Journal of Humanitarian Assistance
7.1. Early warning
The Puzzle
Did those charged with the responsibility for making decisions in the various agencies and states that comprise the international community know that the assassination of Hutu political opponents and genocide of the Tutsi in Rwanda would take place? No. Virtually no-one anticipated the swiftness, scale, thoroughness and unique character of the genocide as it unfolded. In its horrific enormity, it took almost all international observers by surprise.
If they did not draw such an extreme conclusion, did those with the capacity to prevent and mitigate the genocide have the information upon which such a conclusion could be drawn? Yes. Many knew that organized extremist forces existed; increasingly, they even gave public proof of their existence by words and deeds. A pattern of violence was discernable, and the state apparatus itself was clearly implicated in arms distributions to para-military groups and extremist propaganda advocating the need to rid Rwanda of all Tutsi and their supporters. By early 1994, specific information about plans and conspiracies towards this end was picked up by the UN system, most significantly in the notorious Black File of January 1994.
Close observers commonly interpreted this information in the Rwandese context to mean that large-scale ethnic violence against civilians was likely if the civil war was renewed, as seemed likely in early 1994. However, in the higher echelons of state and international agencies, the early warning signals were hardly heard at all, and news of the genocide was an even greater surprise. But pieces of information were available to permit policy-makers on any level to draw the more radical conclusion that both politicide and genocide might occur on a scale quite different from past patterns of "just" hundreds or thousands of victims.
Should the alarming indications have resulted in more thorough assessments and preparations? Absolutely, given the political commitment and actual involvement of the UN in Rwanda, the large-scale violence expected by close observers, and the possibility of genocide. The legal right and the normative obligation to act to prevent genocides according to the Geneva Convention, and the enormous cost of a miscalculation, made contingency planning clearly imperative.
Then why did states, international organizations and other parties who had assumed some responsibility for regulating the Rwandese conflict and had the capacity to act not draw the appropriate conclusions? Since close observers did anticipate mass violence and advocated swift interventionist action, analyzing the problems that afflicted the central decision-makers is critical to understanding the failure to act.
The problems of receiving even clear and unequivocal signals are found in four areas: contradictions in the international system; the UN structure; attitudes of senior officials towards messengers and inadequacies in the messages sent; and interference.
7.1.1.Contradictions in the international system
Neutrality and Intelligence Gathering
Unlike nation-states, the United Nations does not collect and analyze information to protect itself from ostensible enemies. The world organization has no clear-cut security agenda and is supposed to be a neutral body. Yet in matters of threats to international peace and security, the neutrality principle does not necessarily apply, as in Chapter VII enforcement actions. By the same logic, the UN should not be neutral towards genocide, or towards parties threatening civilians whom the UN has placed under its protection. Moreover, once the UN assumes responsibility for conflict management, it needs a capacity for information collection and analysis dealing with military and political issues of member states. Nevertheless, member states are reluctant for reasons of national security to let the UN develop such an intelligence function.
Financial, Remote and Regional Interests
In the area of conflict management, the UN is particularly beholden to the United States, which pays almost a third of the budget for peacekeeping. When regional states and organizations, which are closer to the conflict and its consequences, and in this case were willing, indeed eager, to involve themselves, they were not given the wherewithal to do so, including a structure to gather and analyze information. The OAU, which was intimately involved in the diplomatic process to settle the conflict, only had a skeletal staff for these purposes. The lack of resources also characterized Tanzania, in this case the most non-partisan regional state. The major motive of Tanzania was to regulate the conflict, rather than further one party's interests.
Zaire had access to the Rwandese regime but no interest in collecting, let alone sharing, any information that could be used against the Rwandese government. Uganda had such an interest, and through the RPF also had access to information, but its early warning capacity supported the RPF. RPF early warning could be dismissed as propaganda. France had both interest in and a significant capacity for intelligence collection on Rwanda, but its interests were clearly partisan and led to defining the role of the RPF in adversarial terms. (This included even the French Ministry of Foreign Affairs, which was committed to the Arusha process.)
