CHAPTER THREE

Rwanda's Internally Displaced:

A Conundrum within a Conundrum

Larry Minear and Randolph C. Kent

THE HUMANITARIAN CHALLENGE of Rwanda's internally displaced population represented a conundrum within a conundrum. The problem of internal displacement was a conundrum in its own right, involving issues of morality and international politics on the one hand and operational and resource allocation matters on the other. Yet the IDP conundrum was part of a larger and even more intractable set of problems. These included government-sponsored genocide and its aftermath, the obligation and capacity of the international community to assist traumatized societies, and the inadequacy of humanitarian institutions and response mechanisms.

Regarding the genocide, the international community failed from the outset to identify what was taking place, responding to a fundamental crime against humanity as if it were simply a civil war between rival ethnic groups. Having ignored the early signs of the Rwandan government's plans and having moved slowly to provide assistance and protection during the early months of the genocide, the international community–especially members of the UN Security Council–shares responsibility for the slaughter of between 500,000 and 1 million people.¹

Yet the international community was hardly more successful in resolving the conundrum of the internally displaced. Distinctions made between internally displaced persons (IDPs) and refugees and among relief, recovery,

(end of page 57)

(map of Republic of Rwanda on page 58 is not available online)

and development assistance hobbled efforts. Disproportionate attention and assistance to refugees at the expense of IDPs complicated the task of the new Rwandan authorities to reestablish a viable post-genocide polity and society. The focus on refugees concealed the extent to which their return would continue to be delayed until the situation of IDPs themselves was improved. Identification of individual Rwandans' responsibility for the crisis through equitable judicial proceedings, also a precondition for social and economic reconstruction, was not accorded appropriate priority. The regional nature of the humanitarian problem, including its connections with political and security factors, was also underestimated.

The fact that the massive humanitarian response to the major human cataclysm of the mid-199Os produced such unsatisfactory results raises troubling questions about the strategies pursued and the activities mounted. Its failure rightly leads to probing questions about the strategies of protection and assistance adopted and the institutional machinery activated. The implications range far beyond IDPs and far beyond Rwanda's borders. This chapter reviews these questions, concluding with specific recommendations for the future.²

The Genesis of the Crisis

The emergency that culminated in genocide had its antecedents in colonial and even precolonial times. Yet the immediate spark that led to the government-promoted holocaust was the shooting down on April 6, 1994, over Kigali of a plane carrying the presidents of Rwanda and Burundi. The perpetrators of the genocide were principally the Presidential Guard, the Rwandan military, and the paramilitary forces and militias known as "interahamwe." However, the morbid frenzy stirred by many local leaders and the threats made by the government's radio, Radio Mille Collines, against any Hutu who refused to participate in eliminating the Tutsi made the true number of "genocidaires" difficult to calculate.³ Like the number of victims, the number of Hutu involved in carrying out the campaign, carefully planned in the months before April, is a matter of dispute.

The Historical Context

The backdrop to the genocide was formed by generations of fluctuating tensions between Hutu and Tutsi in Rwanda and elsewhere in the Great

(end of page 59)

Lakes region. Some historians have portrayed precolonial Rwanda as a bucolic accommodation among Hutu, Tutsi, and Twa, based upon a feudal structure dominated by cattle-owning Tutsi and supported by Hutu agricultural labor. "The truth," concludes one study, "is that the present can be explained only as a product of a long and conflict-ridden process, where many factors contribute to the total picture."⁴

The long history of Rwanda is one of a series of periodic Tutsi conquests interspersed with periods of Hutu absorption. Clan hierarchies rather than ethnicity, however, characterized Rwanda's social structure until the mid-nineteenth century, when the Tutsi king, Kigeri Rwabugiri (1860-9S), assumed the throne. It was he who molded Rwanda into a Tutsi-dominated structure to consolidate his own power.

The fact that in both Rwanda and Burundi the Hutu represented the majority of the population did not deter German colonialists from perpetuating Tutsi domination when Germany established rule over the territory in 1899. On the contrary, Germany–as Belgium would do later from 1916 virtually to 1959–fostered Tutsi dominance as a means of maintaining control. From a political point of view, neither had any objection to strengthening the minority group's hold over the majority; from a racial point of view, both clearly felt more comfortable dealing with the European-featured Hamitic Tutsi than the predominantly Bantu Hutu.

In the decolonizing atmosphere of the late 1950s, the Tutsi grip on the country began to erode and Belgium shifted its support to the increasingly vociferous Hutu majority. Nineteen fifty-nine was a catalytic year in the modern history of Rwanda. It was the year of the jacquerie, or "peasants' revolt" of Hutu against Tutsi, and also the year that Belgium, for intents and purposes, adopted a pro-Hutu policy. That year and also marked by another catalytic moment–the massacre of hundreds of Tutsi and the flight of tens of thousands more across the border.

The years between 1959 and 1973 were punctuated by at least three distinct and bloody crises (in 1959-61, 1963-64, and 1973) during which approximately 600,000 Tutsi sought refuge in neighboring countries.⁵ And although there was a discernible trend by the government of Rwanda to lay the foundation for some kind of accommodation between Tutsi (both within and outside the country) and the majority Hutu. government-perpetuated exclusion and demographic reality gave the minority Tutsi little cause for relief.⁶

(end of page 60)

The sense of Tutsi frustration and hopelessness was in no small part the result of the Rwandan government's lackluster efforts to deal with the issue of Tutsi roles and rights in Rwanda. This failure in turn explains to a significant extent the motivation that led eventually to the creation of the Rwanda Patriotic Front (RPF) and its military wing, the Rwandese Patriotic Army (RPA), as well as to the onset of civil war. Between 1990 and 1994, the RPF launched incursions into the country's northwestern and northeastern prefectures from Uganda in order to deal with what was called "the refugee crisis"–that is, the determination of the Rwandan refugees to return to their homes.

Supported by France, the army under President Juvenal Habyarimana brought the RPA invasion to a halt. However, the government lacked the capacity and resources to repel the invading force. The stalemate witnessed brave and exhausting efforts by regional leaders, the OAU, and thc United Nations to bring both sides together in Arusha, Tanzania, to negotiate a settlement that would enable refugees to return and to participate in government on a mutually acceptable basis.

The Arusha Accords, agreed to initially in 1992 and reinforced by a series of cease-fire agreements and UN peacekeeping arrangements the following year, were finalized at a point when both sides had become trapped in an escalation of violence and shrill rhetoric. Nevertheless, there were formal agreements on a "broad-based transitional government," future elections, and a peacekeeping force that eventually would be called the United Nations Assistance Mission for Rwanda (UNAMIR). Yet below the surface of this diplomatic success, the day-to-day violence, and the exchange of hate messages between the government's Radio Mille Collines and the RPF's Radio Muhabura was a government solution to refugee demands. It involved annihilation of the problem and the people who embodied it.

Socioeconomic Context

The historical context that led from interaction between nomadic peo pies and agriculturists in the fifteenth century to genocide in the twentieth cannot be divorced from central socioeconomic issues. Prominent among these were demographic, economic, and regional factors. The most densely populated of any African country, Rwanda suffered from a sharp downturn during the 1980s in world prices for coffee, its major export. "Combined with the effects of the civil war from October 1990, continued

(end of page 61)

Table 3-1. Number of Rwandan and Burundian Refugees as of March 1995

demographic pressure on available resources and decreasing agricultural yields," observes one analysis, "the economic crisis introduced yet another element of stress and instability into the Rwandese political and social fabric."⁷

Ironically, Rwanda received higher per capita inputs in official development assistance than other sub-Saharan African countries throughout recent decades. The fact that substantial aid flows did little to prevent the violence–and may have even exacerbated tensions by reinforcing the exclusionary politics of Rwanda's political elite–has also given the international community pause. In any event, a return to development assistance-as-usual, even if high levels of aid could be reestablished and sustained, seems out of the question.⁸

The fact that the conflict between the Habyarimana regime and the RPF in the beginning of the 1990s was sparked and fueled by Tutsi refugees in Uganda and that Hutu-Tutsi relations in Burundi were affected by events in Rwanda underscores the intensity of the regional dynamics at play. The accompanying table indicates the numbers of refugees from Rwanda and Burundi in various countries of the region a year after the beginning of the genocide in Rwanda.⁹ The presence of refugees in each of the countries of the Great Lakes region from each of the other countries underscored the reality that in this part of the world, events do not respect national borders (see table 3-1).

The Anatomy of Displacement and Response

The effects of Rwanda's genocide and civil strife were staggering. Out of Rwanda's population of roughly 8 million at the beginning of the 1990s, some 2 million had become displaced within Rwanda's borders during the last eight months of 1994 and close to an additional 2 million

(end of page 62)

had fled as new refugees to neighboring countries. The displaced included Tutsi, some of whom had remained in Rwanda during the genocide and others of whom were among the 600,000 "old caseload" refugees who entered with the victorious RPF. The displaced also included Hutu, who, as the military and political tide turned, feared reprisals from the new Tutsi regime and army.

