Author: Cosgrave, J; Gonçalves, C; Martyris, D; Polastro, R; Sikumba-Dils, M.
This report summarises the real time evaluation (RTE) of the response to the Zambezi river floods and Cyclone Favio in February 2007. The RTE was commissioned by a group of Inter-Agency-Standing-Committee agencies and it formed a pilot real time evaluation for a broader programme of such evaluations. A team of five carried out the RTE in April 2007.
The evaluation was intended to have a primary focus on lesson learning rather than on accountability and the terms of reference sets out two objectives:
• in the context of humanitarian reform, assess the overall appropriateness, coherence, timeliness and effectiveness of the response,
• provide real-time feedback to support senior management decision-making and to facilitate planning and implementation.
The team opted to use focus group meetings with beneficiaries to ask them about their experience of the floods and cyclone and their views on the response. The team met with over 400 beneficiaries at 16 different sites (with multiple meetings on some sites) to ask them about their views. The team visited both flood affected and cyclone affected areas. These meetings were held using a prepared script covering what happened during the emergency, what assessments were made, what assistance they got, and their view of the future.
Findings and Conclusions
The relief response to the 2007 floods and cyclone in Mozambique was a success. While the operation was not perfect, there was no widespread suffering nor were there avoidable deaths. The reasons for this success included:
• Effective preparedness and coordination by the government.
• The impact of humanitarian reforms such as the cluster approach and the CERF.
The floods and cyclone in Mozambique were a relatively small emergency. This not only simplified the response but highlighted some general lessons about emergency response that a larger emergency might obscure. The introduction of the cluster approach was uneven, but it did add to the quality of the response. Interviewees who had experienced both the 2001 and 2007 floods, or the 2000 and 2007 cyclone responses, commented that coordination among the international humanitarian community was far better in the 2007 response.
The early deployment of an OCHA staff member from the Humanitarian Reform Support Unit helped to get the cluster system and the CERF application up and off the ground quickly. However, OCHA did not deploy a large enough team to properly support the roll out of the cluster system in Mozambique. Cluster coordination is different from normal sectoral coordination. In normal sectoral coordination, agencies coordinate over the areas where their agencies’ work overlaps, while each keeping their individual agency objectives. With clusters, the focus is not just on removing gaps and preventing duplication, but on jointly moving towards commonly agreed cluster objectives. The cluster lead role therefore demands both participatory leadership and a broad presence in the field. Some clusters achieved this and others did not. NGOs and the Red Cross movement are responsible for most humanitarian response. The participation of NGOs and of components of the Red Cross movement in clusters is voluntary. At the start of the response NGO participation was threatened by:
• Press releases and situation reports that minimised their role.
• Directive rather than participative management by cluster leads.
• A lack of transparency in dealing with applications for CERF funding.
While these issues improved over time, they were critical issues, as the cluster approach can only work if there is full participation by NGOs and components of the Red Cross movement. One specific issue for IFRC was that the procedure for activation set out in the guidelines was not followed.
Cluster leadership demands investment by the leads. They need to have suitable staff to lead clusters; the capacity to provide support for cluster coordination at field level; and the staff to promote information flow within the cluster. Not all cluster leads had a presence in the field limiting their ability to capture information from the field level or to support coordinated action. It was notable that the clusters regarded as the most successful had seen a large investment by the cluster leads. Cluster leadership places many responsibilities on agencies. Sharing these responsibilities throughout the cluster could lead to more attention being paid to some of the lead’s responsibilities that are getting less attention, such as quality control. Delegating specific cluster lead responsibilities to specific cluster members could lead to higher levels of participation, and a more even participation. Overall, the cluster approach was a success in Mozambique. It encouraged a cooperative ethos between agencies that led to a better quality and more effective response. The cluster approach is not a magic bullet, it takes investment by the cluster leads and commitment from the cluster members to make it work.
The CERF helped to ensure a rapid response, and a larger programme of assistance than would otherwise have been possible. The poor support for both the IFRC appeal and the Flash Appeal suggests that without the CERF, relief funds would have been significantly less, especially in the first month of the response. However, agencies need to be better prepared so that they can prepare their CERF request faster. Because of the time pressure, the CERF secretariat cannot vet applications for funding except in the most rudimentary way. Some clusters thoroughly discussed CERF applications before submitting them, others did not. Some of the activities funded seem much more appropriate than others and some control is needed to ensure that applications for CERF funds are of a higher quality than Consolidated Appeals have been in the past Growing local capacity is a key disaster preparedness measure. The strong performance of the Mozambican Red Cross in the response showed this. For capacity to grow, it needs nurturing with access to funds. The CERF needs a mechanism to nurture such capacity during an emergency response. In this emergency, it was very clear that there was a payback from preparedness, both in terms of community disaster preparedness, general agency preparedness, and specific preparedness for the floods. The level of contingency planning varied between agencies in Mozambique. It was led by the National Institute for Disaster Management (INGC). There is severe flooding on the Zambezi every five to ten years. This poses a risk of death and demands immediate action from outside the flooded area to prevent widespread distress and suffering. People have their traditional coping mechanisms for dealing with floods, but such coping mechanisms are not free. The government has over the years, made several attempts to resettle the population on higher ground, but the lowlands have the most fertile land. Resettlement needs to be based on an assessment of both the benefits and risks of settlement in flood prone areas.