The United States also had the capacity, but it was initially truly disinterested, in the sense of being both objective and remote, and utilized its capacity only sporadically. The CIA undertook a January 1994 desk-level analysis of the Rwanda situation as a worst-case projection of the course of current events, which included scenarios of deaths in the order of half a million casualties. This indicated that specialist analysts who focused on a problem area could use current information to develop reasonably accurate scenarios. Covering a country of marginal concern to the US, the report was not distributed widely, nor did it reach the higher decision-making echelons. As there was no "smoking gun" or verification from personnel on the ground, it was not taken seriously. Moreover, since Rwanda posed only a hypothetical problem rather than being an actual crisis, there was little inclination by higher officials to pay attention.
For the UN, both the interest and capacity to cover a crisis in a peripheral state are critical. Caught between the absolute disinterest of the major powers and the need to cater to dominant financial interests, the UN has been unable to establish the appropriate balance between disinterestedness and utilizing regional interests to advance the peace process.
7.1.2. Structure and culture
Though the UN was inhibited from systematically collecting and analyzing critical information and communicating the analyses to those with the power to take action against the genocide, the Secretary-General has had a mandate to engage in preventive diplomacy and peacekeeping even in intra-state conflicts. He has also been given an explicit mandate to engage in early warning. That mandate can be easily extended to cover the analysis required for preventive diplomacy and effective peacekeeping. Why has this capacity not been appropriately developed?
The individual relevant units are thinly staffed, given the exponential increase in global responsibilities assumed by the UN. For instance, only one person in DPKO consistently monitored Rwanda, and that person carried a heavy weight of operational responsibilities. But there are deeper reasons for the failure.
With a changed mandate since 1990, the Secretary-General reorganized the Secretariat, but in so doing, sacrificed an independent information and strategic analytic arm separate from any operational functions by dissolving the Office for Research and Collection of Information (ORCI), because of its shortcomings. The Secretary-General distributed its responsibilities among the Political, humanitarian, and Peace-keeping Affairs Departments, while the Office for Human Rights monitors human rights abuses. There was no central unit in the Secretariat or elsewhere charged with collecting even "soft" intelligence, including the information available in the vast structure of UN agencies and organizations to translate assessments into policy options and strategic planning. As the DHA task force evolved into the focal point for early warning in the UN system for mitigating the consequences of conflicts, there was no parallel tool to generate early warning signals required for contingency planning in preventive diplomacy and peacekeeping operations. DPA, which was expected to undertake this task, was too understaffed and had no systematic link to strategic planning.
Further, the cultural milieu of the Secretariat had reverted to being politicized in two senses. In the Rwanda situation, the DPKO at a critical juncture seemed to be too subservient to the concerns of the major powers. In this case, the DPKO proposed only "what the traffic would bear" in the Security Council, as they put it, not what the situation on the ground demanded. Further, some personnel suggested that others were carriers for dominant state players. Politicization versus professionalism in a context of relative great power disinterest by one major power and a partisan interest by another proved to be a devastating combination.
7.1.3. Messengers
In spite of these deficiencies resulting from the current mandate of the UN and its structural problems, UN headquarters did receive crucial information that should have led it at least to undertake some contingency planning. However, the UN Secretariat had a propensity to discount the information and warnings received from within its own system while paying inordinate attention to media analyses, if the Secretary-General's 20 April report, which prepared the ground for UNAMIR's withdrawal, is any indication.
Using information the various humanitarian organizations obtained from their own field operatives, an inter-agency task force in Geneva was assigned the task of sharing information about impending human disasters. It had many problems. It was ad hoc. Each of the agencies showed an understandable reluctance to collect and divulge sensitive political information that might jeopardize its operations in a country. More importantly, the inter-agency task force was oriented towards anticipating the humanitarian consequences of a crisis (e.g. population movements, the requirements of relief operations, etc.) rather than development of the crisis itself. The task force had no access to the sort of political-military information so critical in discerning escalating conflict. Nor did it have adequate support from qualified analysts. As a result, while the UN system has an early warning capacity, this capacity is inappropriate to questions of early warning for conflict management purposes.
Even greater suspicions were cast on the human rights organizations, which discerned patterns of violence they characterized as genocide. Part of the problem was the manner in which the warnings were conveyed. These organizations did not put human rights abuses in a political context necessary for understanding the nature and probable evolution of the conflict, including its likely points of culmination, or translate them into strategic options. Further, the human rights machinery of the UN itself was virtually irrelevant to the early warning process, not only lacking the capacity to relate human rights violations to a dynamic analysis of the social forces that produced them, but utilizing a monitoring process that was sporadic and provided no follow-through on reports.