The majority of the almost 2 million who had become refugees in neighboring countries between April and July 1994 fled in two distinct waves.¹⁰ The first wave of approximately 300,000 crossed into Tanzania over a two-day period at the end of April 1994, as a substantial RPF force moved down from Uganda. The second wave of approximately 1.4 million fled during mid-July from what had been a temporary protection zone known as "Zone Turquoise" established a few weeks before by France in southwestern Rwanda.

The refugee population in and near camps could be classified in three groups. One was the leadership, including elements from the former Rwandan army (ex-FAR) and interahamwe, who were fundamentally responsible for the genocide and now engaged in political and military maneuvers to reassert their hold over Rwanda.¹¹ A second group comprised those who were coerced, brainwashed, or threatened into participating in the genocide. Members of a third group were innocent but were nevertheless swept up in the flood of humanity that crossed the borders in 1994. After 1994, many of this last group remained under the oppressive authority of "intimidators."¹² The makeup of the camps for internally displaced persons within Rwanda paralleled in most respects the refugee camps on the borders.

In light of the fact that a significant percentage of refugees were assumed to be guilty of crimes against humanity–or at least continued to bear arms–the prima facie refugee status accorded to all who had fled to Zaire and Tanzania was questionable. For the new post-genocide government of Rwanda, such a blanket determination was not only questionable but also highly objectionable.¹³

Major international organizations and the new authorities also had different assumptions about the composition of the IDP camps. The government assumed that a significant proportion of the IDPs, the overwhelming majority of whom were Hutu, were complicit in the genocide and "a dagger" at the government's throat. In contrast, the agencies viewed the internally displaced principally as people entitled to standard humanitarian and human rights attention and downplayed their roles in

(end of page 63)

the genocide. According to international law, the right to international protection does not extend to persons suspected of complicity in crimes against humanity.¹⁴

By September 1994 the majority of IDPs had returned to their homes or settled elsewhere. Although accurate estimates of the number of remaining IDPs in late 1994 are not available, the internally displaced were generally members of one of four major groups. The first were those who decided to remain in the former Zone Turquoise in the southwestern part of the country after French forces withdrew in July. They were unable or unwilling to cross the border but did not feel able to return to their home communes. This group numbered approximately 350,000 in September 1994 and formed the populations that crowded into some twenty IDP camps around three southwestern prefectures.

A second group represented a large but difficult-to-quantify portion of "old caseload" refugees, principally from Uganda but also from Burundi and from areas in the Horn of Africa. A substantial number settled in north and southeastern Rwanda, the former bringing with them 400,000 to 600,000 head of cattle that wreaked devastation in the parklands.¹⁵ The old caseload refugees posed a very complex problem. An embodiment of the discontent that led to the creation of the RPF and RPA and the new regime's loyal constituency, these returnees after so many years in exile had high expectations. Those among the 600,000 who lacked housing, employment, and land–or whose homes and lands had been occupied in the interim–represented a potentially explosive political and emotional issue.

The third group of IDPs was more amorphous and difficult to quantify They were the impoverished and dispossessed in one of the poorest countries in the world. They included innumerable street children, those traumatized by the war, and the destitute, all of whom had been uprooted and received no assistance from a barely functioning social safety net.

Finally a fourth group were "rescapés," principally Tutsi who did not flee the genocide but chose to stay in the country even during the massacres. Ironically, these "survivors" were objects of suspicion by Tutsi who had felt compelled to flee. Some became objects of revenge by those who feared that the survivors would pinpoint the "genocidaires." Often the only recourse for the rescapés was to abandon their homes and seek shelter in different prefectures. They, too, became part of Rwanda's displaced population.

(end of page 64)

By and large, the IDPs in the camps were neither the government's natural constituency nor a priority item for attention or resource allocations. If there were any for whom the government had sympathy, it was those from the old caseload group, a sympathy not fully understood or appreciated by the international community at large. Conversely, the international community's preoccupation with the displaced from the former Zone Turquoise baffled the authorities.

Internally Displaced Persons and Refugees:

An Intricate Interrelationship

The refugees who had fled to Tanzania and Zaire in the summer of 1994 triggered an immense international relief effort. The initial commitment to assist almost 2 million people generated resources of some US$1.1 billion between April and December 1994, excluding the costs of military contingents and direct contributions to NGOs. For the year beginning in January 1995, resources were provided at the level of about US$1 million per day.

The refugees commanded the media limelight and the lion's share of the resources, upstaging the difficult straits of the displaced within Rwanda's borders. The media focused intensely upon their plight, their exhaustion, and their losing battles with cholera. Little attention was paid to conditions in Rwanda or to those who had survived the genocide or, for that matter, to the interrelationship of the 1994 realities of Rwanda, refugee repatriation, and IDP return.

Of the officially reported $1.1 billion in aid available for the crisis in 1994, Rwanda itself received an estimated $372 million, or about one third. Of this sum, $50 million was allocated to Rwanda between January and April 1994–that is, before the genocide. The $372 million also included funds earmarked for IDP camps in Rwanda, funds that, to the distress of the Rwandan government, were not available for use elsewhere in the country.

Refugee-focused media attention and humanitarian activity isolated a single element of a complex problem that in reality linked together refugees, IDPs, and the state of Rwanda itself. Refugee repatriation was closely tied to the return of the IDPs to their home communes, the latter being a sort of vanguard for the former.¹⁶ Yet the willingness of IDPs to return to their communes depended upon the extent to which they felt

(end of page 65)

that security could be ensured and they could return to property that had not been confiscated and to a life reasonably free of persecution.

Such security depended in turn upon a functioning system of justice under which the innocent were protected and the guilty prosecuted. Also essential were central and local administrative systems and the delivery of services. Priorities included the restoration of damaged infrastructure, including roads, water and electricity systems, and housing, along with the provision of basic health and education.

Yet these components were not approached as integral elements of a common whole. No single international actor present in the Great Lakes region had the authority or, for that matter, the perspective to develop policy that transcended the borders of individual states. While the special representative of the secretary-general (SRSG) from his base in Rwanda undertook regional missions on the secretary-general's behalf to Kenya, Tanzania, and Zaire during late 1994 and early 1995, the resources of the UN system–including those of the peacekeepers, for which he was also responsible had no overarching strategy or capacity to deal with the intersecting elements of the problem.¹⁷

Similarly, a regionwide meeting of governments, multilateral organizations, and UN agencies convened by the Organization for African Unity and UNHCR in Bujumbura in February 1995, represented a step in the right direction. Yet the meeting, which after much debate adopted a plan of action focusing on refugee return, did not address the broader issues that the crisis of the Great Lakes region required.

Misperceptions and Misunderstandings

To one extent or another, most humanitarian and peacekeeping operations experience problems in relating to the host authorities. These often stem in part from the government's misperceptions of the motives and intentions of the intervening actors, and vice versa. Problems reflect the highly charged nature of complex emergencies and also the reality that in attempting to bring stability to highly volatile situations, international actors may preempt, or appear to preempt, decisions by the resident political authorities.

When the new Rwandan government assumed power in luly 1994, it was perhaps inevitable that the advice of the United Nations about a range of matters from economic policy and priorities to control over the Zone Turquoise would be viewed as co-opting Rwandan sovereignty. Yet

(end of page 66)

the depth of misunderstanding that evolved was never anticipated or even recognized. The international community never fully comprehended the consequences of the genocide or the resentment harbored by the government and most Rwandans toward those whom the UN Security Council belatedly dispatched to preserve the peace. For their part, the Rwandan authorities showed scant regard for the dilemmas in which the peacekeepers and UN and other humanitarian organizations found themselves. The misunderstandings are crystallized in the issue of internal displacement.

DIFFERING PERCEPTIONS OF THE PROBLEM. There was a fundamental perceptual clash among major actors over IDPs and refugees. While the government regarded a considerable portion of the displaced as criminals responsible for the genocide, the humanitarian community took a position that reflected its mandated responsibilities for assistance and protection.

Determined to exact justice, the government did not find the humanitarian claims of IDPs and refugees and of those who articulated them compelling. The IDPs in particular were seen to be occupying vital agricultural land and faring far better than local people who were not receiving international aid. Furthermore, the administrations being established in the communes and prefectures where the IDP camps were located saw the international relief organizations and operations as challenges to their authority. There was persistent concern that IDP camps were the spearhead of the ex-FAR to regain control over Rwanda.

For their part, humanitarian organizations were on hand first and foremost to provide emergency assistance. Human rights issues were not of overriding concern, although, not having been present during the genocide many were more concerned about what they saw as Tutsi violations of Hutu rights. Initially they were reluctant to reduce food rations and other services as a means to encourage IDPs and refugees to return to their home communes, viewing such actions as causes of additional hardship. Having worked diligently to assist the uprooted, many relief workers developed bonds with them. Over time, the approach of the agencies reflected a growing "clientism." Humanitarian organizations showed themselves reluctant at first to rethink their traditional aid strategies in light of the special circumstances related to the genocide.

