Review of the Cross-Mandate Approach in Ethiopia

SCOPE OF THE REVIEW

This review is intended to identify and analyse the main components of the cross-mandate approach developed in Ethiopia since 1991-1992. It examines how the guiding principles spelled out in the course of this process have influenced the planning and the implementation of assistance programmes in an operational environment characterized by the diversity and interrelation of the populations' needs. The impact and effectiveness of the initiatives that have constituted the cross-mandate approach are also evaluated.

In carrying out the evaluation, key documents were reviewed and in depth discussions were held with UNHCR staff at Headquarters and in the field. Extensive discussions were also held with officials of other United Nations agencies, donors, the Ethiopian government, and staff of non-governmental organizations.

This report is based on the field work of Everett Ressler, Consultant, and was finalized by Constantin Sokoloff, Senior Officer, Inspection and Evaluation Service.

ORIGINS AND BACKGROUND

(1) In 1991 and 1992 when the cross-mandate approach was developed, UNHCR's official caseload in Ethiopia fluctuated between 350,000 and 1.2 million persons. It included long-stay Somali refugees for whom UNHCR was attempting to provide care-and-maintenance support, newly arriving Somali refugees and Ethiopian repatriants fleeing armed conflict in Somalia for whom emergency assistance was required. Concurrently, in the east of the country UNHCR offered limited assistance to some Somali families returning to North West Somalia, and in western Ethiopia was aiding refugees from war in Sudan.

(2) The refugees and returnees whom UNHCR was attempting to assist in Ethiopia were living among a host population which, for reasons that included drought, armed conflict, forced movement and poverty, was in need of the same basic assistance and services. Selectively providing emergency assistance to refugees and returnees, when the general population with which they were living had equal needs, proved both ethically difficult to justify and impossible to implement. It created social animosity and antipathy towards the agency.

(3) Categorizing people as refugees or returnees did not in this situation have a strong social validity. It proved vague enough to cause continuing ambiguity within the organization as to objectives. Most of the refugees and repatriants were nomadic or semi-pastoralists who cross national borders traditionally and with impunity. Discrete categories of beneficiaries could not be confidently established in many situations. "Drought-affected" and "returnees" were many times the same people. "Refugees" and "returnees" were often living among, and on the basis of culture, clan and family ties, were indistinguishable from local residents.

(4) Establishing distinct categories of recipients also proved problematic in that beneficiaries understood categorization as entitlement to international assistance. Based on precedents established in both Ethiopia and Somalia, to be labelled a "refugee" was understood as entitlement to perpetual food relief and access to basic services until such time as one chose to repatriate. Being categorized a "returnee" was understood as a permanent entitlement to assistance such as transportation grant, food and reintegration assistance.

(5) Registration did not prove a valid basis for assuring that assistance was provided to those deemed eligible. The general population attempted to register, validation of claims was impossible, and any rejection was interpreted as discrimination. Delivery of individualized assistance based on ration cards proved to be an inequitable and unaccountable service delivery approach. Some of the poorest families had no access to assistance while many non-refugees gained access to ration cards through barter, sale and forgery.

(6) The provision of different amounts and types of international assistance throughout the region caused migration of people, including cross-border movement, to whichever aid delivery programme was perceived to be most generous. It created demands by various groups for equal treatment, inter-group conflict and antagonism to aid organizations.

(7) Providing assistance through relief camps was observed to have the undesirable consequence of drawing refugees and returnees away from family and clan support systems to distribution sites. In addition, large numbers of non-refugees and non-returnees joined the settlements to avail themselves of aid and services. Once recipients were established in camps, many tended not to reestablish lives outside camps as long as camp systems were available, and as long as alternative services did not exist in home villages. UNHCR was keenly aware that once camp-based services were established withdrawing or terminating them was difficult. Therefore, in spite of a sizeable flow of refugees in the Bale area of region 5 no camps were established either in Dolo organization elsewhere. Spontaneous settlements were set up, but camp organization was not made available.

(8) Provision of travel grants to returnees from Somalia to facilitate return to home villages proved ineffective in stimulating movement from relief camps where many settled temporarily after entering Ethiopia. Furthermore, the precipitous and unexpected return of more than 300,000 returnees claiming a right to the reintegration package, when only just over half that number were planned for, made it impossible to provide the generous repatriation programme originally devised. Similarly, care and maintenance programmes was believed to create a dependency on the continuation of the goods and services provided, while contributing nothing to healing the causes of beneficiaries' lack of self-sufficiency. UNHCR staff realised that unless root causes were addressed, the need for relief was potentially unending.

(9) Concerted development assistance was recognized as sorely needed in the regions where refugees and returnees were living. Developmental efforts in these areas were virtually non-existent. Underlying issues of poverty, insufficient food production and lack of employment opportunities, as well as need for water and basic services were not being addressed. While resolution of such conditions went well beyond the mandate and capabilities of UNHCR, or any single agency, they nevertheless caused distress among the local population and hampered opportunities for the successful reintegration of returnees.

(10) The use of large amounts of scarce resources for relief without contributing to the longer term sustainability of the population was considered increasingly indefensible. Unless programme strategies were altered to end dependency on international relief, UNHCR suggested that the $40 million annual cost of maintaining refugees in Somalia for 14 years might be repeated in Ethiopia. The amount allocated by UNHCR for relief activities in Ethiopia between 1988 and 1992 amounted to about $375 million.