Within the UN Secretariat, far too little attention was paid to early warning emanating from human rights organizations. The Secretary-General's report justifying withdrawal on 21 April reflected the misrepresentations of the media rather than the analysis of informed or more professional observers of Rwanda. When the Secretary-General by mid-May did perceive the situation correctly, he lacked the widespread and sustained support for policy engagement that adequate media coverage can generate. With a few notable exceptions, the media in general perpetuated the misrepresentation of the slaughter in Rwanda as one of anarchic ethnic violence. It took almost a month for most of the media to "get the story straight".
The import of the messages from the Rwandese media was downplayed. Hutu extremist calls in the hate media to rid the country of Tutsi were interpreted by some foreign observers as efforts to put pressure on the RPF to make concessions at Arusha, or as exaggerated rhetoric. Even members of the RPF politburo acknowledge in retrospect that they underestimated the significance of the extremist radio broadcasts.
7.1.4. Interference factors
Functional contradictions and structural deficiencies compounded by discounting some messengers and exaggerating the value of others are, however, only part of the explanation. The Rwandese Patriotic Front was closest to the scene, had networks in Rwanda to provide information, and was in close contact with human rights and refugee organizations as well as its own diaspora around the world. If any party had a stake in anticipating mass murder and its extent, and with developing contingency plans to prevent, or, at least, mitigate such a disaster, this Tutsi-dominated rebel organization surely did.
True, the leadership issued public warnings accusing the government of severe human rights violations and identifying those responsible for killings in 1992-93 as guilty of genocide. But leading Front members acknowledged in retrospect that they did not anticipate the magnitude of the genocide and the RPF did not develop contingency plans for such an eventuality. Why not?
Five factors are suggested that afflicted everyone involved in the Rwanda crisis to different degrees: incredulity, mind-blindness, shadows, noise, and desensitization.
Incredulity
Genocide is rare. Its path of development was unfamiliar and difficult to discern. Despite the precedent of Cambodia, many associated the concept of genocide with the highly mechanized Nazi holocaust. The idea that hundreds of thousands could be slated for execution in a poor, agricultural society, and that this could be carried out in a short time, seemed incredible.
Mind-blindness
RPF leaders admit that on the eve of their 1990 invasion they expected that thousands of Tutsi might be killed in retaliation, as had happened in the early 1960s. But to consider genocide as both a possibility and one they had very limited ability to mitigate, let alone stop, would have either frozen them into impotence or forced them to accept that they had a measure of historical co-responsibility in the slaughter. Hence, a sort of mind-blindness developed, i.e. an inability to use the information available to deduce the appropriate conclusions.
A similar blindness (but in the nature of "wishful thinking") may help to explain the failure of more remote actors to anticipate the genocide. For the French government to acknowledge that genocide was in prospect would be to accept a far greater degree of responsibility for its own involvement in backing the Habyarimana regime and its connected extremists. For the United States, it meant having to confront the fact that it had a legal right and a moral responsibility to intervene. For the UN Secretariat it would have meant preparing for a very different kind of mission in Rwanda at a time when its first principle of action was "not to cross the Mogadishu line". The UN mind-blindness carried a double level of guilt and denial by instilling a false sense of confidence among the Tutsi in Rwanda, who could not imagine that the world in general, and the UN in particular, would flee the scene when a UN peacekeeping force was already in place.
Shadows
The tardy and inadequate response of the UN and its leading members was also influenced by external blinkers, by previous events that haunt decision-makers, distort perceptions and constrain their willingness to act. These are in the nature of shadows.
One such blinker was the shadow of hope. Arusha became an inertial force. The peace accords so preoccupied almost all the players that they tended to shift into the background the growing organization of extremists intent on undermining that peace. The pursuit of the agreement, the hope invested in its effectiveness, and the eagerness to see it implemented dominated their attention right up until 6 April.
The other blinker was the shadow of despair and the propensity to read the present through traumas of the past. Somalia certainly directly colored the American propensity to view (and deform) the Rwandese operation through the Somalia lens. The issuance of PDD 25, a new directive in May of 1994 strictly limiting American involvement and funding of peacekeeping, was a culmination of the shadow of Somalia. But Rwanda was not a failed state; on the contrary, it was an example of a state that, having been taken over in an extremist coup d'état, was executing a massive genocide.