The sharp clash of perceptions spelled trouble for both the government ; and such organizations. In the government's eyes, much of the population

(end of page 67)

perceived by aid agencies as "relief beneficiaries" were "genocidaires." That avowedly neutral and impartial humanitarian efforts to assist uprooted people would be read by the authorities as highly political was not, of course, unprecedented. In this instance, however, the postgenocide landscape on which they worked frustrated agency efforts. Conversely, the ambivalent commitment of the authorities to equitable treatment of the uprooted Hutu populations received particularly close international scrutiny.

THE INTERNATIONAL MILITARY AND ITS MANDATE. The crisis in Rwanda represented a watershed in the number of national military contingents involved, in the permutations of multilateral and unilateral authorities under which they served, and in the types of humanitarian support functions they performed. Yet in a broad sense, the deployment of military assets showed the same imbalance–in timing, scale, and effects–as the humanitarian resources. Their net effect, not surprising although not planned, was likewise to complicate the challenges faced by the new Rwandan authorities.

When outside military forces were most needed–that is, during the onset of the genocide–the Security Council moved to reduce their ranks. The UN's peacekeeping presence, the principal international military presence within Rwanda during the early months of 1994, shrank from 2,500 on April 6 to 270 on April 21. Although on May 17 the Security Council authorized an increase to 5,500, only 1,257 soldiers were in place three months later. The response reflected a lack of understanding about what was actually taking place on the ground, the concern in national capitals about the safety of peacekeeping contingents, and the cumbersome international response machinery.

Despite limited ranks and lack of political support, UNAMIR troops in Kigali in the early weeks worked energetically to protect civilians and to support the work of the few aid groups on the scene. They sought to shelter vulnerable Tutsi in the Kigali stadium, patrol tense neighborhoods, and accompany or rescue beleaguered aid teams around the city. UN troops were later joined by French soldiers under Operation Turquoise and then by national contingents responding to requests from UNHCR for specific packages of services.¹⁸

During the Rwanda crisis, the presence and accomplishments of the panoply of international military personnel were far more evident during

(end of page 68)

(Figure 3-1: International Troops Responding to Rwanda Crisis, 1994) is not available online)

the mass exodus to Zaire than during the dark days of the genocide. In retrospect, however, and despite media portrayals from Goma and Bukavu to the contrary, the troops did not contribute in a timely fashion what was most needed and what was most beyond the ability of other actors to provide: a climate of security in the refugee camps within which humanitarian activities could be carried out. However visible in Goma during the peak of the emergency from July through September, most troops had left the region by the time that disturbances in the refugee camps threatened refugees and aid operations alike. Assistance from the military, as from the aid groups, went disproportionately to refugees outside of Rwanda, to the detriment of IDPs within (see figure 3-1).¹⁹

In the aftermath of the RPF/RPA victory in Rwanda, UN peacekeeping forces by October had attained their requisite force strength. More than 5,500 troops and civilian support personnel were stationed in Rwanda with a mandate to monitor the security situation and assist in

(end of page 69)

the repatriation and protection of refugees and returning IDPs.²⁰ By then, the logistical and security challenges that had deterred the commitment of military assets to Rwanda itself had been reduced.

Throughout most of the period from October 1994 through its December 1995 mission completion date, UNAMIR was in a difficult–many would say invidious–position. Its mandate was viewed by many in Rwanda as protecting the very people responsible for the genocide. UNAMIR, well aware of such perceptions, did all in its power to counter them. For example, it joined forces with the RPA to arrest intimidators in the Kibeho IDP camp.²¹ UN peacekeepers also deployed their meager assets to support police training schools, infrastructure reconstruction, and even prison expansion programs. Yet such efforts could not overcome the prevailing contempt for an interventionist force perceived to be protecting the perpetrators of genocide.

THE HUMANITARIAN COMMUNITY AND ITS MANDATE. Humanitarian organizations affirm an obligation to provide assistance to those in need. In the case of Rwanda, however, the community of aid organizations faced a dilemma that affected nongovernmental, bilateral, and multilateral institutions alike: whether "they should feed the victims at the risk of giving the killers new strength."²²

In late fall 1994, Médecins sans Frontières-France answered the question in the negative, terminating its programs rather than providing assistance. Other agencies weighed the options differently, reasoning that the presence of killers in the camps did not absolve them of their humanitarian obligations. By November 1994, however, some that had decided to stay the course announced that they would withdraw from the Goma camps until adequate humanitarian space for effective operations was reclaimed.²³

The government of Rwanda clearly supported their decision to withdraw from Goma, for it saw humanitarian efforts there as inappropriate and contradictory under the circumstances. Its view of similar humanitarian efforts in the IDP camps was equally negative, if not even more so, for it maintained that the IDPs had no compelling reason to remain in camps. With the conflict ended and the killings stopped, the continued presence of IDPs in the camps was tantamount to a confession of complicity in genocide. The government's position that the innocent had nothing to fear in returning home ran counter to the fears IDPs expressed in the camps and to rumors of revenge killings heard from the communes.

(end of page 70)

Humanitarian organizations remained skeptical of the government's view of stability throughout the country and security at the local level. Humanitarian agencies saw their constituency consisting of frightened people with little hope about the future, few resources, and no means to provide for themselves.

Such contrasting perceptions were clearly paralleled in the refugee camps. The positions adopted by international relief workers there as well as within Rwanda created a deep and unbridgeable schism between a significant portion of the international community and the Rwandan government. International assistance–the major vehicle for outside intervention in the crisis–was perceived as making the government's assertion of its authority and performance of its tasks far more difficult.

THE COORDINATION MANDATE. A review of coordination arrangements among the major international actors calls into question whether senior UN officials at headquarters fully appreciated the disarray caused by their unwillingness to make basic institutional decisions about field structures. Fully eighteen months into the crisis, for example, no decisions had been made about the relationship between the special representative of the secretary-general and the UN humanitarian coordinator or about UNAMIR's relationship with the humanitarian community. The ability of agencies to adjust to the environment in the field, an environment itself highly subject to change, proved limited.

The relationship between UNAMIR and the humanitarian community suffered from a basic lack of understanding of each about the working methods and objectives of the other. The UN military, which had a reasonably clear organization and structure, was bemused by the seeming chaos among the UN agencies, let alone NGOs. As frustrations grew within UNAMIR over its difficulties in achieving productive working relationships with the RPA, some peacekeepers conveyed the view that these antagonisms would not be so great if the humanitarian community were assisting the Rwandan government effectively.

For its part, the humanitarian community was at times exasperated by UNAMIR's efforts to become involved in humanitarian activities, an area in which it had little expertise or comparative advantage. Many aid workers viewed the peacekeepers as underemployed, unreliable, and insensitive to the complexities of the situation facing post-genocide Rwanda. Certain UN agencies were also wary about the respective roles of the SRSG and the humanitarian coordinator. UNHCR for one was concerned

(end of page 71)

that either might enter areas of its own specific mandate. Hence suspicions had to be surmounted even when the need for full cooperation was acknowledged.

A complicating factor was a divergence in priorities. Not all organizations were convinced that in post-genocide Rwanda IDPs should be the primary issue of concern. Many believed that primary attention should be given to restoring Rwandan society, promoting national reconciliation, and dealing with the endemic poverty that had predated the war. Differing perceptions about priorities also affected the development of common strategies for addressing IDP needs.

BILATERAL CROSSCURRENTS. The fundamental issue for many governments concerned the legitimacy of the new authorities, including their capacity to survive and carry out the requisite functions. Some governments were prepared to give them the benefit of the doubt and sought to provide assistance as quickly as possible to strengthen their hand. The Netherlands, for example, perhaps the most creative and effective donor in the circumstances, saw the need for disbursing funds quickly with few administrative requirements. Other governments moved with less dispatch, overlaying funds with heavy conditionality and delaying disbursement. The French lobbied other donors against providing substantial assistance for reconstruction, an approach widely viewed as playing out a bilateral political agenda in the region.²⁴

Such crosscurrents in the disparity of views and interests among major donors had two important consequences for IDPs and the crisis as a whole. First, there was insufficient donor harmony to support the broad UN humanitarian objectives. Hence concerted pressure by donors to constrain the RPA's approach toward IDP camp closure and commune security could not be relied upon.

Second, there was little concerted action to address problems that were highly dependent upon resources and administrative efficiency. This was sadly evident in the results of the U.S.-driven Rwanda Operations Support Group (ROSG). Despite occasionally productive dialogue, the ROSG, comprising all major donor representatives at senior levels, failed to move on fundamental issues affecting Rwandan operations.²⁵ No specific proposals were adopted to ensure cross-border security or to stem the flow of arms into the Zairian refugee camps.

Despite pledges made during two 1995 Round Table exercises convened by the government and supported by the United Nations Devel-

(end of page 72)

opment Programme (UNDP), neither the level of actual resources nor the speed of their delivery increased significantly. Here the ROSG could at least have pushed either to control the level of soaring Rwandan government expectations or to make good on well-intentioned donor commitments. The gap between the expressed concerns of donors for Rwanda and their actions exacerbated the sense of alienation and grievance felt by many Rwandan officials.