(11) Relief and development needs went well beyond the mandate and capabilities of UNHCR. Furthermore, the involvement of other UN organizations in the region was minimal and NGO assistance was focused almost exclusively on relief services within camps. UNHCR Ethiopia concluded that UNHCR's lead agency role with regard to refugees and returnees, caused partner UN agencies to redirect their efforts to needy people in other areas, rather than the needs of the region.

(12) Correspondingly, rigid organizational response according to formal mandates, for both national and international organizations, was perceived to stymie effective problem-solving in humanitarian crises. UNHCR staff strongly held that budgetary, personnel and logistical requirements for both relief and developmental initiatives could more effectively be met through flexible, collaborative inter-organizational efforts rather than by independent initiatives dictated by narrowly defined organizational mandates.

(13) UNHCR also came to believe that its basis of collaboration with the Ethiopian government required alteration. The long-standing arrangement by which UNHCR worked almost entirely through a specialized governmental body created to address exclusively refugees concerns was observed to minimize the involvement of broader governmental services in refugee and returnee areas, thus depriving refugees and returnees, and the regions more generally, of much-needed government services. It was also observed that government refugee programmes funded by UNHCR in the past appeared to have been perpetuated rather than brought to closure. Two reasons explain this fact. Under the previous regime there was a deliberate and asserted will to divert all such humanitarian resources towards other objectives. For the new administration, there was a risk to disturb a status quo which in spite of all its imperfection held the population's attention directed at the humanitarian agencies.

(14) Complications in programme delivery, such as those mentioned above, served as the impetus for review and change by a committed UNHCR staff. Other contributing factors included the increasing international emphasis on inter-agency collaboration between United Nations agencies, and the recognized need for addressing relief needs through more developmental approaches.

(15) Consequently, in an effort to establish a more effective basis for meeting the needs of refugees and repatriants, UNHCR Ethiopia in 1992 articulated and became the principal advocate for a new strategy for aiding refugees and returnees labelled the "cross-mandate approach." Considerable inter-organizational and national discussion was generated by the proposal of this approach and the underlying relief and development problems that brought it into being.

(16) After considerable inter-organizational consultation, the cross-mandate approach was formally recognized in a memorandum of understanding jointly signed in November 1992 by the Government of Ethiopia and United Nations' organizations, represented respectively by the Relief and Rehabilitation Commission and UNDP. The cross-mandate approach thus became and remains the stated intra-organizational policy and programme framework for displacement emergencies within Ethiopia.

OVERVIEW

(17) The cross-mandate approach has resulted in substantial changes in relief delivery strategies in Ethiopia that have contributed to alleviate the plight of many. It has achieved greater equity of assistance to mixed beneficiary populations. The dependency of the displaced and the refugees on camp-based services has been reduced and many have returned to their areas of origin. In addition, it has encouraged communities to organize themselves to benefit from self-help projects and from development-geared activities.

THE CROSS-MANDATE APPROACH ORIGINATED IN THE FIELD

(18) In adopting a pragmatic course that developed into the cross-mandate approach, UNHCR has successfully managed to maximize the use of limited resources in response to a complex humanitarian crisis in Ethiopia. Faced with an explosive situation and insufficient funding, the Government of Ethiopia, UNHCR, other United Nations agencies and most NGOs had come to share a number of assumptions. They recognised basic objectives such as the need to stabilize socio-economic conditions in an area historically prone to natural and man-made disasters that generated populations displacement, and to minimize the disruptive effects of camp-based assistance. Furthermore, all parties also agreed that such objectives could be more effectively achieved through interagency coordination and cooperation mechanisms and by exercising greater flexibility in the implementation of each agency's mandate.

(19) The "cross-mandate approach" may be better understood as a set of principles for organizational arrangements and programme planning. It is a framework to guide action, not a specific programme. The cross-mandate approach has gradually developed from four central notions. First, intervention on behalf of refugees and repatriants should be a collaborative inter-organizational effort drawing on pooled resources and expertise. Second, assistance should address root causes rather than merely meet relief needs. Third, aid should be provided on the basis of need rather than on categorizations of recipients. Fourth, assistance should be provided in communities and through community structures rather than as individual or family packages provided through relief camps. Cross-mandate activities are considered in Ethiopia to be those which put into practice any of the guidelines of the cross-mandate approach.

INTERAGENCY COOPERATION WAS A MAIN FEATURE OF THE APPROACH

(20) Coordination and cooperation efforts by all the parties involved in Southeast Ethiopia have been significant under the cross-mandate approach. Regular task force meetings in the field have involved NGOs and government structures. In addition, United Nations agencies have made effective use of the coordinating capability of the UN Emergency Prevention and Preparedness Group (EPPG), which fostered cooperation and analysis, and occasionally co-managed programmes. Notwithstanding, some observers point out that while a disaster may have been avoided, cooperative programmes fell short of the rehabilitation and development goals originally envisaged.

(21) Similarly, though agencies proved more flexible with regard to their formal mandates, few programmes have earned the "cross-mandate" label. Some UNHCR staff have argued that as a field-oriented organization, UNHCR has naturally extended beyond its formal mandate while other agencies remained more reserved.

