IASC Inter-agency real time evaluation of the Pakistan floods/Cyclone Yemyin



The Pakistan floods of 2007 devastated large swathes of rural Sindh and Balochistan Provinces in southern Pakistan, destroying homes, crops and roads, and causing the temporary displacement of over 2.5 million people. The Government of Pakistan (GoP), through its newly-created agency the National Disaster Management Authority (NDMA), and with the help of the Army, launched a major relief operation, and the UN, with other members of the international humanitarian community and local NGOs, mobilised resources to help.

The decision was taken by the IASC Country Team (IASC CT) to launch a full scale humanitarian response. An application was made to the Central Emergency Response Fund (CERF). Clusters were set up and started work, a joint rapid assessment was carried out with NDMA, and a Flash Appeal was announced and promoted.

For a variety of reasons, the GoP did not fully support the IASC CT's decision and approach. In addition, the assessment was delayed, the Flash Appeal was issued three weeks after the onset of the emergency and raised only 26%(1) of its target, and the Clusters failed to achieve their full potential as coordinating mechanisms. As a result, and despite substantial efforts, the humanitarian community did not succeed, to the extent it considered appropriate, in delivering humanitarian relief to the already-impoverished people of Sindh and Balochistan. This RTE, staffed and operating independently of the UN, was set up to help understand the reasons, and suggest improvements for the future.

The reasons for what many saw as the disappointing results are many and various, and are analysed in the report as fully as possible after only a short 2-week visit to Pakistan. They include:

- The One UN reforms (being piloted in Pakistan), the Humanitarian Response reforms (uniquely mobilised twice in the country), and the setting up of NDMA were all still in their transition phases, and needed more time to take root, and for all parties to understand new mandates, roles and modes of operating, when the floods struck.

- There were huge expectations within the UN that the success of its own Pakistan earthquake operation, mounted jointly with the GoP and fully supported by the international humanitarian community, could be repeated; but the contexts were very different, and expectations were disappointed.

- Balochistan in particular is a highly politically sensitive part of Pakistan, and there have been restrictions on access for non-Pakistan nationals for some time on safety grounds.

- The GoP, and specifically NDMA, was uneasy about launching a full scale international humanitarian response, including the Flash Appeal.

- The UN did not really grasp the implications of this GoP unease, and decisions were made by the IASC CT (to establish 12 Clusters, for example) with the best of intentions which, with hindsight, were over-ambitious and over-complex in all the circumstances.

- Lessons from the 2005 Pakistan earthquake, particularly in relation to the operation of the Clusters, had not been learned or implemented, and many of the issues that were identified in the earthquake RTE re-emerged this time.

Our recommendations reflect both upon the implications of the "systems in transition" context, and on the actual floods response, and may be summarised as follows:

1. The UN, and indeed the GoP and the international humanitarian community as a whole for that matter, must continue to invest in the 3 new structures, so that they achieve their objectives; make greater efforts to understand each others' mandates, roles and operating procedures; and develop a real sense of partnership in working towards common humanitarian goals.

2. Common assessment tools, an effective management information strategy and systems, and shared operating procedures and contingency plans, standards and principles are needed.

3. The RC/HC needs, in some circumstances, a special budget for immediate emergency response, or fast-track access to the CERF; and the capacity for there to be a "quick Flash Appeal", followed by a later full assessment-based appeal update, could achieve greater response from donors while media attention is still focused on the emergency.

4. The RC/HC role is extremely testing, and the management and decision-making structures at country level are labyrinthine, particularly during an humanitarian response. We suggest as an option the separation of the 2 roles by the appointment of an HC (as deputy to the RC) with Disaster Management (DM) and leadership skills and experience, and that the RC should be empowered during the period of the response to exercise overriding authority over the country heads in exceptional circumstances, and if necessary for the purposes of the response.

5. The decision-making structures should be simplified by the creation of a senior level Disaster Management Team (DMT), jointly with the GoP/NDMA and a representative(s) of other agencies as appropriate, empowered to make all the key strategic response decisions quickly and effectively.

