Journal of Humanitarian Assistance
The transition from peace negotiations to implementation was accompanied by mounting tension. One fundamental problem was related to the domestic dynamic: the closer the peace agreement came to be realized, the more desperate grew the extremists. While the Accords provided for a new government that many hoped would subdue the militant opponents, the installation of the BBTG was repeatedly delayed by parties who manipulated the transition instruments for narrow gains. As the implementation process floundered, unforeseen external events in neighboring Burundi heightened ethnic tension and strengthened the appeal of the extremists, thereby making implementation of a delicate political compromise even more difficult. The UN Assistance Mission in Rwanda (UNAMIR) had been designed to oversee a smooth transition and was not equipped to handle the situation. Nor was it strengthened despite increasing recognition that civil war might resume and large-scale civil violence erupt.
3.1. Implementing the peace agreement
To assist implementation, the Arusha Accords provided for the establishment of a neutral international peacekeeping force to be deployed in Rwanda within 37 days from the date of signing (i.e. by 10 September). The UN promptly sent a technical mission to Rwanda for needs assessment. Three issues arose out of that mission: the timeliness of deployment, the size of the force, and the precise mandate as well as rules of engagement for the force. Their resolution depended in part on the team's assessment of the situation in Rwanda, in part on what the UN members were willing to invest.
The Reconnaissance Mission, headed by the designated Force Commander, General Romeo Dallaire, reported a mixed picture. It was a foregone conclusion that the operation would come under Chapter VI peacekeeping, and the situational prerequisites were present: a peace agreement was signed, and the local parties consented and cooperated. In the meantime, both sides were observing the cease-fire agreement, which was monitored by both regional (OAU-NMOG) and international (UNOMUR) observers. To those in the UN who looked for a success to counter the troubled UN mission in Somalia, "Rwanda seemed like a winner", as one participant noted. In an unprecedented move that foreshadowed an easy mission, a joint delegation from both sides of the Rwandese conflict came to New York to lobby the UN to endorse the peacekeeping operation.
Yet Dallaire's team was also warned of instability and dangers. The schedule for deploying peacekeepers was unrealistically short, so time was of the essence. A transitional government required the participation of RPF members as designated, but the RPF absolutely refused to enter Kigali until the French forces had left, and the French would not leave until the UN arrived. Thus, the UN force held a key to the entire transition.
Four standards are relevant to measure the rapidity of deployment. On the standard of need, the troops were required as soon as the Arusha Agreement was signed. Closely following was the expectation of the Agreement itself, i.e. 10 September, which was generally agreed to be an unrealistic deadline. Third, there was the timeline the UN Secretariat set for itself as spelled out by the Reconnaissance Mission. In some important respects it was not met.[51] Fourth, however, standard UN procedures meant that 3-6 months would be needed to assemble and deploy a peacekeeping force, depending upon the size. According to this standard, UNAMIR was deployed fairly rapidly. The UN Secretariat started assembling the Reconnaissance Mission a few weeks before the Arusha Accords were signed on 4 August, and the first battalion reached Rwanda in late November.
Delays in establishing a transitional government would increase the likelihood that the entire peace agreement would be derailed, the Mission was told by several foreign diplomats and the Rwandese parties to the Agreement.[52] It was known that militant forces remained vehemently opposed to the Accords, and Rwanda's history of ethnic and political strife cast a long shadow. Recent violence and the growing strength of paramilitary groups added to the formidable problems of demobilizing some 35,000 soldiers in a society with a shattered economy, famine, and large numbers of persons displaced by the war. In this situation, concluded the UN Secretary-General, it was imperative that the international community demonstrate its commitment to implement the peace agreement it had helped to mediate. Delays would "seriously jeopardize" the Accords (S/24688/1993 para.65).
One military expert in the Secretariat estimated that ideally, a mission of this kind should have 8,000. Dallaire proposed 4,500 as his maximum option. In the end, the Secretariat recommended a force only half that size, anticipating that this was the maximum that the Security Council would approve. Having exercised an anticipatory veto, the Security Council on 5 October authorized a force level of 2,548 military personnel without much discussion.