THE GOVERNMENT'S GOOD DISPOSITION PLAYED AN IMPORTANT ROLE

(22) Improvements in the delivery of assistance in Ethiopia were to a large extent bolstered by the Government's commitment to refugee issues. Its progressive asylum policy adequately matched UNHCR's protection concerns and allowed the Office to focus on assistance matters. Furthermore, the consolidation of governmental relief efforts under the relief and Rehabilitation Commission (RRC) greatly enhanced programme efficiency. The RRC provided a committed coordinating body that often had the authority to take action more effectively than line ministries.

(23) Under the cross-mandate approach, UNHCR has sought to shift to the Government some of the coordination and implementation responsibilities which it had assumed in the vacuum of the previous government Thus, the Government was encouraged to take on a more active role and UNHCR assumed a less directive and more supportive one. While overall positive, this shift has also demonstrated that adequate organizational support must be provided before reliance on national structures can be established.

MANY LESSONS LEARNED ARE APPLICABLE TO OTHER OPERATIONS

(24) While the cross-mandate approach is not a discipline, the guiding principles that developed from the Ethiopian experience may be applicable in many of the refugee situations UNHCR faces nowadays. It is undeniable that the success of the cross-mandate approach in containing a complex humanitarian crisis is to be assessed against the specific circumstances prevailing at a specific time in Ethiopia. In that sense the cross-mandate approach is not replicable. Nevertheless, forasmuch as it demonstrates that principles of action could be developed and flexibly implemented in collaborative endeavours that include the Government and other international agencies, the cross-mandate approach offers important lessons.

CHARACTERISTICS OF THE CROSS-MANDATE APPROACH

(25) The "cross-mandate approach" may be better understood as a set of principles for organizational arrangements and programme planning. It is a framework to guide action, not a specific programme. Cross-mandate activities are considered in Ethiopia to be those which put into practice any of the guidelines of the cross-mandate approach. Any activity involving interagency collaboration, for example, is likely in Ethiopia to be categorized as a cross-mandate activity, as is any assistance project, involving other players, that goes beyond relief to make a community improvement.

THE APPROACH REFLECTS A GENERAL CONSENSUS OVER BASIC NOTIONS AND PRINCIPLES

(26) The cross-mandate approach has gradually developed from four central notions. First, intervention on behalf of refugees and repatriants should be a collaborative inter-organizational effort drawing on pooled resources and expertise. Second, assistance should address root causes rather than merely meet relief needs. Third, aid should be provided on the basis of need rather than on categorizations of recipients. Fourth, assistance should be provided in communities and through community structures rather than as individual or family packages provided through relief camps.

(27) It has also been summarized as an approach based on the premise that "the effective delivery of relief and rehabilitation assistance is best achieved through well coordinated joint action on the part of the Government, UN agencies, NGOs and the affected communities whereby resources are pooled and assistance provided on the basis of need not entitlement. This 'cross-mandate approach,' ... , places a strong emphasis on the adoption of innovative, community-based interventions and working closely with national institutions to establish a real link between relief and development" (EPPG, Plan for 1993-1994, Draft, 18 May 1993).

(28) In practice, the cross-mandate approach has been developed along the following eleven major guiding principles:

1. The national government has the lead role and responsibility for addressing the needs of refugees, returnees and other people requiring emergency, recovery and development assistance. This clarification was necessary because UNHCR rather than the government had come to be seen by many as holding the welfare responsibility for returnees and refugees.

2. An encompassing governmental service, rather than a specialized governmental unit dedicated only to refugees, should hold responsibility for coordination and implementation. This aims at avoiding parallel service structures rather and instead at extending existing government services.

3. The management of emergency and recovery assistance efforts should be decentralized. The decentralization of emergency management efforts so as to involve and make local actors responsible for implementation and coordination counters a long tradition of highly centralized service delivery in Ethiopia.

4. Different assistance packages should be harmonized so that each offers similar types and standards of services. The underlying notion was that reducing differences in aid packages delivered to different groups of beneficiaries would minimize migratory movement to aid sources and discord caused by perceptions of preferential treatment of one group over another.

5. Relief, recovery and development assistance should be provided as collaborative inter-organizational efforts. If government services, UN organizations and NGOs collectively respond, relief, recovery and development needs could be met more effectively. A major component of the cross-mandate effort involves encouraging aid-giving parties to get involved in relief and recovery activities and to work collaboratively.

6. Agencies should flexibly go beyond their formal mandates when necessary to "fill gaps" to meet urgent humanitarian needs. Some organizations are able to flexibly contribute in a crisis, others do not go beyond rigid interpretations of their mandates, thereby contributing little. This guideline promoted flexibility in emergency situations.

7. In some circumstances, assistance must be provided on the basis of need without the discrimination that arises from categorization of recipients. The common practice of attempting to single out particular groups of people, such as refugees or returnees, to receive emergency assistance when the population among whom they are living is in equal or greater distress should be avoided.

8. Emergency assistance should not be provided in ways that create or perpetuate relief camp situations. This position counters what was perceived as UNHCR's standard response of providing assistance predominantly at relief camp sites. Relief camps tended to draw people to relief resources, created dependency and did not aid or encourage self-sufficiency.

9. Emergency and recovery assistance is to be provided through community structures rather than by aid organizations attempting to select and provide assistance to specific individuals or families. The community-based approach was intended to meet more equitably the needs of those who most need assistance while enhancing community/clan coping efforts. The community approach is seen as an alternative to ration card systems which have a long history of abuse in the region.

