Scope and method
Study III, which this chapter summarizes, examines the provision of humanitarian aid and physical protection by the international community in response to the Rwanda crisis. It combined a detailed technical assessment of the principal sectors and phases of the response with an attempt to draw out the principal conclusions and policy lessons from the experience. The period covered was broadly that from April 1994 until late 1994 for operations inside Rwanda, but for refugee operations in Tanzania and eastern Zaire it extended to July 1995. Humanitarian operations prior to April 1994 were described but not evaluated and insecurity in Burundi and the limited time available resulted in refugee operations in Burundi not being evaluated. Time pressures also obliged the study to focus on the main refugee concentrations in Ngara, Goma and Bukavu. Consequently, refugee movements into Karagwe in Tanzania and Uvira in Zaire were not considered.
The study was undertaken by a team of 21 people, representing eight nationalities and a wide range of backgrounds and experiences. Initial consultations with key UN agencies, the International Red Cross and Red Crescent Movement and NGOs began in January 1995. A reconnaissance mission by five Team members to the Great Lakes region was undertaken in April and the principal block of fieldwork by more Team members was undertaken during June and July. Within the Great Lakes region a total of 235 donor, UN, NGO and government personnel were interviewed and approximately 140 beneficiaries of assistance. These were complemented by interviews with 245 personnel of donor organizations, UN agencies and departments, the International Red Cross and Red Crescent Movement and NGOs in Europe and North America and a document collection that eventually exceeded 2,000 items. A database to enable analysis of financial flows during 1994 was created, and two sub-studies on the 1994 dysentery epidemic and UK TV coverage were commissioned.
Overview of humanitarian relief operations
The protection and humanitarian crisis of 1994 did not begin with the shooting down of the Presidential plane on 6 April, but was preceded by at least three and a half years of developing operations inside Rwanda and in neighbouring countries within the Great Lakes region. By mid-1992, for instance, attacks by the RPF in the north of the country and ethnic violence and insecurity elsewhere had created 200-300,000 IDPs. Following the February 1993 advance by the RPF, this number increased sharply to perhaps 900,000, though, by the end of the year, 60% of these had returned to their homes. Large-scale relief operations were mounted, particularly by the ICRC and the Rwandese Red Cross and WFP undertook a massive airlift of food that transported twice the tonnage carried by the 1994 airlift operations. In October 1993, the attempted coup and subsequent wave of ethnic violence in Burundi resulted in the death of 50,000 to 100,000 and an influx of almost 700,000 refugees to neighbouring countries, principally southern Rwanda and eastern Tanzania. Documentation reviewed by the study indicated that the international community's response to the refugees in eastern Tanzania was poor and exceptionally high rates of mortality were experienced as a result of the combined effects of inadequate water and sanitation, food supplies and health care. For those who moved into Rwanda the response was better as relief agencies involved in the IDP Programmes were able to rapidly divert personnel and resources to the Burundian refugees.
The events that followed 6 April were an extraordinary human tragedy consisting of genocide and civil war that caused the violent death of between 500,000 and 800,000 people, the movement of over two million Rwandese into neighbouring countries and the temporary displacement of well over one million people inside Rwanda. This study estimates that approximately 80,000 people died in the refugee and IDP camps in Zaire, Tanzania and inside Rwanda during 1994, principally from cholera and dysentery. This figure would probably exceed 100,000 among Rwandese refugees in Burundi and Rwandese outside the IDP camps if data were available for these populations.
It is highly significant that the number who died as a result of causes that could be considered avoidable (had the humanitarian response been more effective), was several times lower than those who died as a result of the genocide and conflict. The critical failings in the international community's overall response, therefore, lay within the political, diplomatic and military domains rather than the humanitarian domain. Had the international community responded more effectively in the months prior to, or in the days immediately following, the shooting down of the Presidential plane on 6 April, many, perhaps most, of those who died would probably have survived and much of the massive expenditures on the provision of humanitarian assistance been unnecessary.