The decisive restraint on the overall size of the mission was financing. Being assessed 31% of the costs of UN peacekeeping, the United States insisted on a minimal force. The Clinton Administration had just started an executive review of its UN policy, and was sensitive to Congressional concern over the mounting American share of peacekeeping costs, which had increased a stunning 370% from 1992 to 1993.53 Emphasizing the bright aspects of the Rwandese situation, the cost-conscious US delegation in New York suggested in September that a token mission of some 500 men would suffice. The French mission in New York recommended a small force of around 1,000 men, noting that the French contingent in Kigali was merely 6-700 men. The end result of 2,548 was more than a token force, and at that time considered quite acceptable by the Force Commander. UNAMIR was estimated to cost about US$10 million a month, a very modest amount compared to other UN peacekeeping operations.[54]
The signals from the Rwandese parties concerning force size were conflicting. A skeptical RPF reinforced calls in the Security Council for a minimalist force. Recalling earlier French lobbying at the Security Council to have a UN rather than an OAU force in Rwanda, the RPF suspected that France was using UN peacekeeping operations as a shield to protect its political allies in Kigali and secure a foothold for the future. The RPF consequently claimed that a few hundred men would do. By the same logic, the government of Rwanda wanted several battalions.
Despite a speedy start,[55] the deployment of UNAMIR did not fully meet the schedule laid out by the Reconnaissance Mission. The special representative of the Secretary-General did not arrive in late November, or three and a half months after the Arusha Agreement was signed. The first UNAMIR battalion entered Kigali shortly thereafter, enabling the French paratroopers to leave. Putting together the rest of the mission also proved difficult. The force never received the equipment authorized and required even at the reduced level (including an armored unit and helicopters).[56] UNAMIR's budget was subject to the standard, lengthy UN decision-making process and was not formally approved until 4 April, 1994.[57] As a result, the mission was constrained by numerous shortcomings in personnel, equipment and disbursable funds (including petty cash), and even basics such as ammunition. Repeated field requests to New York to bring the Mission up to authorized strength were to no avail.
The problem was partly structural, reflecting cumbersome UN procedures for budgeting and deployment of peacekeeping operations generally, as well a the political system in the UN. For all but one of the major powers on the Security Council, Rwanda was of peripheral interest. This meant that UNAMIR lacked a powerful patron in the Council to help cut through the bureaucratic-political morass that governed normal procedures for UN peacekeeping. Without this political pressure, UNAMIR moved to the bottom of the UN security agenda. Nuts-and-bolts issues as well as more fundamental mandate questions were neglected. Only the French government had a direct interest and considerable presence but, with some reason, claimed that such involvement restricted France to a low profile in the Council deliberations on UNAMIR so as not to compromise the neutrality required by a UN peacekeeping force.
The barebones UNAMIR had weak or no capacity in two areas that were significant for its operation. With only a small civilian police unit and no human rights cell, the mission had very limited ability to investigate violent incidents. The force also lacked an official intelligence unit, a fact that was deplored by the Force Commander and led to some improvization on the ground (see 3.2.)
3.2. Civil violence and the subversion of the peace agreement
In early autumn 1993, the Rwanda situation did permit some optimism that the peace agreement would be implemented in an orderly fashion. Yet underlying tensions soon surfaced to reinforce earlier warnings that the process was being subverted. The splitting of political parties, which had started already in late 1993, became more profound as factions competed for positions in the transitional government, thereby undoing the formula for power-sharing so carefully negotiated at Arusha and playing into the hands of those who opposed the Accords altogether. Repeated failures to form a transitional government made the diplomatic community in Kigali and states associated with the Arusha process renew their efforts to break the deadlock. The Tanzanian government took the lead, but Belgium, Germany, the United States, Canada and the Vatican also weighed in with strong language and sometimes high-level representations to repair the process. During February and March, and even right up until the crisis erupted on 6 April, concerted efforts were made to avoid a rupture.
The diplomats who labored to save the Arusha Accords saw implementation as a means to control the extremists. With Hutu moderates and RPF in charge of key ministries in the planned transitional government - including Interior and Justice - and a new national army to be formed, many expected that the Hutu extremists could be subdued. Also leading members of the RPF politburo appeared to take this view. However, the implementation process was fundamentally a race against itself: the closer the Arusha Accords came to be realized, the more the extremists were prepared to sink it in violence. The point was formulated unequivocally in a cable from the UNAMIR force commander to DPKO/ New York on 11 January 1994: A "very important government official" turned informer had told him that "hostilities may commence again if political deadlock ends".[58] Previously, "hostilities" had often been interpreted to mean renewed civil war. By early 1994, the term clearly conveyed the meaning of civil violence. It was quickly recognized by the SRSG, Jacques-Roger Booh-Booh, who reported to the Secretary-General soon after his arrival in Kigali on 23 October that the extremist adversary to the peace process was a powerful, subversive Third Force - ruthless, well-organized, well-equipped and centrally directed.