10. Rather than gratuitous hand-outs, relief assistance is to be provided in exchange for activities by recipients that enhance the well-being of all in the community. The underlying notion is that if emergency goods are provided based on participation in community enhancement projects, the assistance not only meets immediate needs but also reduces vulnerability to future emergencies. It is also believed that returnees who had settled in camps rather than in their villages of origin would voluntarily return home if community enhancement projects in association with food relief were available there.

11. Social and economic development programmes should be developed in the regions which refugees flee from and return to. The regions hosting refugees and to which returnees were attempting to reintegrate in Ethiopia are among the poorest in the country. Proponents of the cross-mandate approach argue that relief needs and difficulties in reintegration are not independent of the general poverty and lack of opportunities in the region, and that only by addressing development needs can the continuing cycle of displacement and dependency be broken.

THE APPROACH HAS ENCOMPASSED SIGNIFICANTLY MORE THAN ITS LABEL SUGGESTS

(29) The adoption of a label for the new approach was helpful in communicating that assistance was being provided on the basis of new concepts. The choice of the name "cross-mandate approach," however, has probably created unnecessary uncertainty. It has meant that in certain situations where emergency and recovery assistance is required, aiding parties must flexibly respond to needs without limiting specific humanitarian actions to certain mandated categories of beneficiaries or types of activities. However, by drawing attention only to the legal basis of action (mandate), the label does not sufficiently reflects the breadth of issues which the approach encompasses.

(30) Despite the official adoption of the cross-mandate approach and continuing advocacy efforts to explain and encourage its implementation, considerable ambiguity remains within UNHCR and among partner agencies as to what precisely the cross-mandate approach is and how it is implemented. In this context opinions tend to differ, with proponents considering it widely applicable and others remaining somewhat sceptical. This divergence of opinion appears to arise from the fact that the cross-mandate approach includes many concepts relating to quite diverse concerns. The absence of quantifiable criteria for clearly defining a cross-mandate activity also contributes to uncertainty, as do the many quite different explanations of the cross-mandate approach offered by UNHCR staff; it has been described as an approach, a programme, a concept, an operation and an attitude.

(31) As reflected above, the cross-mandate approach was envisaged by its originators as a flexible framework of guiding principles to be applied and adjusted as appropriate to each situation. The significance of this effort to the broader work of the organization, however, arises from the fact that the principles underlying the cross-mandate approach in Ethiopia address so many of the same issues with which UNHCR is grappling globally.

IMPLICATIONS FOR ASSISTANCE STRATEGIES

(32) The cross-mandate approach has resulted in substantial changes in relief delivery strategies in Ethiopia that have contributed to alleviate the plight of many. It has achieved greater equity of assistance to mixed beneficiary populations. The dependency of the displaced and the refugees on camp-based services has been reduced and many have returned to their areas of origin. Communities have organized themselves to benefit from self-help projects and from development-geared activities. Indeed, cross-mandate activities have not only improved the delivery of assistance to those who critically needed it, but have also relieved some of the emotional despair caused by the debilitating sequence of destitution, flight and dependency

NEED RATHER THAN ENTITLEMENT SHOULD GOVERN ASSISTANCE PROGRAMMES

(33) The cross-mandate approach aimed at providing assistance on the basis of need rather than entitlement linked to the categorization of recipients. In practice, this principle has been selectively applied to emergency and rehabilitation assistance when it was not possible to provide assistance to specific groups of people. This was the case for instance, when persons of direct concern to UNHCR lived among other groups of equally needy people. Although more equitable, due to the larger number of beneficiaries a smaller amount of resources was received by each party. This resulted in criticisms from groups who believed they had a right to preferential treatment and a more complete ration and from some donors who saw the lower standard of assistance as an abrogation of responsibility.

(34) Harmonizing assistance delivery under the cross-mandate approach has meant seeking greater parity of relief and rehabilitation assistance packages between different groups of refugees, repatriants and other groups of people receiving emergency and recovery assistance. It has also meant encouraging the standardization of packages among different government agencies, and UNHCR offices. Originally, the RRC's package of assistance for drought victims and locally displaced persons differed from UNHCR's various packages for people classified refugees or repatriants. Similarly, each national UNHCR office in the region offered different types of aid.

(35) Under the cross-mandate approach, assistance is not to be provided in ways that create or perpetuate relief camps. Despite political and media pressure, UNHCR and EPPG resisted establishing relief camps or centralizing the distribution of relief. As an alternative strategy, relief goods and services were provided to the whole population. No registration or categorization was attempted.

THE APPROACH TARGETED COMMUNITIES, NOT INDIVIDUALS

(36) Whenever possible, assistance was dispensed through community structures rather than directly to selected families or individuals. This was perceived as a way to more equitably meet needs, enhance community/clan coping efforts and, at the same time, avoid the abuse of individual and family ration systems. In Jijiga, for example, food was transported to distribution points in the area and handed over to community elders for allocation to the neediest families. In practice, however, simply transferring resources to community leaders did not guarantee equitable distribution. While they could now be held responsible, their legitimacy could not always be ascertained. Often the assistance provided may have been often distributed on criteria that differed from what the Organization had intended.