Over the period April to December 1994, approximately US$1.4 billion was allocated by the international community to the response. Of this amount, approximately 85% was from official sources with the remainder being provided from private sources. By a substantial margin, the European Union (principally ECHO) and the US Government (USAID, Department of Defense and the State Department's Refugee Bureau) were the largest official sources of funds, accounting for 50% of total allocations. Approximately 50% of the total allocations were expended by, or channelled through, UN agencies, with just two agencies, UNHCR and WFP, accounting for over 85% of these. A substantial proportion of the resources channelled through these two agencies were allocated onwards to NGO implementing partners. The Red Cross Movement accounted for 17% of all flows.
At least 200 NGOs were involved in the response, but estimation of their relative role (i.e. their direct contributions and as partners to UN agencies) proved difficult as a result of inadequate data and an incomplete response to a questionnaire survey undertaken by Study III. It was clear though that many NGOs played critical roles and that overall NGOs formed an important part of the response.
Main findings
The response contained many highly commendable efforts, notably: the initial response in Ngara; the impressive performance of UNHCR Emergency Response Teams in Ngara and Goma; the work of ICRC inside Rwanda, mainly between April and July 1994, particularly in the field of protection of survivors and with its hospitals in Kigali and Kabgayi; and the courage and commitment shown by UN, ICRC and NGO personnel in extremely difficult and often dangerous situations. Widespread starvation did not occur. For the refugees and many of the IDPs the food aid supply system, dominated by WFP and to a lesser extent the ICRC, was vital to their survival and performed well. Given the magnitude and scale of the population movements and the distance of the beneficiary populations from coastal ports, this was a substantial achievement. For the non-_displaced population within Rwanda the combination of a good crop and the dramatic reduction in population meant that locally-available foods were comparatively plentiful.
Humanitarian operations in Kigali and in FAR-controlled areas after 6 April were severely constrained by the high levels of violence. Only ICRC (with MSF support) and the UN Advance Humanitarian Team were able to operate in Kigali and, though valuable, the volume of humanitarian assistance and protection they were able to provide was limited. The critical need was for security and physical protection, which the much-reduced and ill-equipped UNAMIR force was unable to provide, though it did succeed in protecting perhaps 25,000 threatened civilians. Between April and the end of June, only ICRC, CRS/Caritas and to a lesser extent WFP were able to provide humanitarian assistance in the south and west of the country, though again, the volume was severely limited. In the RPF-controlled areas in the north and east, ICRC, UN agencies and NGOs had greater access and were able to deliver quite substantial volumes of assistance, though their freedom of operation was closely controlled by the RPF and many agencies were not allowed to remain inside Rwanda overnight.
The French-led Opération Turquoise that pushed into western Rwanda on 22 June and then concentrated on the creation of a so-called Safe Zone in the south-west remained in the country for two months. The operation protected approximately 14,000 threatened civilians within Rwanda and the improvement in security in the south-west enabled a dramatic increase in humanitarian assistance activities by the three agencies that operated during the April-June period to at least 15 agencies by August. Such efforts served to spread out over a longer time period the number of displaced Rwandese crossing into Bukavu and to limit their eventual number. Had this not been done, it is highly likely that the mortality rates experienced in Bukavu would have been much higher.
Despite this, judgements of the benefits of Opération Turquoise have to be highly qualified. By concentrating forces in the Safe Zone after the end of June the operation
The positive contribution of Opération Turquoise in reducing and spreading out the movement of IDPs into Bukavu has to be balanced by the fact that the several hundred thousand Hutu who were encouraged to remain in IDP camps in the Gikongoro area presented the new government and the UN with an extremely difficult problem. Though the majority were eventually returned to their home communes, several thousand IDPs were killed at Kibeho camp in April 1995. The south-west has arguably remained the most insecure area of the country.
The response of humanitarian agencies to the needs of those concentrated in IDP camps in the Gikongoro area was initially slow as a result of: the reluctance by some NGOs to be closely identified with the French military; the time needed to establish operational capacity in the area; the focus of international attention during July and August upon the situation in Goma; and a lack of technical coordination capacity at field level. The initial lack of food and water and inadequate sanitation resulted in very high rates of dysentery in many of the camps and the death of perhaps 20,000 IDPs.