The UN was also haunted by Somalia. Fear of taking causalities in a UN peacekeeping mission made the Security Council and the Secretariat willing and ready to follow the Belgians out of Rwanda once the hostilities broke.
In the shadow of Somalia, the American and UN observers misinterpreted events on the ground and exaggerated difficulties into an impotence to respond effectively. Not only were the wrong lessons learned from Somalia, the right ones were ignored, In fact, the lesson of Somalia should have been that too few forces inserted tardily can undermine a whole operation.
Noise
This is a standard element often used to explain intelligence failures. At the same time as the situation was particularly tense in Kigali in March of 1994, the UN was preoccupied with a long list of crises or problem areas - in Russia, Angola, Burundi, El Salvador, Georgia, Haiti, Liberia, Nagorny-Karabakh, Somalia, South Africa, Tajikistan, Eritrea-Sudan, Mozambique, Hebron, North Korea, and especially the former Yugoslavia (including the security and safety of UNPROFOR). Because of Rwanda's peripheral relationship to the major issues and structures of international relations, the signals from Kigali were weakly heard.
In a context crowded with actual crises, it was particularly difficult to get busy decision-makers to shift their attention onto a potential crisis.
Confusion about the Message
Noise may overwhelm a message, but this propensity is exacerbated when the message itself is confusing. At various points, different actors monitoring the conflict - human rights organizations, the RPF, the Special Rapporteur of the Commission on Human Rights - used the term "genocide". But the term conveyed different meanings, which created confusion among the listeners.
The problem with the term "genocide" as a signal comes in the different implications and illustrations of the two uses of the term. The use of the legal definition of the term in an accusatory sense arguably diminished the impact of the term in its function as a warning signal. If the killing of 300 Tutsi constitutes genocide (in the legal sense), then warnings about potential genocide signal the potential death of a few hundred more. The linking of the deaths of 300-1000 people to the terms "apocalypse" and "genocide" diminished their impact as warnings. While significant in and of itself, early warning about genocide defined in legal terms leads to very different thinking about consequences and reaction than would a clear signal of an impending genocide as defined in the popular sense.
In the popular consciousness, the concept of genocide is linked to the massive slaughter of Armenians by Turkey at the turn of the century, the extermination of millions of Jews in Europe by the Nazi regime, and the wiping out of an entire generation and class of people by the Khmer Rouge in Cambodia. The numbers involved in each of these cases are measured in the hundreds of thousand or millions. Unfortunately, events propelled Rwanda into the same tragic class.
Desensitization
Only six months before the genocide occurred, massacres in Burundi claimed the lives of an estimated 50,000 to 100,000 persons. There was no worldwide outrage and little response. This failure of the international system to be startled or to respond reflected a pre-existing propensity to expect disasters out of Africa. The threshold for international response seemed to increase accordingly: if 100,000 persons could be killed with impunity, the possibility of another massive slaughter did not seem so extraordinary, nor did it require particular responses. Hence, there developed a mental indifference to the possible consequences of the rising tensions in Rwanda.
This was reinforced by the tendency to view the struggles within Rwanda as the recurrence of ancient conflicts between Hutu and Tutsi rather than as a product of a centrally led, political murder machine. As one Belgian official later put it: "We've learned from our experience that whenever there is trouble in Rwanda, the Hutu and the Tutsi kill each other."
7.2. Conflict management
The international community might have responded better had the early warning systems generated a clearer anticipation that genocide was on the horizon. Yet conflict management is a function of capacity, interest and commitment as well as information. In the Rwandese case, the relevant actors knew at a critical stage that the situation was unstable and dangerous. Unforeseen detrimental events did indeed occur to give the situation a turn for the worse - in particular, the October 1993 ethnic massacres in neighboring Burundi. But the major powers that controlled UN peacekeeping operations paid only cursory attention to Rwanda and there were no contingency plans except for withdrawal. The sustained and careful attention so necessary to successful conflict management was lacking.