(37) Under the cross-mandate approach the basis for receiving assistance was altered. In a bid to reduce dependency on relief and to increase participatory action in community enhancement efforts, the government supported food for work projects as an alternative to gratuitous relief. Food for work based on community development projects was also linked to the assumption that improving services and distributing food in communities, rather than in camps, would stimulate repatriants to move to their communities. Although new both to refugee and returnee populations in Ethiopia, this assumption was supported by local leaders because in so doing they regained social recognition.

(38) Communities where large numbers of returnees would possibly return could apply for UNHCR assistance by proposing to implement community development projects. The proposal was evaluated by UNHCR field staff. Labour and material costs were subsequently calculated and an equivalent value of food provided. It was generally expected that some of the food would be sold to raise the cash needed for items which had to be purchased.

(39) The food for work initiative demonstrated that populations were interested in undertaking community development projects. Once the programme came into being, UNHCR received large numbers of proposals and experienced considerable pressure to support such projects throughout the region. A small number of projects, including clinics, schools, a jail, and many water catchment systems, were built under this programme. Not all were completed or operational at the time of the assessment but, for example, some of the schools built by local residents were, with concurrence from the Ministry of Education, in operation using local teachers paid directly by families.

(40) UNHCR, however, never made this a major initiative: no additional staff or budget were allocated. Food for work projects were difficult to administer. They required reliable managers within the communities. Conversely, moving beneficiaries to project locations was seen as disrupting the overall reintegration objective. In addition, beneficiary groups in primary areas were too small for any significant public works projects to be realistic. Finally transportation costs made for very high total project costs.

NEW PROBLEMS EMERGED THAT HINDERED REHABILITATION ACTIVITIES

(41) Shifting to village food for work projects also created new sets of problems. The concept that only villages with a high number of returnees were eligible for the aid was unacceptable to other villages who did not receive assistance. This caused conflict and some violence against UNHCR staff. Without additional capacity it was impossible for existing staff to screen all proposals. Furthermore, the considerable amount of work required to monitor and supervise food for work projects over-extended the capabilities of available staff. The lack of community development staff hindered the development of such activities.

(42) To counter the almost complete absence of developmental activities throughout refugee and returnee receiving regions, a root cause of the continuing humanitarian crisis, UNHCR made a concerted effort under the cross-mandate approach to stimulate support for regional development from national, bilateral and international sources. Despite these efforts, a substantive development aid programme for the region has not materialized up to the time of this assessment. While the needs of the area are widely recognized, bilateral donors have chosen to allocate their assistance to other regions of the country which are not only needy but are perceived to offer a better chance of recovery. United Nations agencies also failed to mobilize substantive resources for development and UNHCR has remained the major contributor in the region.

(43) UNHCR also made an important contribution in redirecting NGOs from camp services to community-based efforts. Organizations such as Save the Children Fund, who had been exclusively oriented to camp-based activities, shifted their programmes to community development and rehabilitation. Few NGOs, however, reacted immediately to UNHCR's new approach. In fact, some of the agencies providing camp services were not interested or capable to redirect their efforts to longer-term community-based activities. Instead they wished to remain "emergency agencies."

(44) UNHCR's effort to stimulate NGOs was further stymied by a government directive in mid-1993 forbidding international organizations to fund international NGOs. Weak local government structures in many areas, few national NGOs in the refugee and returnees areas and a prohibition to fund international NGOs, hampered the establishment of the mechanisms usually required for stimulating community improvement activities.

(45) Proponents of the cross-mandate approach advocate and defend its openness. What is needed, they argue, is flexibility to address humanitarian needs at hand. Accordingly, a wide range of projects have been supported. Written criteria have not yet been defined and the lack of a programme management framework in the cross-mandate approach has caused continued questioning of the approach itself and of possible organizational consequences. Some observers believe such flexibility could result in UNHCR funding projects it should not be funding or in long-term commitments UNHCR may not be ready to make.

(46) Clearly, translating guiding principles into programme management policies requires a sustained effort. To implement the cross-mandate approach, an active advocacy effort and a continuing commitment to creative policy and programme developments have proven necessary. Achieving the desired changes must be an ongoing process.

REDEFINING UNHCR AND GOVERNMENT ROLES

(47) Under the banner of the cross-mandate approach, UNHCR sought to facilitate greater governmental involvement in addressing relief and recovery needs of refugees, returnees and other needy persons in the regions where refugees and returnees are living. This required reaffirmation that the government, not the international community, holds responsibility for humanitarian assistance and for recovery. Corresponding changes in thinking, organizational arrangements and programmes were necessary.

THE GOVERNMENT WAS ENCOURAGED TO TAKE ON A MORE ACTIVE ROLE

(48) Implementing this concept involved shifting to the new government some coordination and implementation responsibilities which UNHCR had assumed in the vacuum of the previous government. In brief, the government was encouraged to take on a more active role and UNHCR assumed a less directive and a more supportive one. As part of this redefinition, the national relief commission of the Ethiopian government assumed the management of refugee and returnee affairs by incorporating the specialized governmental agency for refugees within its responsibilities. Thus the Agency for Refugee Affairs (ARA) became a part of the Relief and Rehabilitation Commission (RRC) and UNHCR became a principal partner of the RRC.