Because of the insecurity inside Rwanda and the access problems facing not only humanitarian agencies but also the international media, the large-scale movement of Rwandese into neighbouring countries enabled readier access, at the same time as creating substantial humanitarian needs. The international response to the first major influx, that of almost 200,000 into Ngara District at the end of April, which was led and closely coordinated by UNHCR, was highly impressive. Substantial loss of life was avoided.
However, despite the initial successes and the continued impressive performance of most agencies working in Ngara, the programme has remained fragile as a result of a number of factors. Unlike Goma and Bukavu, where the initial influx was not followed by new arrivals, refugees have continued crossing into Ngara. The refugee population in Ngara District in May 1995 was 500,000 - double that of May 1994. Consequently, the situation has never quite stabilized and agencies have been continually needing to increase the scale of their programmes. Another factor contributing to the post-emergency situation in Ngara was that from mid-July onwards the focus of international attention moved to Goma, resulting in the transfer of resources and personnel away from Tanzania. The water sector was one where the initially impressive emergency response was not maintained; on a per capita basis, the amount of water available to refugees by June 1995 was less than half that of July 1994.
Factors contributing to deterioration have been the constantly expanding refugee population, deterioration of emergency boreholes that were not designed or equipped for long-term service, and a lack of investment in more sustainable supply systems. Initial expectations that the refugees would repatriate, the high capital costs involved in developing sustainable supply systems and the government of Tanzania's reluctance to see investments that seemed to confirm that the refugees would be in the country for a long period, have all served to deter the necessary investments.
The number moving into Bukavu during July and August was approximately 300,000. The influx was not as intense as the initial influxes into Ngara and Goma and, because of the lack of camp sites for them to immediately move to, the town effectively served as a huge temporary transit camp until UNHCR, NGOs and the local authorities were able to identify and open new sites. A combination of the continued operation of the municipal water system, substantial levels of initial assistance from the people and local agencies in Bukavu, and the fact that many refugees arrived with disposable assets (much of it looted on leaving Rwanda), meant that disease outbreaks were limited and substantial loss of life did not occur. This result is somewhat paradoxical, because of poor overall coordination and because Bukavu received substantially less financial and human resources than were being deployed to Goma.
The Goma influx
The influx into Goma was of unprecedented scale and rapidity: in the space of just five days between the 14th and 18th of July, approximately 850,000 refugees crossed into Goma town and at points further north. The capacity of the agencies present in Goma was quickly overwhelmed despite an unprecedented and rapid response. Within the first month approximately 50,000 refugees died as a result of a combination of cholera, dysentery, dehydration and violence. Given the massive scale of the influx, many deaths were likely and the fact that there were not substantially more is a credit to the agencies involved in the response.
The study assessed the performance of the system both in terms of providing warning of the event and in preparing for a large influx. This assessment identified a fundamental weakness within the humanitarian system in that it did not possess a mechanism for monitoring and analyzing information to provide warning of population movements that was either sufficiently integrated or capable of gathering information in areas that were poorly covered by relief agencies. UNREO and its daily Sitreps came closest to performing such a role, but UNREO's capacity directly to collect information was wholly inadequate and it had to rely heavily on relief agencies in different locations providing it with any monitoring that they were carrying out. The reduced UNAMIR force was not able to monitor the situation in the north-west and the system was therefore reliant upon the ICRC operating out of Goma, whose monitoring of the build-up of IDPs was confined to the area around Ruhengeri, though within this area there were already 250,000 IDPs by early June. It was not until the first week of July, when an Oxfam Assessment Mission visited the area between Ruhengeri and Gitarama, that information became available on IDPs in this area. The Oxfam Team "discovered" another 200,000 and also estimated that another 300,000 were moving westward, following the RPF capture of Kigali.