There was considerable international investment in preventive diplomacy in Rwanda, but when this failed and genocide ensued, the international community effectively disengaged. Humanitarian relief within Rwanda and to refugees across the border soon resumed, but these efforts were designed to deal with the consequences of conflict rather than the violence itself. Re-interventions to save civilians had very limited impact, as indeed might be expected from retroactive conflict management.
The reasons for this fundamental failure to respond are found, ultimately, in a structural mismatch between institutions and interests in the contemporary state system. Revitalized by the end of the Cold War, the United Nations in the 1990s rapidly expanded its peacekeeping operations throughout the world. Rwanda was added to the list in October 1993. The framework for peacekeeping, however, was set by the distribution of power in the Security Council, which in form still reflected the world as it was half a century ago. Except for France, the major powers on the Council were basically uninterested in a small Central African country that was marginal to their economic or political concerns, and peripheral to international strategic rivalries. Through their power of veto and financial commitments, these states also controlled the peacekeeping or enforcement operations of the United Nations. Preoccupied with crises elsewhere, especially in Bosnia and Haiti, and haunted by the memories of Somalia, they decided not to engage in Rwanda until it was too late. Other potential actors were either unwilling or, at that late stage, unable to respond on their own. Hence, a principal lesson from the Rwandese conflict is that in a world of multiple crises, even major disasters in a seemingly peripheral state fall victim to neglect.
Another lesson refers to the general principle that respect for international law and norms will tend to diminish conflict, whereas violations will tend to stoke it. The behavior of state and presumptive state actors in the Rwandese conflict was in this respect less than adequate, and mostly counter-productive. International law and associated principles designed to uphold international order were repeatedly violated, including sanctity of national borders and arms embargoes. International refugee law was not observed. The legal right and moral obligation to intervene to stop genocide was acted upon. Human rights law was repeatedly and severely transgressed with impunity. Donors continued to give economic aid, and, in one case, also substantial military assistance, to a government linked to systematic violations of human rights. While this might be expected in a world of competitive nation-states observing the rules of Realpolitik, it should also be recalled that the consequences of lawlessness in this case were exceedingly costly - both for the Rwandese parties and people, and for outside states concerned.
International efforts to manage the conflict - as distinct from addressing its humanitarian consequences - were a failure when judged against international norms governing the prevention or mitigation of genocidal violence, as well as other UN standards. By explicitly asserting the right to intervene in a world otherwise governed by principles of national sovereignty, the Genocide Convention implicitly affirms the associated moral obligation to act. The United Nations and its members had already assumed some responsibility for conflict management in Rwanda by establishing a peacekeeping operation to help implement the Arusha Accords. When the Accords failed and the genocide commenced, the UN formally acknowledged an obligation to help protect civilians. This recognition came belatedly in the form of a Security Council resolution on 17 May 1994 - i.e. six weeks after the genocide commenced. The tardiness in authorizing and deploying a protective force weakened its operational effectiveness, but not the validity of the principle.
7.2.1. Critical actors at critical junctures
While policies were shaped by underlying structures of power and interest, the various actors nevertheless operated within a framework of options that permitted a certain choice. Among these, a few stand out as commendable efforts to reduce or mitigate the conflict; others are conspicuous for failing at critical moments. Significant in this respect are:
The Tanzanian government and the OAU efficiently and patiently spearheaded a sustained international effort to mediate the civil war. While the Arusha Accords in retrospect were not as perfect as the sponsors claimed, the agreement did provide a reasonable basis for settling the civil war. Partly undermined by unforeseen developments in neighboring Burundi, the peace agreement also carried some seeds of its own destruction by failing to take care of the losers. Having been excluded from the settlement and not dealt with otherwise, the Hutu extremists became an obstacle to its implementation.
UNAMIR on the ground was an exercise in doing the most with as little as possible. The Force Commander struggled in vain to bring the mission up to authorized strength and to alert an unresponsive UN Secretariat (DPKO) to the fact that the situation in early 1994 was seriously deteriorating. Once the civil war resumed and genocide commenced, UNAMIR was able to save lives by protecting some 15-20,000 persons in the Kigali area, despite lack of supplies and heavy equipment and a drastically scaled-down force. Its credibility and hence ability to modify the conflict in other respects (e.g. by promoting a cease-fire) was critically undermined by the decision to withdraw.