(49) Implementation of the cross-mandate approach also included a shift in policy whereby UNHCR began to more directly support line ministries to enable them to extend their services in localities where refugees and repatriants were living. Previously, UNHCR had focused almost exclusively on providing services through the specialized governmental agency for refugees and NGO partners. Under the cross-mandate approach UNHCR developed agreements with the Ministry of Health, the Ministry of Education and the Ministry of Agriculture, and provided them with financial support to carry out projects in refugee and returnee areas.

(50) As a further consequence of the shift to greater recognition of the governmental role, UNHCR field staff were encouraged to involve and work more closely with local authorities and through local task forces. This included strengthening the local consultative process and providing administrative support, transport, minuting task force meetings, and delivering basic supplies. In coordination and operations, UNHCR adopted a less visible profile.

(51) The consequence of integrating refugee and returnee matters into the national relief structure and of the government taking a more active role appears, in the main, positive. It broadens central governmental involvement in refugee and returnee concerns and enhances ownership of issues dealing with refugees and returnees. The national government assumes greater responsibility for ensuring the equitable of distribution of assistance among groups requiring relief assistance. Greater government involvement also frees UNHCR from some of the dependency that had been created by long-term support to a single government body.

(52) However, an analysis by UNDP revealed that despite the shift in emphasis, the government is still not required to contribute to the financial costs of UNHCR supported programmes. At present, refugee and repatriant programmes are completely funded by UNHCR. Some observers taking a developmental perspective argue that a greater contribution by government is necessary for ownership and capacity building. In this regard, while recognizing the financial and operational constraints of many national governments, the issue of cost-sharing of programmes deserves greater consideration.

GREATER GOVERNMENT INVOLVEMENT REQUIRES ADEQUATE INSTITUTIONAL SUPPORT

(53) Increased reliance on central and local government also brought with it complications. With greater governmental involvement, implementation efforts were often mired in problems stemming from reliance on a resource-scarce governance system in transition, with a high turnover of local officials who face innumerable difficulties implementing virtually any project. The government assumed that the concept of pooling of resources would mean that greater financial support would be put at its disposal. When increased resources did not materialize, the government expressed its disappointment in the cross-mandate approach.

(54) Depending more on national and local systems may require greater emphasis on organizational support to strengthen governmental partners. Without such support, governmental efforts on behalf of refugees and returnees are likely to be only as quick and as effective as a particular government entity is in its usual functioning. While strengthening national systems is imperative, it will not serve refugee and returnee interests to over-estimate government systems where limitations exist.

(55) Notwithstanding, the cross-mandate approach's aim to enhance involvement of the central government, to step up the participation of line ministries and of local officials in emergency and rehabilitation efforts is justified and deserves every consideration as a standard approach.

INTERAGENCY COORDINATION AS A COLLABORATIVE RESPONSE

(56) Collaborative action is a central tenet of the cross-mandate approach and in Ethiopia, to the credit of the effort, the cross-mandate approach has come to mean "working together." In situations where there are mixed populations of needy people requiring emergency assistance, relief and development needs are often so multifaceted and vast in scale that they are beyond the capacity of any single agency or government department. Therefore, collaborative action, described as the "pooling together of financial and material resources, as well as the expertise and manpower, of all agencies involved," is required.

(57) In practice in Ethiopia, the " pooling together of resources " remains a fluid concept and various implementation modalities continue to be explored and attempted. Implementation efforts have included enhanced interagency coordination, joint management of operations, joint assessment and planning exercises, and joint funding of mutually agreed upon projects.

INTERAGENCY COOPERATION WORKED OUT EFFECTIVELY

(58) At a working level, coordination between United Nations agencies in Ethiopia has been exceptionally positive. Collegial and cooperative working relationships developed between the United Nations agencies. Regular information and problem-sharing systems are operative and well supported. This spirit and practice of interagency cooperation became a reality several years before the implementation of the cross-mandate initiative, contributing to its acceptance and adoption.

(59) It is of consequence, therefore, that donor government and NGO representatives still perceive the United Nations' response in Ethiopia to be poorly coordinated. In part, this dichotomy reflects an inadequate information sharing and consultative process between UN agencies and donors. It may also reflect different conceptions of coordination and perhaps some donors' reluctance to fund an unconventional approach. Finally, donors may not fully understand the complexity and difficulty inherent to the coordination of agencies with different capabilities and programme priorities. It is of significance that UNHCR did not sign the MOU with the Ethiopian government. It was felt that in allowing UNDP to sign it, other agencies would fall into line more easily: at least as regards the principle of cooperation.

TECHNICAL COORDINATION MECHANISMS ENHANCED COOPERATION

(60) Special mention must be made of the United Nations Emergency Prevention and Preparedness Group (EPPG), a facilitator of United Nations' efforts in Ethiopia and a contributor to the formulation and implementation of the cross-mandate approach. EPPG is an interagency support unit housed within UNDP. It was established in 1987 to act as a "facilitator, consensus builder, and information broker" to strengthen the capacity of the government and the UN system to meet emergency needs in Ethiopia. In cooperation with national authorities the unit identifies needs and mobilizes resources; facilitates, monitors and reports on emergency situations; and supports efforts to reduce vulnerability to emergencies.

(61) The EPPG has played various roles in the implementation of the cross-mandate approach. It acts as the secretariat for the UN heads of agencies meeting, bringing critical humanitarian issues to the attention of the agencies. It provides authoritative technical review of emergency situations which are well-regarded within the aid community. In addition, early in the implementation of the cross-mandate approach, EPPG took a more direct role in service delivery, acting sometimes on behalf of United Nations agencies.