6. OCHA needs to be properly resourced for a humanitarian response in country, and quickly, if it is to do its job effectively. The general view is that this was not the case in Pakistan.

7. The lessons of the 2006 earthquake RTE, particularly relating to clusters, and reinforced by the floods RTE, should be learnt and implemented.

8. A full list of recommendations, grouped according to responsibility for considering and (if agreed) implementing them, is given at Appendix J

The issue at the heart of the findings from the RTE relates to the role of UN in a sovereign state with a strong government, and an humanitarian crisis to which the humanitarian community feels impelled to respond, but where the government does not wish to seek or receive international assistance at the level which the humanitarian community believes is appropriate.

This fundamental issue, the so-called "humanitarian imperative" is a delicate and sensitive one, and gives rise to fundamental issues of international law, as well as to passions that run deep on both sides of the argument. One person's imperative can easily become another's imperialism.

Careful negotiation and discussion is required and patient advocacy, based on good quality information and great sensitivity.


There were spells of heavy rainfall for a month over the two southern coastal provinces of Balochistan and Sindh in June 2007, aggravated by the cyclone Yemyin. (See map Appendix E). Given the nature of the terrain, vast volumes of water cascaded down hills and a sloping landmass, either towards the river Indus to the east or towards the Arabian Sea to the south. The flash floods swept across and inundated 6500 villages in 23 districts and affected 2.5 million people. The flooding caused 420 (officially confirmed) deaths, destroyed over 88,000 houses, displaced 377,000 people and destroyed or damaged standing crops, irrigation systems, schools, medical facilities, roads and bridges. Most of the area suffers from extremes of endemic poverty, with nearly 40 percent of the rural population of Sindh and over 50 percent of the rural population of Balochistan living below the poverty line(2). As a result, there were few community preparedness measures in place at the local level, beyond those they have developed over the years by the communities themselves, or by central or local government.

The international community's humanitarian response mechanisms were activated, and made a significant contribution to the relief efforts. There are, however, concerns about the way the flood relief operation was conducted. Following the practice adopted in the wake of the relief operations during the Pakistan earthquake and the Mozambique floods, an Inter-Agency Real Time Evaluation (RTE) of the relief efforts during the Pakistan floods was launched. The objectives of the RTE are two-fold: (a) to assess the overall relevance, effectiveness and coherence of the response, in the context of humanitarian reform and (b) to provide feedback to support management decision-making and to facilitate planning and implementation.

It is important to capitalize on the learning from these RTEs. The international community has had the opportunity to test out the efficacy of the humanitarian reform package twice in Pakistan. The difficulties in meeting the challenges have produced an opportunity. There is a vibrant, analytical and frank debate on the factors responsible for the successes and challenges of the earthquake and flood situations. This debate needs to be formalized, and the lessons learnt need to be institutionalized particularly to enable the UN to perform effectively as One UN in future emergency situations, especially when dealing with sovereign states in the midst of establishing new institutions. We hope that this RTE will contribute to that debate.

This evaluation is the second in an IASC-endorsed one year initiative to pilot inter-agency real-time evaluations. It was managed by OCHA New York and guided by a headquarters-based steering group of evaluation staff of participating agencies and an NGO representative. In Pakistan, an Advisory Group provided advice to the evaluation.

This report is organized as follows: Section 1: Introduction; Section 2: Systems in transition; Section 3: The Floods - assessment and response; Section 4: Financing; Section 5: Coordination mechanisms; Section 6: Moving on.

The purpose of this section is to analyze the context in which the recent Pakistan floods occurred in relation to three systems/structures, which are all in the course of major change and transformation. These are the One UN system (globally and in Pakistan as one of the pilot countries for UN Reform), the Humanitarian Reform package and its elements, and the GoP's newly established NDMA. Understandably, the pace and nature of the internal changes, and transformations within these structures, impact upon their complex interactions with one another, changing previously relatively smooth working relationships.


(1) 26% Correct on 30th September 2007

(2) Social Policy & Development Centre, Social Development in Pakistan, Combating Poverty: Is Growth Sufficient?, Annual Review 2004