be in a better position to answer our colleague from the Iranian Ministry of Health when he asked how we could organise ourselves better in the face of disaster.

Jenty Wood is the Humanitarian Programme Manager for Iran, Oxfam GB. The views expressed in this article are her own, and do not necessarily reflect those of Oxfam GB or its affiliates. Her email address is: JenWood@oxfam.org.uk.

References and further reading

The Humanitarian Charter and Minimum Standards in Disaster Response, Sphere project, second edition, November 2003.

The preliminary report from the UN Lessons Learned workshop on Bam is available from the OCHA website at www.unocha.org.

The UN Disaster Assessment Coordination (UNDAC) Mission Report following the Bam earthquake is at www.reliefweb.org.

Nicola Reindorp and Peter Wiles, Humanitarian Coordination: Lessons from Recent Field Experience (London: ODI, June 2001).

Ben Watkins, ‘Information Exchange for Humanitarian Coordination in the Horn of Africa’, Humanitarian Exchange, no. 19, September 2001, www.odihpn.org.