This gap also led to a peculiar donor dynamic that frustrated the Rwandan government as well as aid officials. In all too many instances, donors pledged resources to fund vitally needed programs and, when funds were delayed, then imposed conditions reflecting lack of confidence in the programs. To the embarrassment of several major donors, the urgency of translating pledges into disbursements was either lost on their capitals or a victim of cumbersome bureaucratic procedures.

The justice system was a case in point. Lack of resources, including some pledged by governments, greatly delayed the establishment of procedures for trying those suspected of participation in the genocide. The delays resulted in an increase in "rough justice," the rise of which in turn led donors to specify that funds would not be released if such "extrajudicial procedures" continued. Whether or not the rise of conditionality was a smoke screen for more sinister intentions, it reflected the profoundly disjointed approaches of bilateral donors.

Conventional Responses to Unparalleled Circumstances

Although misperceptions took their toll, the international humanitarian response was also marked by operating procedures and institutional conventions unsuited to the situation. Inadequate commitment, institutional incapacity, and lack of practical post-conflict recovery strategies were particularly problematic.

Inadequate Commitment

A persistent complaint of those who deal with internal displacement issues is that there is no single organization with clear-cut responsibility for IDPs. In Rwanda the UN humanitarian coordinator was given indirect responsibility for resolving the IDP problem but not the resources for doing so. The lack of resources would not necessarily have been crippling had the UN agencies been instructed to support him actively

(end of page 73)

with personnel and material aid. Yet while credit must be given to individual agencies for assisting his initiatives, there was little institutional will or flexibility of mandate to provide the requisite human and material resources.

An instructive example of the strengths and weaknesses of the response was the Integrated Operations Centre (IOC), established in late November 1994 in the Ministry of Rehabilitation and Social Integration to provide an integrated approach for returning IDPs to their home communes. The IOC tackled four tasks.

First, it sought to deal with the political dimensions of IDP issues by adopting a community-wide strategy. A particular aid agency taking the lead could be singled out by the government for failure or by the humanitarian community for complicity in violating fundamental human rights; the IOC provided protective cover and a degree of security based upon common agreement and approaches.

Second, the IOC sought to engage the government in the planning process and in implementing agreed upon strategies. If the authorities remained unengaged, it was likely that the government would either act on its own–perhaps creating a rift between itself and the aid community–or leave the issue of the IDP camps to the agencies, opening them up to accusations of connivance with the perpetrators of the genocide or of incompetence.

The third challenge was to restore the balance between peacekeepers and the humanitarian community. Although given the contentiousness of closing the IDP camps the matter might have been left to UNAMIR, the humanitarian community felt obliged to become engaged in this issue, whatever the hazards. The agencies were concerned lest the political and peacekeeping roles of UNAMIR dictate the methods used to deal with the IDPs. They also believed that, with relative peace restored, the time was right to make clear that the peacekeepers were in Rwanda to support the work of humanitarian organizations.

Finally, the IOC was an instrument for addressing the practical dimensions of the IDP problem. In a world used to providing humanitarian assistance to millions, attention to an estimated 350,000 IDPs might seem a relatively simple exercise. In fact, the numbers were relatively large, and the complexity of the operation necessitated a fully integrated approach. At issue were not only humanitarian assistance in the camps and transport home but also preparations in the communes to receive returnees and publicity campaigns to encourage return.

(end of page 74)

Sensible in concept, the IOC encountered various difficulties, many from the agencies themselves. Although the International Organization for Migration (IOM) and a host of NGOs were fully on board, the UN organizations themselves were not. Despite considerable resources, UNHCR felt that its mandate would not allow it to commit its energies fully to IDP programs. The World Food Programme (WFP) was engaged in a wide range of activities, including but not limited to IDPs. Like UNHCR, it was concerned that the IOC solution fell between institutional stools: that is, it was neither multilateral nor governmental but a hybrid. UNICEF, too, though heavily involved in the IDP camps, had nationwide commitments and priorities well beyond IDPs. It was also reluctant to get involved in IDP solutions that might have negative human rights consequences.²⁶

For its part, UNAMIR was fully engaged, although it temporarily withdrew support for the IOC in the midst of the Kibeho crisis of April 1995 (described below). The peacekeeping operation's mandated responsibilities for the IDPs made its involvement institutionally simpler and, far more than the individual agencies, it had personnel experienced in "operation-room" activities as well as spare computer and communications equipment fundamental to the IOC's information and operational functions.

From the UN side, the IOC thus fell short of its promise. Despite its efforts to facilitate cooperation, it encountered restrictions imposed by agency mandates and by lack of priority to IDP issues in agency headquarters and on the UN's Inter-Agency Standing Committee. Also problematic were uncertainties that stemmed from venturing into unconventional institutional arrangements. Rather than becoming an innovative and effective instrument for dealing with IDPs, the IOC mirrored the shortcomings of the agencies and the government in planning, although the humanitarian operations it orchestrated were highly effective in response to the Kibeho crisis.

Whatever the deficiencies of the IOC with respect to the agencies, its fundamental and almost fatal flaw resulted from its inability to bring the Ministry of Defense and the RPA into its orbit as active players. The only occasional participation of their representatives reflected a lukewarm institutional commitment. The assumption that the endorsement of its plans by the Ministry of Rehabilitation and Social Integration, which was ostensibly responsible for IDPs, and more particularly the minister himself, was sufficient ignored political realities within the government.

(end of page 75)

For all practical purposes, the Ministry of Defense and the RPA were the government in an otherwise enfeebled structure. They rapidly lost patience with the slow pace of returning IDPs to their communes and generally supported only those specific proposals they had endorsed in advance. The IOC's assumptions about the importance of the Ministry of Rehabilitation did not take into account the real intentions of the Defense Ministry in delaying camp closure in November 1996. The delays, it turned out, did not represent a basic change in approach but only a tactical concession to the international community, upon whose support and resources the post-genocide government depended.

Institutional Incapacity

While it does not necessarily hold that the problems of displacement in Rwanda and throughout the Great Lakes region would have been solved if the UN's institutional structures had been more coherent, those structures did little to facilitate international efforts.

Within Rwanda itself, there was lack of clarity about relationships among the SRSG, UNAMIR, and the UN humanitarian coordinator. Ostensibly, UNAMIR's force commander reported to the SRSG, but in effect UNAMIR's operational requirements and resources led at best to a very loose-knit UN command-and-control structure in the field. Similarly, although the UN humanitarian coordinator was also to report to the SRSG, his primary accountability was to the under-secretary-general for humanitarian affairs. In other words, the largely harmonious relations among these three entities did not translate into the operational coherence essential for moving IDPs from the camps.

Moreover, UN headquarters played too large a daily role in the affairs of UNAMIR and, for that matter, of the SRSG. The able and experienced UNAMIR force commander on crucial occasions was prevented from using his best judgment in responding to evolving challenges in the field by officials in New York who were less aware of local realities. For example, written instructions received from headquarters just before the Kibeho massacres forbade the force commander from using peacekeeping troops to intervene between IDPs and Rwandan army soldiers. Despite this order from New York, the force commander felt morally obliged to keep UNAMIR's Zambian battalion at Kibeho to maintain a presence that might lessen potential tensions.²⁷

(end of page 76)

At the same time, the SRSG, an engaging individual with decades of diplomatic experience, needed greater control over the overall UN structure in Rwanda. He required authority to mobilize the UN peacekeepers and humanitarian organizations quickly and coherently to plan for moving IDPs. He also needed resources to implement recovery and stabilization activities quickly, which required having a trust fund in Rwanda rather than access to funds only through the tortuous administrative procedures of headquarters. Each of these would have contributed in a major way to the UN's overall response.²⁸

The fact that the problems of refugees and IDPs were part of the same complex web of issues should also have spurred greater interest in a more permanent structure for dealing with the regional issues. UNAMIR's portfolio was limited to dealing with problems within territorial Rwanda, although on occasion in 1994 it did take a look at the evolving problems in the refugee camps outside. The humanitarian coordinator's brief likewise stopped at the border; UNHCR's priority tasks began there. The division of labor among UN organizations did not correspond to the regional dimensions of the problems at hand.

For various reasons, including sovereignty as well as institutional rigidity within the UN, more comprehensive approaches were not seriously attempted. Although UNHCR had its own special regional representative, it was not until November 1995–eight months after the initial proposal–that a regionwide information structure was established, the UN's Integrated Regional Information Network (IRIN). IRIN reported for the first time on all aspects of political, economic, humanitarian, and military events throughout the region that affected IDPs and refugees.²⁹ As a source of regionwide and systemwide information, IRIN was successful. Yet the international community needs to go beyond creating a center for information exchange and develop a more proactive and creative regional approach for interagency strategizing and implementation.