Of the humanitarian agencies, ICRC in particular helped provide protection and assistance to thousands of civilians.
After a bungled start and initial paralysis by the DPKO, the UN Secretary-General recovered the initiative by proposing that the UN re-engage itself to mitigate the conflict. The Under Secretary- General of Humanitarian Affairs helped focus attention on the crisis during the early period and was the first high-level representative from New York to visit Kigali after 6 April.
While failing to anticipate that some 1.5 million persons would flee across the border (see Study III), UNHCR soon recognized the need to deal with the problem of militarized refugee camps so as to avoid the start of a new conflict cycle. The High Commissioner's innovative proposal was bogged down in discussions at UN/New York, where the Secretary-General sought to address the problem comprehensively as a peacekeeping matter. This failed, however, and the result was a critical delay of several months before UNHCR could move ahead to deal with the camps issue.
Of the regional states, Uganda and Zaire at various times provided support that served to escalate the conflict rather than reduce it, though Uganda also contributed to the mediation process.
Through its military and economic assistance, as well as diplomatic support, France gave significant and sustained aid to a regime that was linked to systematic human rights violations and ultimately the genocide.
The Belgian decision to withdraw its UN contingent at the time of crisis crippled UNAMIR and drastically reduced its options for the future. Belgian lobbying for the Security Council to withdraw the remaining force altogether helped shape the final decision to this effect.
By acts of omission, the United States ensured that neither an effective national response nor a collective UN effort to mitigate the genocide materialized. Citing financial restraints, the United States wanted a bare-bones UNAMIR before 6 April, argued for withdrawal soon afterwards, and delayed the authorization as well as deployment of an expanded UN force in May-June.
The media by and large covered events in a fundamentally irresponsible manner. Notable exceptions were reports from correspondents for the BBC, Le Monde, Libération and The Times (London). The rest of the media reported in a highly selective and initially misleading way, _although some later were quick to charge Western governments with complicity in the genocide. Genocide and politicide were at first depicted as tribal-ethnic conflict with ancient and typically African roots. Coverage did not become intense until the genocide was over, and huge refugee flows streamed into Goma. The failure of the media to report accurately and adequately on a crime against humanity significantly contributed to international disinterest in the genocide, and possibly to the crime itself.
In a more general sense, the failure to respond was also linked to inadequate structures or procedures in decision-making. Some problems arose from the relationship between the OAU and the UN, and point to the importance of including - one way or another - the relevant regional forces in the decision-making order to achieve a durable solution to a local conflict.
7.2.2. Decision-making structures
Regionalism and the Disjuncture between Mediation and Implementation
From the beginning (1990), regional states participated in diplomatic efforts to deal with the conflict. While two of them were also involved in the war (Uganda and Zaire), that was considered all the more reason to include them in the process, which was spearheaded by another neighboring state (Tanzania) and the OAU. The latter's interests were primarily in a higher rational order that, among other things, would spare the region another large refugee problem. Over time, all external parties to the conflict appeared to gain a stake in the success of the mediation effort, and all signed the final Arusha document. There was, however, no continuity between the mediation and implementation phase of the peace accords. The UN took over the peace operation, essentially closing off the regional effort and cutting out the OAU. The consequent structural disjuncture between mediation and implementation affected three critical aspects that made it difficult to maintain the momentum of the Arusha peace process. These were:
Mandate and expectations: The Arusha Accords presumed and specifically called for a peacekeeping force with a mandate broader than the UN was willing to provide.
Timing: The timetable of the Arusha process presumed a speedy international presence, yet this did not sufficiently take into account, nor was it adequately coordinated with, the time-frame for establishing UN peacekeeping operations.
The extremists: The existence of an extremist group that was excluded from power in the peace agreement, but which continued in positions of power in the interim, should have been dealt with by other means. In shifting responsibility from one set of actors in the mediation phase, to other actors in the implementation phase, the critical issue of extremists was neglected.