UNHCR had deployed a substantial Emergency Response Team to Goma in April but, with the influx into Ngara, part of the Team was redeployed in early May. At the end of June, just two weeks before the influx, the remainder of the team was withdrawn and the Sub-Office in Goma reduced to a staffing level that the Acting Head of the Sub-Office termed "skeletal". Following the Ngara influx the agency had begun contingency planning measures in early May that had included the build-up of stockpiles of non-food items in Amsterdam for 500,000 refugees. The team in Goma had begun preparing a Contingency Plan for North Kivu that used a planning figure of 50,000. Identification of a contingency site was hampered by the reluctance of the local authorities to consider the possibility of a large influx. The difficulties of making adequate preparations in Goma, coupled with the fact that the Goma airport was able to cope with heavy-lift aircraft, appears to have led UNHCR to rely more on its ability to respond rapidly by air rather than on the ground preparations, such as local stockpiling. This relative emphasis on rapid response rather than on-the-ground preparedness may also have reflected the agencyÕs conception of the term ÒpreparednessÓ, which traditionally within UNHCR has effectively meant "contingency planning aimed at facilitating a rapid response once an influx occurs". This more narrowly conceived than that used by other UN agencies.
The North Kivu Contingency Plan was finalized in the third week of June. Follow-up on the numerous action points by UNHCR Headquarters and the (much reduced) Sub-Office in Goma to convert the plan into reality was slow. Staff were severely overÐstretched and a rapid sequence of events in the three weeks following the finalization of the Plan, including the RPF capture of Kigali and the creation of the Safe Zone in the south-west, generated additional work. Consequently, the contingency plan was not "ready-and-waiting" for an influx of even 50,000 by mid-July.
During June sufficient evidence was available from two sources, the ICRC Sub-Delegation in Goma and the figures being used by an interÐagency contingency planning process led by UNREO, to warrant a substantial increase in the planning figure. Poor relations between the ICRC Sub-Delegation and the UNHCR team in Goma appear to have prevented the ICRC estimate of 250,000 IDPs around Ruhengeri reaching the UNHCR Team. The UNREO-led process was initially taken seriously by UNHCR and the agency went to considerable lengths to ensure that a critical meeting in Nairobi was attended by key staff from Geneva and Goma. However, the meeting ended before it had considered the implications of the various scenarios and despite UNHCR requesting that the meeting resume the following day (a Saturday) this was not supported by representatives of other UN agencies present. After this fiasco, key UNHCR personnel do not appear to have taken the UNREO-led process seriously and the final document, which included a "worst case" scenario of large numbers of displaced moving into eastern Zaire and Burundi, was not copied to the UNHCR Team in Goma. The coincidence between the completion of the UNREO-led process and UNHCR's North Kivu contingency plan with the start of Opération Turquoise was unfortunate as the French operation quickly altered the situation and dynamic of the conflict. As noted earlier, the concentration of Turquoise upon the Safe Zone in the south-west had a critical impact on the outcome in the north-west.
In the event, the fall of Ruhengeri and the sudden increase of civilians and FAR military moving towards Gisenyi coincided with a joint DHA, UNHCR, UNICEF, IOM and UK-ODA assessment mission into the north-west that resulted in the first steps in mobilizing a major relief effort. Thus UNHCR took the decision to deploy a new Emergency Response Team the day before the start of the influx and, with the exception of a Water and Sanitation Coordinator, the full team was deployed within the next few days.
The scale of the response to the crisis in Goma was extraordinary. Prompted by intense media coverage of the influx and the subsequent cholera outbreak, the international community poured assistance into the area. The response involved not just the usual UN agencies and NGOs, but also civil defence and disaster response agencies from within donor countries, several military contingents providing support to the humanitarian activities and a large number of comparatively inexperienced NGOs. Assessed overall, the results were impressive. The speed with which water was supplied to most camps, health care facilities established and general ration distributions initiated was commendable.
However, there were several aspects of the response where performance of the system was less impressive and the performance of some agencies was poor. Almost all the non-food assistance arrived by air and so management of the airlift and the limited capacity of the airport became a critical _constraint. UNHCR played a central role in the management of the airlift operation using the Air Operations Cell in Geneva, which had been established two years previously to coordinate the Sarajevo airlift. It appears that the Air Operations Cell had difficulty adjusting to a multi-destination operation (Bukavu and Kigali were served as well as Goma), and several agencies complained that the airlift had been treated as a UNHCR airlift and not as a common resource for all agencies. Cargoes arriving did not always conform to the priorities established in the field, though this may have owed more to donors sending whatever was available rather than what had been requested. There is ample evidence also that the airlift, or at least substantial components of it, such as the US Air Force operation out of the Entebbe AirHead, continued for several weeks longer than was required.