Lack of an Effective Strategy

The interlocking relationship among IDPs, refugees, and conditions in Rwanda required that the conditions at the local level be improved enough to attract people to return to their homes. But creating such conditions would have required the international community to understand and support post-genocide recovery. Though donors expressed sympathy for the government's burdens in rebuilding the nation, they

(end of page 77)

were by and large unwilling to provide appropriate assistance in a timely manner.

Rwanda's needs did not fit into conventional aid systems or concepts. Rwanda was indeed in an emergency, but the principal victim of the emergency was the society itself. Blankets, nutritional feeding, plastic sheeting, and other emergency relief supplies were needed, of course, and for the most part were provided for a wide range of vulnerable groups, IDPs, and refugees. But also needed was a rapid but sensitive approach to societal stabilization. Such unconventional emergency relief required the international community to review its very conception of "survival needs."

It raised politically sensitive issues about the criteria of "relief" and challenged the procedures by which donors pledged and committed funds.

While massive funding poured into refugee camps across the borders, Rwanda itself became even more impoverished. Because the climate in Rwanda was so volatile and its government so heavy-handed, donors were unwilling to respond quickly to essential priorities there–the restoration of basic services, including electricity, water, and transportation, as well as social support such as health and education. As UNICEF and a variety of NGOs soon discovered, the needs of a traumatized society and of shattered families and orphaned children required special attention.

Equally conspicuous by their absence in the wake of the genocide were the essentials of law and order and security. The justice system had been destroyed. There was no police system; law and order was in the hands of an army that was essentially a guerrilla force that only recently had emerged from the bush. Local and central administration had been eviscerated as buildings were looted and files destroyed when the former regime retreated. A high proportion of civil servants were dead, internally displaced, or living in camps in Tanzania or Zaire. There were few facilities and personnel to register returning IDPs; where they did exist, staff usually lacked paper or pencils to perform even the simplest administrative tasks.

The situation was nothing less than an emergency, a society on the verge of collapse. While discussions ensued in political circles about the legitimacy of the new Rwandan regime, several key donors were equally perplexed on a more technical level about mechanisms for providing resources suitable for stabilizing the situation. Three problems loomed large.³⁰ The first was determining what types of support could actually

(end of page 78)

be provided. Prisons, police forces, and military facilities–even salaries for civil servants and soldiers–were essential. Yet these items were regarded as too politically sensitive and in any event were not within the standard portfolios of aid groups on the scene at the time.^3l

Second, less politically sensitive needs, such as the justice system, were problematic. What was required? Courthouses? Legal texts, judges and magistrates, defense attorneys and prosecutors? More traditional adjudication systems? Funding for salaries, technical assistance, or training? The uncertainty stemmed in part from the government's own lack of clarity about its requirements. But there were more fundamental concerns not fully appreciated by outsiders. The government wanted "to put something in place" as a clear demonstration that there was a system, however basic, to which accusers and accused had recourse. Yet the urgency was lost amidst standard development discussions and broad-based projects. As a result, the most obvious change in the judicial system between late 1994 and late 1995 was an increase in the prison population and the explosive potential of the situation.

Third, the response mechanisms of even well-intentioned donors were too slow and cumbersome to provide an effective stabilization program. Such programs fell between donor stools, being neither traditional relief nor traditional development activities. There were no "funding pockets" that would readily meet such requirements. The absence of quick disbursement resources was a source of obvious frustration to the government. Meanwhile, donors–some with ambivalent political agendas– faulted the government's maddeningly ineffective ministries for their inability to deal with the surge in local violence and the resettlement imperative.

Misperceptions and Conventional Responses Converge

The April 1995 massacre of several thousand IDPs in Kibeho camp by the RPA was, in a sense, the culmination of months of misperceptions and conventional responses.³² Kibeho was one of the major camps among two dozen that in early 1995 were to be closed under Operation Retour. Devised by the IOC in December 1994, Operation Retour used a combination of "push" and "pull" methods to encourage IDPs to return home. The operation's implementation arrangements used UNAMIR and the RPA for security purposes, the transport capacities of UNHCR and IOM, and the humanitarian resources of the major UN agencies and

(end of page 79)

NGOs for the packages that returnees would take to their home communes. Over half of the camp populations had made the transition by mid-February 1995.

Despite that promising beginning, increased insecurity in the communes led to a reverse movement back to the camps. Operation Retour ground to a halt as those who returned, though modest in number, fueled the fears of the remaining camp occupants. Disregarding publicity campaigns and other inducements, approximately 190,000 people refused to move. A "new" Operation Retour was needed. From the end of February until the first week in April, the IOC worked on devising new plans and approaches.

Finally, on April 5, 1995, the IOC submitted a plan to the government for closing the camps. Elements included timetables for ending food distribution in the camps and beginning them in the communes, enhanced security measures in the communes, and cordoned-off camp sites for IDPs who insisted on remaining. Though harsher than the original Operation Retour, the plan still adequately protected the human rights of the occupants. Through the UN humanitarian coordinator, UN agencies insisted that this be the case or they would withdraw their support. The coordinator emphasized the point to the SRSG as the latter prepared to introduce the plan to the prime minister.

It was agreed that the plan would be launched by the minister of rehabilitation and social integration on April 18, with implementation to begin during the following two weeks. These plans were overtaken by events, however. By April 18, the RPA had already surrounded the main remaining IDP camps as a prelude to their imminent forced closure. Thus the operational plans made by the IOC and presented by the SRSG to the government could not be implemented.

The resulting violence at the Kibeho camp demonstrated that the gulf between governmental and international perceptions and responses was too wide to bridge. Rwanda's resentment over the ineffectiveness of the international response paralleled the outside world's frustrations with the government's seeming lack of flexibility and humanity. Where the former did not understand why resources for UNAMIR would not better be used to assist the government, the latter wondered how a supposedly broad-based regime offered only prison as an alternative to justice.³³

The authorities increasingly felt that the international community offered only promises without results. Pledges to provide massive assistance to post-genocide Rwanda, including almost US$600 million

(end of page 80)

pledged at the January 1995 Rwanda Round Table, had not fully materialized. International efforts to deal with the IDP problem seemed halting and ineffectual. Conversely, the international community saw itself, too, receiving only promises with no results. The government's frequent assurances that the essential "criteria of guilt" for the genocide would be elaborated remained unspecified after two years; trials based on those benchmarks had yet to begin. Its lackluster participation in the IOC planning process for IDPs, as promised, was viewed as another indicator of lack of serious intent.

Recommendations for the Future

The experience of the international community in providing assistance and protection to IDPs in Rwanda from 1994 to 1996 raises a number of key policy issues. The issues constitute something of a microcosm of the IDP challenge in other complex emergencies as well. We present these recommendations in the hope of improving humanitarian responses in general and in situations involving internal displacement. Few are new, but the recent experiences underscore the need once and for all to come to terms with the challenges identified.³⁴

A Commitment to Address the Fundamental Issues

It is often observed that humanitarian action has become an alternative to political will rather than an embodiment of it. Indeed that appears to have been the case in Rwanda. Although large numbers of vulnerable groups required emergency relief, the central issue–and the central failing–of the international community's involvement in Rwanda was its unwillingness to confront the political issues that were the fundamental cause of the crisis.

The reluctance of the international community to become politically engaged was reflected in its unwillingness to support efforts to separate out those Rwandans in Zaire who should have been excluded from refugee status from those who were genuine refugees.³⁵ Its political timidity was also apparent in its willingness to tolerate, despite an arms embargo, the buildup of weapons along the Zairian-Rwandan border.³⁶ The failure of the major bilateral actors to agree on objectives that would have lent direction and support to UN efforts also demonstrated a lack of political will.

(end of page 81)

Although humanitarian resources are largely contingent on the good will of a small group of bilateral donors, humanitarian organizations need to be far more vocal than they were in this instance in demanding political action to resolve inherently political crises. Aid organizations need to be more willing to engage and challenge the governments on which they depend for resources. Although the humanitarian community cannot abandon those in need and still remain faithful to its mission and mandates, it must resist being used as a Potemkin village to protect the sensitivities of those who hold the key to real solutions.

In an effort to generate more effective advocacy among donor governments, the Inter-Agency Standing Committee (IASC) could play a far more potent role. It represents a forum for agreeing on common strategies and for conveying these to the political and peacekeeping elements of the UN system. NGOs, in addition to playing a more assertive role on the IASC, should use their sometimes greater access to the corridors of power to press for political solutions.

From a review of the recent past, it is apparent that in Rwanda and other major crises too much attention has been paid to "early warning systems" and too little to what might be called "early implementation systems." Before the cataclysmic events in Rwanda and before most such catastrophes, warnings were and are available about societies and social groups in distress. Even with such information, however, the international community has been slow to respond. Linked closely with the capacity to read the early warning signs, therefore, must be a capacity to mobilize advocacy efforts and to act with dispatch.

The Rwanda crisis, both the genocide and its consequences, such as the flood of the uprooted, also points to another indicator of international commitment: an in-depth understanding of the issues in a given emergency. The resources and responses first masked and later highlighted the profound lack of understanding of the nature of the problem. The fundamental issues must be better understood before they can be effectively addressed. The willingness to intervene is no substitute for a working understanding of the problems to be addressed.