In important respects, Rwanda obtained a peacekeeping force rather different from the one the negotiators of the agreement had anticipated and deemed necessary for the implementation of the Accords. The disjuncture was partly caused by organizational competition between the OAU and the UN. The OAU actively sought to obtain a leading role in the peacekeeping phase as well, but the Security Council insisted that the UN would not pay unless it was given command-and-control of the operation. Except for France, which lobbied hard to make it a UN force, and the Rwandese government, which happened to be a member after January 1994, no other states in the Security Council gave the issue much consideration. The UN Secretary-General weighed in on the side of his own organization. Moreover, the dismal record of the OAU in managing its minor military mission in Rwanda (NMOG I and II), gave ammunition to critics in the UN who maintained that the OAU had neither capacity nor the required impartiality to play a major role in a peacekeeping operation. As it turned out, however, UN Headquarters did not run an effective peacekeeping operation in Rwanda either, and UN neutrality in the face of genocide became a matter of criticism rather than approbation.
Issue of Stakeholders
While the United Nations had formal responsibility for helping to implement the Arusha Accords, the main stakeholders in the peace agreement were in the region, not in the Security Council. The latter consequently gave UNAMIR a narrow mandate, a limited budget, and scant attention. When the Arusha Accords appeared to unravel, the Security Council threatened to withdraw rather than strengthen UNAMIR. When the crisis erupted after 6 April, there was a "stampede to get out", as one member of the Security Council described the reaction of the chamber. At this critical juncture, the department of the UN Secretariat most directly responsible (DPKO) showed neither initiative nor an ability to rise above its assessments of what "the traffic would bear" in the Council.
African states showed more interest, partly reflecting the notion that African states had a special responsibility for solving their own conflicts. States in the region had particularist interests as likely receiving countries for massive refugee flows. Except for the military observers, the African battalion in UNAMIR I was the only unit that stayed put during the crisis; Ghana decided to keep its contingent while the Belgian and Bangladeshi governments withdrew theirs. During subsequent UN efforts to manage the conflict retroactively by reintroducing a force, only African countries offered to send troops. Operating under severe resource restraints, they required external financing and some equipment, to which the major powers only reluctantly agreed.
In sum, the regional contribution to conflict management was undervalued and underutilized despite recent and formal recognition by the UN leadership, including the Secretary-General, of the need to involve regional organizations to promote international peace (cf. Agenda for Peace, 1992).
7.2.3. Decision-making process
The United Nations
UN procedures governing peacekeeping operations made for slow and incomplete deployment. While this was not specific to Rwanda, the consequences were more acutely felt because the country had no "patron" in the Security Council that could cut through the political and bureaucratic morass. The Secretariat for the most part proved unwilling or unable to compensate. Before the acute crisis erupted on 6 April, DPKO, which had responsibility for UNAMIR, made no contingency plans or efforts to strengthen the mission's preparedness for worse-case scenarios despite clear evidence of mounting tension. The problem was partly due to limited institutional capacity in the face of a rapid increase in peacekeeping operations worldwide. The restraints were also more deep-seated, as indicated by the failure of DPKO to overcome a near-paralysis when the crisis broke and Rwanda moved to the top of the agenda. DPKO's communications to the Security Council were tailored to expectations of what the Security Council would approve (thus giving the Permanent Five anticipatory vetoes); options were formulated in terms of standard operating procedures, rather than the unique needs of the situation; and instructions to the field were heavily influenced by a concern to reduce risk so as to avoid "failures" - which were defined as the death of UN peacekeepers. Only when the Secretary-General in late April decided to provide some leadership did the Secretariat play an innovative and proactive role, based on more comprehensive and independent assessments of the requirements of the Rwandese situation. At that point, however, the limitations of retroactive conflict management, and inability of the UN to respond with dispatch, rendered the operation largely irrelevant.
The critical importance in this case of leadership touches broader issues of accountability and transparency in the continuing discussion of UN reforms. Given the large number of crises in the world, the Secretary-General can not respond equally effectively to all. However, the criteria for selection, and the process that leads the Secretariat to highlight some conflicts rather than others, remain obscure to the public. Similarly, the fateful decision virtually to withdraw UNAMIR was taken by the Security Council in informal consultations. In keeping with normal Security Council procedures, only select formal statements made at the conclusion of the decision-making process (21 April) were recorded in the proceedings and are thus in the public domain.
The effectiveness of UNAMIR also suffered from characteristic features of UN peacekeeping operations:
The field mission had very limited authority to make decisions; routine matters as well as issues heavily dependent upon judgement of the situation in the field were micro-managed by New York.