Coordination of the arrival of critical inputs was not impressive. For instance, while the ability to pump water from Lake Kivu was quickly increased by a US private company supported by the US military, the arrival of water tankers to transport it, particularly to the spontaneously settled camp at Kibumba, which had no water sources, took much longer. Similarly, given the hard volcanic rock in the area, a critical need was for heavy equipment to construct access roads into the camps to enable the siting of health facilities and water storage and distribution systems. However, as a result of commitments by the US Army not being implemented and faulty information flows between Goma and the US Army base in Germany, it was not until the end of September that the heavy equipment capability was substantially increased.
The level of violence within the camps was extremely high, with one estimate based on a retrospective survey in one camp suggesting that 4,000 refugees died as a result of violence at the hands of the militia, undisciplined Zairian soldiers and other refugees. The high levels of insecurity in the camps directly affected the effectiveness of the relief efforts as most foreign personnel were unable to remain in the camps overnight and the ability of medical personnel to maintain continuous care of patients was hampered. The performance of the Zairian authorities and the international community in addressing the violence was also unimpressive. The fact that Western military contingents were in Goma to assist with the relief efforts but were not mandated to address the problem of in-security in the camps appeared illogical. It was not until March 1995 that a satisfactory solution was implemented involving a contingent of the Zairian Presidential Guard, paid and equipped by UNHCR, and supervised by an international monitoring team.
Many of the military contingents, civil defence and disaster response organizations that worked in Goma did so in response to a UNHCR request to donor governments to provide eight "Service Packages". This was a relatively new concept devised as a means of rapidly increasing management and implementation capacity within the system, and the intention was that individual governments should assume responsibility for entire packages. The results were very mixed, with several governments providing capacities that were broadly similar, leading to coordination problems. Within the critical water sector, for instance, there was confusion between the respective roles of the US military, the German agency Technisches Hilfswerk (THW) and Oxfam. At one point Oxfam was informed by UNHCR Headquarters that the US military was responsible for the whole sector and that the very substantial outlays by Oxfam would not be met by UNHCR.
Principal policy conclusions
The close relationship between the level of security and the effectiveness of humanitarian assistance highlights the need for coherence in the strategies adopted by the political/military and humanitarian domains. However, the Rwanda crisis has been characterized by the lack of a coordinated political strategy within the international community for "managing" the crisis. Differences between key members of the UN Security Council and governments of neighbouring countries in terms of their attitude towards the RPF and the former government and an apparent inability to confront and overcome these differences appear to have been responsible for the lack of a coordinated political approach. Despite this lack of an agreed framework, donor countries were prepared to allocate substantial resources, particularly in the second half of 1994, to humanitarian assistance programmes. This readiness with which the international community appears prepared to fund humanitarian assistance programmes contrasts with the lack of concerted efforts to devise coordinated political solutions to the crisis.
In the absence of a coherent political approach, it seems that humanitarian agencies, encouraged by the new government and certain Western political leaders, developed and pursued strategies, such as encouraging the early repatriation of the refugees, that attempted (but failed) to substitute for political solutions to the crisis. In a society that had just experienced genocide, apparently carried out by a substantial proportion of the society, the approach adopted by key elements of the international community of reintegrating Hutu refugees into Rwanda was unrealistic and broadly unsuccessful. Reports or events that questioned the new government's commitment to respecting human rights and threatened to undermine these strategies were suppressed or played down in public, though some governments did press the human rights question in private. Despite the massive loss of life and the expenditure of enormous sums of money, an estimated 1.8 million Rwandese remain in camps outside their country and many observers expect the civil war to be resumed at some point. A solution remains distant.