Clear-Cut Policy and Institutional Responsibility for IDPs

The absence of an effective UN systemwide approach to IDPs was demonstrated in the response to the Rwanda crisis. In 1994 the responsibilities of the UN Rwanda Emergency Office (UNREO), the in-country

(end of page 82)

coordination vehicle for humanitarian activities, were limited to Rwanda proper, while UNHCR had the lead responsibility for Rwandan refugees in neighboring countries. UNREO's responsibilities inside Rwanda stopped at the Rwandan border; refugees were a concern only once they had returned. Most of UNHCR's responsibilities also stopped at the border; IDPs within Rwanda were not its primary responsibility. Missing were clear-cut decisions of policy and institutional responsibility regarding IDPs that embodied integrated and multisectoral approaches and were supported by political authority.

Each of the major policy and institutional alternatives has advantages and disadvantages. An institution specifically designated to deal with IDPs would make unnecessary the tortuous process of identifying an entity to take responsibility with each new crisis. There would also be benefits from having a center of demonstrated experience and expertise at the ready. Some have suggested that the mandate of an existing institution such as UNHCR should be expanded to encompass responsibility for IDPs. There are strong arguments in support of this suggestion, including the similar operational needs of refugees and IDPs, UNHCR's established expertise in both assistance and protection, and the likely economies of scale to be realized by having a single institution serve both categories of uprooted populations.

On the other hand, the experience in Rwanda suggests the alternative of a "controlled ad-hocracy" such as is reflected in an Integrated Operations Centre. Such a structure could facilitate the development of common strategies, including the sharing of both responsibility and resources. It could also be adjusted to the particulars of a country-specific situation and lodge political, human rights, humanitarian, and peacekeeping components under a single operational umbrella. An existing agency deputized to deal with IDPs or an agency newly created solely for that purpose might simply add more bureaucracy to an already overly bureaucratic system. A single agency responsible for IDPs or an agency whose mandate is extended to include IDPs might also duplicate efforts, heighten competition for donor resources, and provoke more debates over definitions relating to the displaced. Any single humanitarian entity would probably also lack the authority and capacity to integrate essential political, human rights, humanitarian, and peacekeeping elements.

Whatever the preferred option, four points are essential to bear in mind. First, support from headquarters by UN political and humanitarian officials, including the Inter-Agency Standing Committee, is a prereq-

(end of page 83)

uisite for effective policies and operations. Second, there must be incountry political support of humanitarian operations to assist and protect IDPs through a coherent UN field structure. Third, the institutional configuration chosen must be aufficiently flexible to meet field challenges affecting each IDP situation. Finally, the structure put in place must incorporate the range of expertise needed to assist IDPs effectively, from peacekeeping to human rights, from humanitarian concerns to those of politics. One lesson emerges clearly: UN operations in the field require greater coherence and attention to structure and more effective command-and-control arrangements than existed in Rwanda. In such a complex emergency, the SRSG should have overall responsibility for all aspects of the operation, with two deputies. One would be responsible for all aspects of humanitarian, recovery, and development assistance, the other for peacekeeping activities. All UN humanitarian and development organizations would report to the former, who would also be in charge of humanitarian and development issues in which the SRSG or the military were involved.³⁷

The utility of this approach depends in part on the effectiveness of future SRSGs, which accordingly requires attention to the procedures by which they are selected. Success also depends upon the capacity of the SRSG office and its personnel. Agencies in the field generally do not resist overall coordination authority as long as priorities are agreed upon, headquarters concur, mandates are clear, and there is evident value to the arrangements. The value of a more structured UN system in the field is that the full weight of the system can be brought to bear on an issue such as IDPs without overtaxing any one component or making any single organization bear the political opprobrium of the authorities.

One factor that has inhibited development of a more coherent field structure has been the desire of individual agencies to preserve their independence. This is one reason for their reluctance to work across borders, as in the Great Lakes region. Even within agencies, "turf" concerns can result in unwillingness of representatives from the same agency in two adjoining countries to share information, let alone to cooperate in other operational ways. In the name of sovereignty, states have reinforced such rigidity by restricting attempts to work together across borders.

Yet if long- as well as short-term solutions to the IDP and refugee problems in the Great Lakes are to be found, greater attention must be paid to regional structures. The view of a geographical border as an

(end of page 84)

operational dividing line proved to be as artificial and dysfunctional in Rwanda as the distinction attempted between IDPs and refugees. "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 of policies inside Rwanda with those relating to refugees in neighbouring countries."³⁸ Regional information-sharing also requires a mechanism with the authority and capacity to monitor, analyze, and implement strategies and programs that will address refugee, IDP, peacekeeping, and even developmental issues from a regional perspective.

More Balanced Use of Resources

A pattern of resource allocation that favored refugees at the expense of IDPs was established in 1994. The halting and uneven international response to the genocide within the country from April 1994 onward contrasted sharply with the outpouring of assistance to those who fled the country. Contrasting international hyperactivity surrounding the cholera epidemic in Goma with earlier inactivity at the time of the genocide, MSF commented derisively, "The people who had sat stony-faced while innocents were massacred were suddenly deeply moved by the damage wrought by a bacterium."³⁹

The subsequent pattern remained largely unchanged, despite considerable comment on the imbalance. "We have received more than $20 million in private donations and government funds to support relief work in the [refugee] camps," observed the president of CARE in February 1996, "and only $3 million to assist in Rwanda." The division of resources, he observed, was "inhumane, wrong-headed, and doomed to fail.⁴⁰

In addition, the allocation of expenditures on assistance has worked to the detriment of protection. A consensus now exists that both the UN and the governments largely ignored the warning signs of the impending genocide that were evident at least as early as January 1994. The failure to take preventive action early in the year, followed by the failure to meet protection obligations once the genocide had begun, allowed the situation to become unmanageable. The world was then faced with unfulfillable protection needs and exorbitant assistance ones. "When ethnic massacres began in Rwanda in April 1994, the U.S. share of the cost of augmenting the U.N. force there would have been about $10 million," observes one

(end of page 85)

analyst. "By July, with the U.N. force drawn down rather than supplemented, and with half a million Rwandans dead and two million refugees in neighboring countries, the U.S. finally felt compelled to commit nearly $500 million to the emergency relief effort."⁴¹ What is true for U.S. investments was equally true for other governments.

Aid agencies such as UNICEF and CARE were often in the forefront of advocacy efforts on behalf of Rwanda's people. Yet they, no less than governments and the UN itself, have responsibilities to establish allocations that match the need. How humanitarian institutions frame the issues and their own roles has a direct bearing on the terms under which resources are provided and accepted. A more probing approach to the causes of human need by the media might also facilitate a more proportionate use of resources.⁴²

Stabilizing Conflict-Affected Societies and

Post-Conflict Recovery

The massive amount of resources expended for humanitarian operations in the Great Lakes region masks the narrow vision and inadequate mechanisms of international support for conflict-affected societies. We offer four specific recommendations:

It may seem curious to recommend placing the funding mechanisms for emergency and recovery activities within the panoply of peacekeeping efforts. Those efforts, however, would not conflict with activities normally

(end of page 86)

carried out by UN specialized agencies, which could themselves play prominent roles in supporting additional post-conflict activities under the overall aegis of the SRSG. The activities added to the UN peacekeeping mandates would be those that could promote post-conflict recovery and use the assets of the peacekeepers themselves in collaboration with UN agencies. To continue to make important aid activities depend on "voluntary" contributions while governments are "assessed" to raise funds for peacekeeping work seems contradictory and counterproductive.

Whether the Rwandan government would have taken a different attitude if innovative UN structures and procedures had met its needs quickly and efficiently will remain a matter of speculation. But it is possible that the government's attitude toward the peacekeepers themselves might also have been more positive had they been an integral part of a comprehensive effort to restore the nation's ability to function.

(end of page 87)

Lessons from Rwanda

In the final analysis, most complex emergencies are about conundrums. In the case of the internally displaced, the reason is all too apparent. The plight of the displaced person–the more so of displaced populations as groups–poses multifaceted and multidimensional challenges. Simply gaining access to provide assistance is one example; the impediments confronted in offering protection are another. Both are frequently complicated by the absence of a perceived vested interest in the fate of those who have abandoned, for whatever reasons, the conventional relationship between themselves and the state.

Herein lies the source of the first conundrum: that the very issue of displacement rarely garners much priority from the authorities whose populations are displaced within their own borders. Where ethnic tensions are involved, IDPs are often a bigger source of political embarrassment than of humanitarian concern to the resident authorities. In Rwanda the state saw no abiding humanitarian interest in the situation, perceiving the IDPs, like the refugees who had fled the country, as linked inextricably to the genocide. Thus the connection between the conundrum of displacement and the overarching conundrum of genocide.