When deployed in late 1993, UNAMIR I had only a small investigative unit and no separate human rights component designed to monitor and report on human rights violations. This limited its ability to gauge a deteriorating situation. It also signalled to the Rwandese parties that the UN placed low priority on human rights violations in implementation of the peace accords.
The normal, slow process of assembling a peacekeeping force had monumental consequences. Once the Security Council decided in May 1994 to re-engage the UN for the explicit purpose of protecting civilians, a period of about three months elapsed before new UNAMIR units arrived in Rwanda. In the intervening period, hundreds of thousands were killed. The deployment was particularly slow because none of the industrialized states that could have provided the UN with fully-equipped units at short notice were willing to do so. The solution was to match Western equipment with African troops, a process that was time-consuming.
7.2.4. Policy issues
The international response to the Rwandese crisis also raised several policy issues of general concern:
National Peacekeeping Operations with a UN Umbrella. The French Opération Turquoise was quickly deployed and efficiently run and some held it up as a model. It should be noted, however, that the smooth operation was not only a function of its being national (as distinct from cumbersome and multilateral). Effective communication with the RPF made it possible to agree on the boundaries of the French "safe humanitarian zone" and, behind this line, French troops were in generally friendly territory. The operation also raised some troublesome issues. Mounted at the time the UN was struggling to obtain troop commitments for UNAMIR II, the French initiative detracted from the collective effort and weakened the credibility of multilateral interventions. The circumstances of the operation cast doubt on official claims that it was a purely humanitarian intervention, thus degrading the latter concept.
Democratization. International insistence that Rwanda's political system be democratized had contradictory implications. The formal institutions of multipartisme provided a political framework for accommodating the RPF and, in that respect, foreign donors promoted a formula that would help terminate the civil war. On the other hand, democratization so defined did not address human rights issues, and in some respects became a substitute for dealing with them. The results of efforts to strengthen civil society were also ambiguous in that they provided space for both human rights organizations and extremist groups. More generally, donor demands for democratization added to the overall pressures experienced by a regime that was at war and faced national economic collapse as well as mounting internal opposition. The combined effect was a sort of system overload that enhanced the power and possibly the appeal of Hutu extremists. These dysfunctional consequences were not sufficiently recognized at the time in the belief that the multi-party system entailed by democratization and endorsed by the Arusha Accords would take care of the extremists in due course.
Preventive Diplomacy. In retrospect it can be clearly seen that the closer the parties at Arusha came to end the civil war, the more Rwanda inched towards disaster. The inter-connectedness of the two tracks - the civil war and the civil violence - was recognized at the time, but the Arusha process was basically designed to settle a war, not to prevent a dimly perceived future catastrophe. If there is a lesson here, it lies in the tested principle that a settlement to end one war can be the beginning of a new one unless the peace agreement is also constructed as preventive diplomacy.
Refugees. Two principles of refugee policy are generally recognized in the international community: refugees have the right to return, and those who cannot return should be given asylum or resettlement elsewhere. Failure to observe such principles typically creates festering refugee problems, and in many cases militant communities who seek to escape from their dilemma by force. The phases of the Rwandese conflict considered here started and ended with festering refugee problems. While cognizant of the problems preceding the 1990 invasion, UNHCR could rely only on its good offices to promote a solution. Significantly, the agency lacked even the capacity to assess adequately the magnitude of the problem, let alone other means to pursue what in the early 1990s came to be known as a "comprehensive refugee policy", that is, an approach that addresses causes and solutions to refugee problems, not only immediate material and protection needs (A/AC.96/799,1992).
The formation of new militant and militarized refugee communities in Zaire during the second half of 1994 indicated the beginning of another conflict cycle. The response of the UN body responsible for refugees, UNHCR, reflected increasing awareness of the problems posed by such communities. _It was recognized that a solution, if any, requires disaggregating the problem. This entails separating the refugees from the military and dealing with the former within a legal-political framework of repatriation or resettlement. The latter constituted a military-security problem and had to be dealt with as such. By early 1995, reasonably effective arrangements were made to improve policing of the camps in both Tanzania and Zaire. Yet the issue of the armed militants and their links with the refugees remained unresolved as the armed groups were free to rearm and regroup near the border.
A solution to this problem required, in the first instance, greater cooperation from the authorities in Zaire.