The Rwanda case demonstrates the need for much closer linkages between humanitarian and political policies in the principal donor countries and the UN system and also with the neighbouring countries and regional bodies such as the OAU. The creation of task forces or contact groups composed of key interested parties may serve to encourage closer linkages.
The response was resourced through a variety of mechanisms but ultimately donor organizations and donor governments accounted for the bulk of the resources provided. The extent to which funding was reactive to events was striking. There was a marked contrast in resource availability between the "tap-on" period from mid-July to September, when funding appeared limitless, and other periods, when it was less readily available. The factors contributing to this reactive characteristic are many and their relationship complex. Media coverage and the concern of almost all organizations (donor organizations and the military as well as NGOs and UN agencies) involved in the response for "profile" and "visibility" were clearly significant. What was clear from the study is that the way the system was resourced was sub-optimal, limiting the effectiveness of the response and substantially increasing eventual costs. Preparedness and contingency planning were not encouraged, a position not helped by variations in conceptualization of preparedness between agencies and donors. Investments that would have yielded substantial savings, such as opening road routes and increasing the capacity of low-cost railway routes, were not made. While donor organizations did provide some "up-front" funding this was quite inadequate in the face of such a large and highly dynamic emergency and in some cases did not even reach the levels previously agreed by donor organizations.
Foreign military forces were heavily involved in the response, with some contingents concentrating solely on provision of security, others concentrating solely on provision of relief assistance or providing support to relief agencies and several other contingents mixing these two roles. For those contingents providing relief assistance and/or support to relief agencies, logistics support (airlifting, trucking, etc.) formed the bulk of the military contribution, though several were involved in intermediate services (water production, laboratory services) and in the actual delivery of assistance to the affected population. Generalizations about the performance of the military are difficult, particularly when they were involved in such a wide range of roles and information on their impact and cost was so limited. The performance of those contingents assessed was very mixed, with some performing critical roles well while others performed poorly in key sectors and thereby reduced the effectiveness of the overall response. Information on costs that could be compared to commercial or NGO activities was difficult to obtain apart from air-lifting, where commercial companies proved considerably more cost-effective. The Rwanda experience suggests that though the military may be able to fulfil a useful role in extreme situations, their comparative advantage is often of short duration and restricted to very particular situations. Their use may be questioned from several stand points, including their predictability, effectiveness, cost and ability to participate collaboratively in operations involving several agencies and numerous NGOs.
The study reviewed available studies on the impact of the large refugee populations upon their host communities and complemented these with additional but still limited investigations. It found that within the neighbouring countries there were gainers as well as losers, with the losers often being those communities in the immediate vicinity of the refugee camps, though farmers were able to take advantage of cheap labour and the gainers often being those involved in supplying goods and services to the camp populations. The international communities mechanisms for compensating the local populations for the detrimental effects on their assets, livelihoods and environment were found to have been inadequate, with losers having to wait long periods before being compensated. In _several cases the level of services available to refugees after the initial emergency period exceeded those available to the local population. Actual and perceived differences contributed to resentment towards the refugees within the host community. In the case of Tanzania these may have contributed to the government's closure of the country's borders with Burundi and then Rwanda during 1995.
The response involved an unprecedented number of agencies and organizations and this must have increased overall costs and the difficulties of ensuring a coordinated response. The unprecedented number of NGOs involved reflects not only a genuine and widespread desire to provide assistance but also the reality that participation in large-scale, high-profile relief operations has become an important factor in the formation and development of NGOs. The performance of many NGOs was highly impressive and many cooperated closely with each other. However, there were numerous examples where this was not the case. Some NGOs sent inadequately-trained and -equipped personnel, some undertook to cover a particular sector or need and failed, and others were unwilling to be coordinated. The conclusion drawn by the study is that the current mechanisms for ensuring that NGOs adhere to certain professional standards are inadequate.
Approximately 50% of total resources allocated during 1994 were expended by or channelled through the UN system, with WFP and UNCHR accounting for 85% of these. With so many UN agencies, NGOs and other organizations involved in the relief operations, there was a critical need for a strong capacity at the centre to provide leadership and overall coordination. In regard to refugee operations, UNCHR came close to fulfilling such a role by virtue of its clear mandate, support from host governments (particularly in the case of Tanzania), highlycompetent technical coordination personnel, and control over a significant proportion of the funds available for agencies and NGOs responding to the refugee problem - in large part due to a bold decision by ECHO to channel all its funds for refugees through UNHCR.