(62) In 1991 and 1992, for example, EPPG acting under the label "the United Nations," co-managed with the government and with UNHCR as a quiet partner the on-site emergency distribution of international goods and services for a mixed population of drought-affected residents, people felling conflict and repatriants returning to Ethiopia. As another example, in 1993 EPPG organized an interagency emergency response, under the rubric of the cross-mandate approach, to provide seeds to flood-affected local farmers in a refugees and returnee receiving area of the Ogaden. In this operation, UNICEF agreed to pay for the locally purchased seeds, WFP agreed to provide an aircraft for transport, UNHCR was asked to cover the cost of the fuel for the aircraft and the government managed the distribution.

(63) Unquestionably, the existence of this technical unit has enhanced interagency coordination, collaboration and analysis. UNHCR, in recognition of this contribution, has supported EPPG efforts. But defining the role and activities of the technical unit continues to be a balancing act. The need and usefulness of independent, non-sectoral defined assessments must be balanced with the need for each agency to carry out its own assessments.

THE UNITED NATIONS 'LEAD ROLE' IN PLANNING SHOULD NOT OVER SHADOW GOVERNMENTAL EFFORTS

(64) The need for independent assessments by the United Nations must be balanced with the desire of authorities to present humanitarian issues as they choose to depict them. The need to strengthen United Nations efforts must be balanced with the need for supporting and strengthening government's efforts. Assuming a more active management role must be balanced with the government's perception that this is a government rather than a UN responsibility. An expansion of the monitoring role of the technical unit to an emergency operations management function, however, created resistance from the government and from United Nations agencies.

(65) Interagency assessments and planning missions to emergency and rehabilitation situations are, in Ethiopia, considered part of the cross-mandate approach. Accordingly, they are more often and more purposefully organized by both the government and United Nations agencies and continue to prove helpful in developing interagency consensus about the nature of humanitarian problems and the type of actions required. Donor country representatives, who are often invited to participate, reported that participation in such visits had resulted in recommendations for their respective governments to support specific operations.

(66) Joint assessment and planning missions also carry a risk. Technical staff reported that when mission personnel agreed upon technical programmes or solutions on the basis of quick visits and without adequate technical input, the proposed solutions were sometimes inappropriate.

(67) Another implementation effort to strengthen coordination and collaboration involved the creation of directories of all relief and development activities in refugee and returnee hosting areas. Previously, it was not known which relief and development programme were being undertaken, by whom, where and when. The UNHCR Field Office in Jijiga, for example, developed a useful field directory of activities for the area and, at the national level, UNDP developed a matrix of agency support levels and activities.

A TASK FORCE APPROACH PROVED APPROPRIATE

(68) Task forces were set up to strengthen general emergency coordination and implementation efforts. Under the task force model, functional task forces are to be formed at the local level whenever management of emergency and recovery efforts are required. Participation in such task forces is open to all relevant parties: local authorities, civic leaders, representatives of UN agencies and NGOs. Whenever a local task force is established, a corresponding task force is to be formed at central governmental level as a supportive unit.

(69) The adoption of the task force approach for emergency management as part of the cross-mandate approach appears to be positive and promising. Organizing emergency management efforts around task completion, participatory problem-solving by those most directly involved in a situations, and divulging responsibility for operations to the local level, have been elements of effective action. The task force approach avoids various problems inherent to the centralized, single agency, and lead agency approaches.

(70) When a task force works well the model is enthusiastically endorsed. However, the task force approach has at least one potential weakness, also demonstrated in the Ethiopian experience. Because the formation of task forces is essentially ad hoc in that no mechanism exists to ensure that task forces are formed at both the local and central level, they are not always created when needed any may not involve the appropriate intervenors.

(71) Under the cross-mandate approach, the flexibility of mandate means either providing assistance to recipients who would under usual mandate limitations not be aided or contributing to a type of service not usually supported. In practice, whenever emergency assistance is needed which goes beyond the scale or nature of assistance for one agency, other agencies are queried as to the type and amount of assistance they can provide. Agencies either accept or decline requests for support, but have no obligation to act. This, too, has been described by United Nations agencies as the pooling of resources. While this form of interagency "pick-and-choose" response is collaborative, it has been criticized by donors as inadequate, for the sum of the response is not based on whether the full range of needs in a situation are addressed but on the willingness or ability of the agencies to respond. In some cases only token responses are made.

THE CROSSING OF MANDATES COULD HAVE BEEN TAKEN FURTHER

(72) "Crossing mandates" also includes the notion that agencies would flexibly provide assistance not usually provided by them. While all agencies agreed in principle, in practice, the response of each UN agency differed. For UNHCR the cross-mandate has meant extending assistance to non-refugees and non-repatriants in certain situations. Under the cross-mandate approach, WFP extended its mandated activities to include food distribution to the general public. It also assumed a greater operational role in food management and distribution. The mandate issue was not relevant to UNICEF because its current mandate and standard approach are broad enough to enable them to respond flexibly. FAO, however, has no standing funds to allow for flexible response but can appeal for special funds. WHO does not usually funds exceptional operation. And UNDP, despite seeking involvement cannot generally address emergency needs due to funding and administrative constraints.