Issues related to displacement examined in this chapter–morality, policy, and resource allocations on the one hand and operational and practical matters on the other–were part and parcel of the reality and the residue of genocide. At issue was not the attempt of one party to best another in a political or military confrontation of wills, but rather of one ethnic group to eliminate another altogether. The fact that both parties were implicated in exclusive policies and politics over time and throughout the Great Lakes region complicated the challenge further still.

Both conundrums posed challenges to international political will; and the international community did not respond effectively to either one, in part because it failed to approach the former conundrum as firmly nested in the latter. To be sure, the world expressed horror at the events of April 1994, but ineffectually. Effective action would have required measures to separate those guilty of crimes against humanity from genuine IDPs and refugees, to respond more comprehensively to issues of regional security and stability, and to deal with the legitimate assistance and infrastructural needs of a society traumatized by conflict.

There is much to be distilled from the experience in Rwanda. Its lessons on displacement are many. Displacement promises to be a recur-

(end of page 88)

ring problem as societies splinter and the state loses its claim on the loyalties of populations and its centrality in international relations. Although it is to be hoped that genocide will not recur in Rwanda or elsewhere, it, too, offers lessons too important to leave unexamined.

Notes

1. The unleashing of genocide within the context of the 1990-94 civil war led to confusion–some of it genuine, some of it intentionally obfuscatory–about what was actually taking place. The UN Human Rights Commission's special rapporteur on summary executions sounded an alarm by pointing out that massacres being perpetrated in Rwanda seemed tantamount to genocide: "The cases of intercommunal violence brought to the Special Rapporteur's attention indicate very clearly that the victims of the attacks, Tutsis in the overwhelming majority of cases, have been targeted solely because of their membership of a certain ethnic group, and for no other objective reason. Article II [of the Convention on the Prevention and Punishment of the Crime of Genocide], paragraphs (a) and (b), might therefore be considered to apply to these cases." See Report of the Special Rapporteur on Summary, Arbitrary and Extrajudicial Executions, E/ CN.4/1994171Add. 1 (United Nations, Commission on Human Rights, August 11, 1993), pare. 79. Human rights groups themselves early on identified the problem as "genocide" and urged the signatories to the Genocide Convention to intervene.

By contrast, the United Nations for a number of weeks studiously avoided using such terminology, although the UN secretary-general did inform the international community during the last week of April 1994 that a "genocide" had taken place. "Even in the face of convincing proof of the true nature of the massacres," reported Human Rights Watch, "a few Security Council members refused to acknowledge that they constituted genocide" and effectively prevented use of the term in Security Council resolutions. U.S. officials were specifically instructed to avoid the term. Describing what was happening as a civil war between traditional rivals offered a convenient rationale for avoiding obligations under the Genocide Convention.

2. Discussions of the Rwandan crisis and the international response have the benefit of a growing number of studies. The most comprehensive is The International Response to Conflict and Genocide: Lessons from the Rwanda Experience, 5 vols. Joint Evaluation of Emergency Assistance to Rwanda (Copenhagen, March 1996). Released after more than a year of work by some fifty-two consultants and researchers, the evaluation contains four separate studies and a synthesis report. Initially proposed by DANIDA (the Danish government's international aid agency), the evaluation was managed by a Steering Committee consisting of thirty-seven members drawn from governments, UN organizations, and nongovernmental organizations (NGOs) who guided the work and provided the $1.7 million necessary to carry it out. The finished product, which includes a useful review of bibliographical resources, represents the most exhaustive review of any major humanitarian operation to date.

A number of other studies provide useful perspectives on the issues. One of the first to appear, featuring first-person accounts of Rwandans involved in the crisis, was African Rights, Rwanda: Death, Despair and Defiance (London: 1994; rev. 1995). A second study is a collection of essays by senior officials involved in Rwanda and other complex emergencies: Jim Whitman and David Pocock, eds., After Rwanda: The Coordination of United Nations Humanitarian Assistance (London: Macmillan, 1996). Of particular relevance are chapters by Major General R. A. Dallaire, "The Changing Role of UN Peacekeeping Forces: The Relationship between UN Peacekeepers and NGOs in Rwanda"; and Randolph Kent, "The Integrated Operations Centre in Rwanda: Coping with Complexity" A third volume, by Larry Minear and Philippe Guillot, Soldiers to the Rescue: Humanitarian Lessons from Rwanda (Paris: OECD, 1996), examines the role of international military forces in responding to the crisis. Various UN organizations and private aid agencies have also produced reviews of their own experiences. For example, see Department of Humanitarian Affairs, "Report on the Coordination of Humanitarian Activities in Rwanda," United Nations, 1995; and International Federation of Red Cross and Red Crescent Societies, "Under the Volcanoes: Special Focus on the Rwandan Refugee Crisis," in World Disasters Report (Geneva, 1994).

3. The moral complexities surrounding the genocide stem in part from the government's use of its official radio not only to urge the elimination of the Tutsi but also to threaten the lives of those Hutu who refused to take part in the slaughter. These threats were reinforced by interahamwe, who were not reluctant to act on the pronouncements of Radio Milles Collines, a privately owned and operated extremist broadcasting system established by members of the government and the wife of President Habyarimana. For examples of such broadcasts, see J. A. and C. Pott-Berry, eds., Genocide in Rwanda: A Collective Memory (London: CARE-UK, 1996). Such radio pronouncements should be put in the context of a Hutu majority, the vast majority of whom were minimally educated peasants, traditionally dominated by the Tutsi minority until the early 1960s. Also a reality was the perceived threat that a Tutsi army might take over the country.

4. Historical Perspective: Some Explanatory Factors, vol. 1 of The International Response to Conflict and Genocide: Lessons from the Rwanda Experience, Joint Evaluation of Emergency Assistance to Rwanda (Copenhagen, March 1996), p. 12.

5. The number of "old caseload" refugees returning to Rwanda generally used for planning purposes in 1994-95 was 600,000. The majority came from Uganda, where almost two generations of Rwandans lived both as official refugees and in some instances as "Ugandans." A much smaller percentage came from Burundi and an even smaller numbers from Tanzania, Kenya, and Zaire. The relationship between this planning figure and the numbers that fled Rwanda between 1959 and the early 1990s is very imprecise. The figure of 600,000 included those born in exile. The multidonor evaluation concludes that "this figure is contested by many [though is probably] the very best estimate available" (ibid., p. 30).

6. When the then-president Juvenal Habyarimana spoke about the possibility of refugees returning to Rwanda, he used the analogy of a "full glass" to suggest the negative consequences of any significant return. Similarly, there are many prominent persons within the Rwandan leadership today who question the need for a mass return of the Hutu who fled the country in 1994. There are also demographic and economic questions that the new government must face. The lack of adequate farmland for an economy based primarily on agriculture inevitably poses a disincentive to return. The ability of the country to support an additional 2 million people not only represents a serious economic issue but also raises questions of "ethnic balance."

7. Historical Perspective: Some Explanatory Factors, p. 11.

8. For discussion of such relationships, see Peter Uvin, Development, Aid, and Conflict: Reflections from the Case of Rwanda (Helsinki: UN University/ WIDER, 1996).

9. Reprinted from Historical Perspective: Some Explanatory Factors, p. 55.

10. During April and May, 580,000 fled from Rwanda to Tanzania; in July, 1.2 million fled to Zaire. An additional 243,000 Rwandans who went to Burundi during April constitute the remainder of the 2 million persons referred to in the text.

11. Reliable sources state that among the refugee population were probably more than 40,000 persons belonging to the former military and political apparatus who had played an instrumental role in the planning or execution of the genocide and who, under the terms of UNHCR's mandate, would be undeserving of international protection. The government of Rwanda, however, estimated that the total number of guilty persons was around 170,000, or about 10 percent of the refugee population. This figure most likely included those in the first two groups referred to in the text.

12. Initially, many among the refugees were subjected to the authoritarian rule of what became known as the "intimidators," who either as former military or interahamwe officials ran some of the camps as virtual states within a state. The intimidators tried to control the exchange of information in the camps perpetuated stories of terror and mass killings in Rwanda, and through their own brutality ensured camp compliance. This military influence was later reported to have been replaced by "more conventional civilian control." One analyst has asserted that "the camp populations have now become so indoctrinated that force is no longer needed to convince them of the terror that awaits across the borders. The 'bush-telegraph' tells them about the killings in the communes, about the prisons, about Kibeho. And this is what keeps them united in the camps and unwilling to return home."

13. The insensitivity of many well-intentioned humanitarian organizations was remarkable. At a mid-1995 reception for a visiting delegation to which senior Rwandan government officials were invited, one major multilateral organization displayed, amidst tables bulging with elegant canapes and drinks, copies of magazines featuring its work with Rwandan refugees in Zaire. Rwandans were also puzzled by the promotion by aid groups of conferences, workshops, and seminars on reconciliation only months after the genocide had ended. While to aid agencies it seemed essential to move quickly to reconstruction, Rwandans believed that generous assistance to refugees and dispassionate discussions of reconciliation begged profound questions of justice.