However, coordination arrangements in relation to other areas and levels of the system were less satisfactory. The fact that the roles of the SRSG, the UNAMIR Force Commander and the Humanitarian Coordinator/Head of UNREO were limited to operations within Rwanda hampered coordination between the policies and operations inside Rwanda and those relating to refugees in neighbouring countries. Within Rwanda UNREO performed several useful functions, though it suffered as a result of its ad hoc status and lack of clarity over its relationship to DHA and UNDP, its relationship with operational UN agencies and its relationship to the SRSG. In addition it did not have adequate resources and some of its personnel (many of whom were UNDP and seconded NGO personnel), lacked emergency coordination experience. Consequently its role was limited, principally to that of information sharing. At the préfecture level within Rwanda, UNREO's Field Offices provided a useful forum for information sharing among NGOs, but technical coordination was the responsibility of other UN agencies such as UNICEF, WHO and FAO. As a result of their initial concentration on developing the capacity of the new government in Kigali, the provision of technical coordination in the south-west with its 300,000 IDPs was slow.
As well as supervising UNREO, DHA undertook a wide range of coordinating actions spanning from the initiation and leadership of the UN Advanced Humanitarian Team to coordination of Consolidated Appeals and the chairing of the Inter-Agency Standing Committee. Whilst assessment of the effectiveness of such non-operational coordination is difficult, it was clear that DHA was substantially more effective in providing coordination than it had been during the Somalia operations in 1992 when DHA was created. Nevertheless, it was apparent that DHA experienced substantial institutional and financial obstacles and faces continuing uncertainty over its future. By virtue of its lack of control over the funding of UN agencies and ambiguity over its representation in the field, it was unable, despite the best efforts of its personnel, to provide strong leadership and directive coordination. As a result of all the above difficulties in the area of coordination, and the dominance in resource terms of WFP and UNHCR, Study III concluded that the term "hollow core" was an apt characterization of the humanitarian relief system during the response.
The performance of WFP and UNHCR, the two largest agencies within the UN humanitarian system, was of critical importance to the overall response. Though the Team was impressed by many aspects of the performance of the two agencies, the relationship between them was subject to unproductive tensions stemming from the division between them of the general ration supply/distribution chain. Despite development of a detailed Memorandum of Understanding between them, these tensions persist and are likely to continue, given their different perspectives on the same pro-blems and the inherent difficulty of splitting such a critical function between the two largest agencies. One aspect of this split is that accountability is diluted as each may shift the burden of responsibility for problems encountered onto the other. Such tensions resulted in unnecessary expenditures and reduced the effectiveness of their combined actions.
Another principal conclusion drawn from the Study is that the present accountability mechanisms within the humanitarian aid system are quite inadequate. The Team found remarkable variation in the amount and quality of information on the situation in a given area depending on the agencies involved. Thus for some areas, especially the refugee camps, detailed information on morbidity and mortality was readily available whereas inside Rwanda such information was extremely patchy. In part this reflected UNHCR's clear coordination role in relation to refugees and the presence of highly-competent technical coordinators, in contrast to the unclear responsibilities inside Rwanda and the lack of technical personnel within UNREO. Thus large parts of the response could not be properly assessed, either because information on process and impact indicators was not available or it had been collected differently by different agencies. This is an unsatisfactory state of affairs. While accountability to donors is important, it should not be forgotten that relief agencies should also be accountable to the populations they are seeking to assist. The Team was struck by the very limited attempts by agencies to obtain the views of beneficiaries on the assistance they were provided with. Finally, a potentially more disturbing problem is that in a context of increased concern for profile by, and competition between, humanitarian agencies, the objectivity of their reporting may suffer as a result of their emphasis on the positive aspects of their programmes and playing down of the negative.
Note:
1. This summary of Study III, Humanitarian Aid and Effects, was prepared by John Borton.