(73) While the willingness of other United Nations agencies to stretch their mandates may appear to have been limited, the underlying reasons could easily be justified. UNHCR's position with regard to interagency cooperation in Ethiopia was that economic competition for limited resources was largely responsible for crating the conflicts that resulted in population displacements. Therefore, other agencies should be involved, whose mandates address durable solution and developmental needs rather than short-term relief. It may have been unrealistic, however, to expert that agencies endowed with limited resources would see developmental programmes in the Horn of Africa a higher priority. The degree of UNHCR's concern for the uprooted populations could not be matched with other United Nations agencies' longer-term planning priorities. Clearly, FAO, WHO, or UNDP were not prepared to invest heavily into a primarily unstable environment.

(74) The suggestion that regional United Nations offices be established throughout Ethiopia represents still another type of "crossing of mandates." UNDP has suggested that to facilitate and coordinate cross-mandate operations, regional offices be established and staffed throughout Ethiopia with UNDP's and other agencies' personnel. These United Nations offices would, on behalf of all UN agencies, liaise with regional authorities and operational agencies, assist in the formulation of regional action plans, facilitate information flow and provide logistical support. At the time of writing, this proposal was still being considered by national authorities and the United Nations.

(75) The relative merits of establishing regional United Nations offices throughout a country are debatable. An advantage might be that such a system could establish decentralized focal points through which the international community could more effectively channel assistance to meet local relief and development needs. The disadvantages might include the lack of funding mechanisms to support such an initiative, and possible interference with the central government and the existing UN agencies' programme structures.

REPLICATING THE CROSS-MANDATE APPROACH

(76) The cross-mandate experience in Ethiopia is of significance to the broader work of the Organization. Indeed the guiding principles of the cross-mandate approach relate to many of the issues with which UNHCR is grappling globally. The role and the relationship of UNHCR with host governments, interagency cooperation, emergency assistance and rehabilitation strategies, the relief to development continuum and durable solutions, are all major concerns for the Organizations.

THE APPROACH OFFERS LESSONS NOT A BLUEPRINT

(77) The cross-mandate approach does not, however, offer an assured blueprint. Though the concept may be sound, its applicability and success can be influenced by various factors, including social context and implementation modalities. Two major lessons concerning social context might be drawn from UNHCR's experience in Ethiopia.

(78) First, many of the problems which UNHCR has encountered in Ethiopia while providing emergency and rehabilitation assistance for refugees and repatriants suggest a clash of cultures between the international donor agencies and the people of the Ogaden. The lack of social validity for designations such as "refugee" and "repatriant," for example, is but one of many indicators that great differences existed between how the Somalis and the international community understood the problem and the best remedial strategies. UNHCR's efforts would benefit from a better understanding of the cultural and social contexts in which its programmes are set.

(79) Second, the complexity of the emergency situation in Ethiopia greatly influenced both the development and the effect of the cross-mandate approach. The cross-mandate approach was formulated and has been implemented in what might be labelled an acute societal distress situation. Acute societal distress situations may be defined as particular types of emergencies where massive and endemic humanitarian needs are combined with the lack of effective governance of governmental systems. Additionally, there may be continuing social strife, an expanding population that exceeds local absorption capacities, increasing destitution, no regional assistance plans and no community-based civic structures.

THE LESSONS LEARNED APPLY BEST TO SITUATIONS WHERE REFUGEES AND RETURNEES CANNOT BE TREATED AS DISCRETE GROUPS

(80) The cross-mandate approach raises the central question of what assistance strategies should be considered in situations when refugees and/or repatriants cannot be treated as a discrete group and when the population among whom they live are as needy, or more so. Such circumstances constitute one of the specific types of situations UNHCR faces which requires unique intervention strategies. The development of a typology of situations and of matching intervention strategies would be helpful.

(81) The originators of the cross-mandate approach have suggested that in situations in which refugees and returnees are not a discrete group and do not have particular protection problems that merit segregation, agencies collaborate in providing assistance on the basis of need rather than on the basis of categories of participants. Often, UNHCR has no valid alternative but to respond to needs. However, it may be helpful for the Organization to establish guidelines to assist field offices in developing and managing programmes accordingly.

(82) The cross-mandate approach, at present, offers no guidelines or goals as to the appropriate type or level of services for emergency assistance in non-camp situations. What standards are applicable in a community setting that includes refugees and/or repatriants, when food, water or health services are insufficient? While obviously difficult to articulate, the establishment of such a standard would provide a much needed reference point for relief activities.

(83) In this regard, the UNHCR emergency handbook has long been recognized as predominately oriented to camp-based emergency situations. A review and rewriting of the handbook to make it more community- rather than camp-oriented could be helpful.

(84) Within the cross-mandate approach, the concept of interagency cooperation for rehabilitation efforts is based on the notion that if all United Nations agencies pooled resources and expertise, the lack of development, conflict and displacement causes could be better addressed. On the basis of the cross-mandate experience, greater interagency cooperation in rehabilitation is possible but unlikely. Furthermore, cooperation alone is insufficient if recovery requires more than joint contributions toward a few small-scale initiatives. Also essential are vision, overall strategies and the development of master-plans to address the fundamental developmental needs of the poorest regions. In the face of rising poverty and social discord, there appears to be a crucial need for cooperative strategies that would address current developmental and political emergencies.