Poverty, rather than disasters, is the real issue in the lower Zambezi Valley and in rural Inhambane. Many of the real needs of the affected population reflected structural poverty rather than the impact of the disaster. However, disasters and their related coping mechanisms contribute to this poverty, and this poverty in turn makes people more vulnerable to disaster. The HIV/AIDS epidemic threatens
poverty reduction efforts in the disaster-affected areas, and this has to be a high priority for any continuing assistance.
Registration is a common problem in humanitarian emergencies and it was particularly difficult in this response. The initial registration was not thorough, but it was made before people were attracted to centres by the possibility of aid. This initial survey was probably the most reasonable basis for planning overall levels of assistance but this did mean that some affected populations got less assistance
than they should have. The team concluded that the real needs for emergency relief were largely met for both the flood and cyclone-affected populations. However, the critical needs for the population now are not relief needs, but recovery needs, specifically the restoration for livelihoods for the flood affected, and shelter for both groups. Some aspects of the response in Mozambique represented the ideal of what a response in a developing country should be. In particular, and despite some constraints, INGC behaved as the very model of an ideal National Disaster Management Institute. This strong national coordination also helped with coordination of the international humanitarian response. Strong national coordination and good international coordination supported each other.
Three types of recommendations are presented in this report: recommendations for the IASC; for the HCT; and other recommendations, most of which are for OCHA.
The IASC should:
• develop further guidance on the scale of emergency at which use of the cluster approach is appropriate for particular sectors.
• develop a checklist for the country team to use when considering which agencies should lead clusters.
• consider introducing a requirement that staff delegated as cluster leads hold delegated budgetary authority within their agencies.
• consider making the issuing of a concept of operations a standard procedure for all service provision clusters.
• consider changing cluster ToRs to make it explicit that field presence is a requirement for all cluster leads.
• amend the generic terms of reference for clusters to promote the delegation of some of the cluster lead responsibilities to other cluster members. Such delegation could include lead responsibility for specific geographical areas.
• discuss measures that could be taken to discourage agencies from making commitments that they cannot keep.
• agree on benchmarks for each cluster. These should draw on existing benchmarks within the humanitarian sector rather than trying to formulate new ones.
• consider establishing response standards for focal point agencies who have responsibility for key cross cutting issues.
• consider introducing a deputy-lead in all clusters to improve participation by agencies from the deputy-lead’s sector.
• develop clearer guidelines for access to CERF funding by cluster members.
• consider introducing a special allowance for lead agencies that don’t have large reserves - or restrict leads to those with significant un-earmarked funds.
• introduce the requirement for automatic independent external ex-post evaluation of the use of CERF funds to encourage discipline in their use.
• consider establishing in each CERF funded response the possibility of having an umbrella fund for NGO response to be managed by the HCT.
Increase support for community and district DRR in the ongoing development programmes.
Specialised agencies should prepare viable contingency plans for the most common emergencies in Mozambique.
All clusters should play a full role in the annual simulation by INGC.
The HCT should:
• continue with the concept of the continuing IASC at country level. The current DMT should be folded into this.
• support targeted resettlement based on a risk benefit analysis for individual communities.
The HCT members should support INGC to develop the capacity to conduct verifiable registrations together with the local authorities.
The broad focus of the post-disaster phase should be on poverty reduction (and on the control of HIV/AIDS), rather than just on resettlement for some of those at risk.
The HCT and the government should agree, at the start of the response, what the minimum package of assistance is likely to be and what the qualifying conditions are, and then publish this widely, including making this information available to beneficiaries.
Members of the HCT should continue and deepen their support to INGC
HCT members should include national civil society actors in their own emergency planning, so that their capacity is developed, and they are included in funding requests.
• quickly deploy teams to support cluster roll-out in other emergencies, and support this with sufficient staff in country for information management and a field presence to support inter-cluster coordination.
• develop a more formal country-based dispute resolution mechanism for intra-cluster disputes. Such a mechanism should probably be based around arbitration with a mutually agreed arbitrator
• establish a web framework that can be used for cluster coordination in future emergencies, and provide ion-country support for clusters to use it.
• provide training for agency information officers on how to operate in a clustered environment.
• broaden its roster to include people who can stay in place for at least six weeks and who have appropriate language skills.
• increase training for potential cluster and sector leads to reach all countries at risk of disasters.
• provide strong support for the Resident or Humanitarian Coordinator for inter-cluster coordination at national level during emergency responses.
The CERF mechanism needs a clear and simple system for reprogramming grants in the event of changed circumstances.
NGOs need to provide information about their operations in a format that makes it easy to include them in situation reports.
Donors, the UN, the NGOs, and the Red Cross should support the development of strong national coordination in other countries.
PDF files require Acrobat Reader.