14. Convention Relating to the Status of Refugees, Article 1F (1951).

15. The cattle brought through the northeastern parklands, or Akagera forest, had important symbolic as well as economic value. For many Rwandans–the very wealthy as well as the very poor–cattle represented bank accounts on the hoof. Cattle wealth is assumed to have provided a considerable portion of the resources used to build up the RPA, although the herds destroyed much of the parklands after the refugees returned. The situation became so politically sensitive that for almost two months the minister of rehabilitation and social integration operated out of the northwest in order to give higher priority to the problem. Human and livestock displacement and resettlement were thus intimately related.

16. In a January 1995 meeting between the UNHCR special representative for the Great Lakes and the UN humanitarian coordinator for Rwanda, the former noted the importance of successfully returning the IDPs to their home communes: "If the IDPs don't make it back safely, we'll find ourselves with at least an additional 150,000 in Zaire that we won't be able to handle."

17. The UN secretary-general also appointed separate representatives to deal with the region, including Ambassadors Aldo Romano Ajello and Jose Luis Jesus. These individuals, however, did not provide full-time presence in the region or have the mandate or capacity to deal with the regional nature of the problem.

18. For a more detailed account, see Minear-Guillot, Soldiers to the Rescue, chap. 4.

19. For an elaboration of these points, see Minear-Guillot, Soldiers to the Rescue.

20. The UNAMIR mandate, revised on May 17, 1994, provided for 6,000 uniformed troops, military observers, and civilian police. UNAMIR was charged with supporting and providing safe conditions for IDPs and other groups in Rwanda (both throughout the country and on the border) and with helping humanitarian organizations to provide aid. UNAMIR was to monitor border crossing points and could take action in self-defense against those who threatened protected sites and populations and the delivery and distribution of relief aid. Despite these wide-ranging duties, UNAMIR had Chapter VI rather than Chapter VII authority. A Chapter VII mandate would have permitted the use of deadly force to impose peace, empowering UN troops to tackle the broader challenge of restoring and maintaining law and order at the commune level. A Chapter VI mandate allowed deadly force to be used only in self-defense and in the protection of UN installations.

21. During the month of December, UNAMIR joined with the RPA in an ostensibly secret operation to root out the intimidators and perpetrators of genocide from the Kibeho camp. This joint operation was planned without the knowledge of the UN humanitarian coordinator, though UN aid agencies as well as NGOs were working in Kibeho. No incidents occurred, largely because word had already spread throughout the camp about the upcoming operation and the apparently guilty had fled. Nevertheless, the fact that no provision was made for the safety of relief workers (let alone for innocent civilians) reflected poorly on UNAMIR's planning and did little to enhance its cooperation with civilian agencies. It was in part to prevent UNAMIR from taking such unilateral initiatives that the Integrated Operations Centre (described later) was created.

22. Francois Jean, ea., Populations in Danger 1995 (London: Médecins sans Frontières, 1995), p. 45.

23. In November 1994, NGOs working in the camps in Zaire issued a joint statement condemning the continued intimidation by ex-FAR and interahamwe of the refugees and the international community's reluctance to address this challenge to recognized principles of refugee law and treatment. On the basis of this declaration, a number of NGOs withdrew their services.

24. The French government was identified with ongoing support for the ancient Hutu regime, which was francophone. Even after the transition from the Hutu to the Tutsi regime, the French were alleged to have continued supplying arms and military training to the discredited army.

25. On occasion, U. S. Assistant Secretary of State for African Affairs George Moose chaired meetings.

26. Given the constraints described in the text, it is nevertheless noteworthy that UNHCR provided a staff member for the IOC secretariat and, during the Kibeho crisis, major transport assistance. WFP, too, supplied a junior member of its staff to the IOC operation. UNICEF was highly supportive and provided a representative but was too overstretched to focus on the IDP issue. IOM and NGO involvement was considerable. UNAMIR, which detailed three people to the IOC, also provided telecommunications and computer equipment. Its threatened withdrawal from the IOC in the midst of the Kibeho crisis reflected the force commander's distress at not having received essential information from the IOC during the operation. As later became apparent, the information had been received but not transmitted within UNAMIR's own chain of command.

27. Conversations between the authors and UNAMIR Force Commander Major-General Guy Tousignant, most recently on October 18, 1996.

28. As early as September 1994 the SRSG proposed a Rwanda Emergency Normalisation Plan designed to address key infrastructural requirements. Such a plan would have gone a long way toward proving the international community's commitment to stabilizing the traumatized society of Rwanda. This type of plan would have been well suited for a quick-action trust fund to be established under the SRSG.

29. It is worth noting that one of IRIN's functions was to counteract the agencies' habit of failing to transmit information across borders on a regular basis to their colleagues. The conventional pattern of communications is from the field directly to headquarters; rarely does the field systematically transmit information to colleagues in neighboring countries. IRIN encouraged the agencies in the field to transmit information for circulation throughout all countries in the region as well as to the international community at large.

30. See Randolph C. Kent, "Rwanda: The Aid Impasse," DHA News, May-June 1995, pp. 27-29.

31. An example of the nature of the needs and the reluctance of the established aid agencies to take them on was provided by the challenge of improving the conditions in the country's prisons. In the absence of a UN organization with competence in that area, the humanitarian coordinator, in his personal capacity, was asked to assume the responsibility of special adviser to the SRSG on prisons.

32. The government's original figure of 300 and other figures ranging from 800 to 8,000 reflected the politics and the confusion surrounding the Kibeho incident. The most accurate figure probably came from an Australian military medical team that counted almost 4,000 corpses in areas into which it was allowed. (The team had covered only about half the area before being prevented from proceeding by the RPA.) Even after the worst two days of incidents, there were periodic shootings as the RPA reportedly continued its own "screening process." The final figures are not known, but somewhere between 4,000 and 8,000 deaths seems likely.

33. Originally it was thought that the government and donors had agreed that the shortage of Rwandan magistrates would be temporarily offset by magistrates from other francophone countries. It was assumed that a functioning justice system would relieve donors of the burdens of assisting the government to build prisons and, more important, would be a major step toward some form of accommodation, if not reconciliation, with Rwandans outside the country After months of apparently successful discussions, the government rejected the idea of using foreign magistrates.

34. Ironically, the dysfunctional division of labor among the agencies regarding IDPs had been a recurring problem even before the crisis in Rwanda. In a 1994 report on displacement issues from a global perspective, the representative of the secretary-general on IDPs noted that "despite greater willingness on the part of UN agencies to develop more coherent collaborative arrangements, a vacuum of responsibility often exists in cases of internal displacement." Francis M. Deng, "Internally Displaced Persons: An Interim Report to the United Nations Secretary-General on Protection and Assistance," Refugee Policy Group and United Nations Department of Humanitarian Affairs, December 1994, p. 40. See also Roberta Cohen and Jacques Cuenod, Improving Institutional Arrangements for the Internally Displaced (Washington, D.C.: Brookings Institution-Refugee Policy Group Project on Internal Displacement, October 1995). See also Lance Clark, "Internal Refugees: The Hidden Half," in U.S. Committee for Refugees, World Refugee Survey: 1988 in Review (Washington, D.C., 1988), pp. 18-24.

35. Clauses in both the Geneva and OAU conventions on refugees allow individuals known to have carried arms or to have committed crimes against humanity to be denied refugee status.

36. The UN at the highest levels was officially informed about the introduction of arms into Zaire and was provided with documentary evidence by April 1995 that included pictures of arms caches. The evidence sparkeu,.... senior UN officials.

37. Such an arrangement was tried for the first time in 1995 in Haiti, where the deputy SRSG served also as the UNDP resident representative and the DHA humanitarian coordinator. See Edwige Balutansky and others, Haiti Held Hostage: International Responses to the Quest for Nationhood 1986-1996 (Providence, R.I.: Watson Institute, 1996), pp. 80-81,107-8.

38. Humanitarian Aid and Effects, vol. 3 of The International Response to Conflict and Genocide, Joint Evaluation of Emergency Assistance to Rwanda (Copenhagen, March 1996), p. 159.

39. Rony Brauman, "Genocide in Rwanda: We Can't Say We Didn't Know," in Jean, Populations in Danger, p. 89.

40. Peter D. Bell, "Rwanda Aid Doesn't Add Up," Christian Science Monitor, February 2, 1996.

41. Kathleen Newland, "U.S. Refugee Policy: Dilemmas and Directions," Carnegie Endowment for International Peace, Washington, D.C., 1996, p. 40.

42. For a review of the role of the international media in influencing humanitarian activities and international policy in Rwanda and other major recent crises, see Larry Minear, Colin Scott, and Thomas G. Weiss, The News Media, Civil War, and Humanitarian Action (Boulder, Colo.: Lynne Rienner, 1996).

43. The "framework" arrangements established in 1994 among DPKO, DPA, and DHA principally concern sharing of information. However, by 1996 promising moves were under way to engage the capacities of each to assist in the preparation of peacekeeping proposals.

44. Despite two years of the donor-driven ROSG, few practical ideas emerged from the donors about how to support post-conflict recovery or societal stabilization.