Additional information provided in the UNAMIR cable of 11 January to the UN Secretariat in New York was quite specific. The Rwandese government informant had revealed the extremists' plan to assassinate politicians at the scheduled ceremony for swearing in of the transitional government. In the process, they would provoke an encounter with Belgian UNAMIR soldiers, expecting that by killing some, the entire UN contingent would leave Rwanda. As the RPF would be instigated to resume war, 1,700 interahamwe who had been training in camps outside Kigali were staged to sow insecurity throughout the city; teams of 40 each were organized within Kigali to kill all the Tutsi who had been registered by the militias. The informer estimated his units could kill 1,000 persons per 20 minutes.
Assassinations on this scale were no longer of the order of a political tactic but clearly genocidal. The message dovetailed with a series of other developments, indicating that the situation was rapidly deteriorating. Both at the time and looking back, foreign diplomats and UN officials serving in Kigali clearly noted the worsening trend. The October events in Burundi represented a watershed. Rwanda's Hutu population projected developments in the neighbouring state onto their own society: a newly-elected Hutu president was assassinated and tens of thousands of Hutu were killed in the ensuing violence, many in situations where Burundi's Tutsi-dominated army was directly or indirectly responsible. With the RPF poised to share both governmental power and half of the national army in Rwanda, the message of fear of the Tutsi being spread by the Hutu extremists gained new credibility.
Additional information was picked up by the small, unofficial intelligence unit set up by the Belgian 1st paratroop battalion serving with UNAMIR under Col. Luc Marchal. The unit was financed directly by Belgium after the UN Secretariat declined to finance it, and reported to the Belgian military headquarters in Evere outside Brussels as well as the UN. During early 1994, the Belgian intelligence channels provided detailed secret information, including vouching for the bona fides of a key informant, detailing secret arms caches, and spelling out plans by MRND leaders, in cooperation with interahamwe units, to isolate, harass and undermine the work of the Belgian UNAMIR contingent.[59] Belgium also had another intelligence channel. The military intelligence agency (SGR) of the Ministry of Defense ran an operation attached to, and under cover of, the small military assistance unit that Belgium maintained in Kigali (mainly for medical training) until April 1994 (De Morgen, 4 November 1995).
Hutu supremacist rhetoric became increasingly vocal and public. Radio mille collines (RTLM), _broadcast hate messages that called on the people to rid the country of Tutsi. African diplomats reported CDR members telling them that in the future there would be no Tutsi left in Rwanda. Referring to the RPF as the "historical overlords", the RTLM whipped up fear by claiming that once the RPF army was in Kigali - as provided for in the Arusha Accords - Tutsi troops would attack the Hutu population.
In the streets, the violence was real. Assassinations of two high-level politicians in late February made the CDR organize a four-day rampage in Kigali with interahamwe mobs (readily recognized by the colors of their "uniforms") roaming the streets and killing dozens of people. The mob violence clearly targeted individuals, and even foreign diplomats considered "pro-Arusha" were threatened. The Tanzanian Ambassador was almost killed on 5 January when a CDR crowd attacked his car. The Papal Nuncio received grizzly death threats.
UNAMIR's Force Commander responded to the deteriorating situation by repeatedly requesting equipment to bring his force up to strength. In January, he sought guidance from the Department of Peace-keeping Operations in the UN Secretariat (DPKO) on actions to be taken if the situation worsened. In early February, he asked for clarification of the mandate to permit searches of arms caches. In late February and again in mid-March, he requested reinforcements.[60] Reports of arms being distributed to para-military groups in the Kigali area were frequently discussed by UNAMIR staff, reported by the Belgian military intelligence unit back to the Belgian government, and were circulated in the diplomatic community. Some went public with their concern, including the UNAMIR commander of the Kigali sector, Col. Luc Marchal, and the SRSG, Jacques-Roger Booh-Booh. Noting that violence was about to overtake the peace process, Booh-Booh complained to the press on 24 January that "weapons are distributed from arms caches around Kigali and even inside town."
Some arms shipments could readily be traced. On 22 January 1994, a planeload of arms from France was confiscated by UNAMIR at the Kigali airport.[61] The delivery was in violation of the cease-fire clauses of the Arusha Accords, which prohibited introduction of arms into the area during the transition period. Formally recognizing this point, the French government argued that the delivery stemmed from an old contract and hence was technically speaking legal. More serious, UNAMIR's Kigali sector commander has reported that a planeload of arms arrived from France in the early morning of 9 April, that is, shortly after the killings started. The French government has categorially denied this shipment.[62]
Arms distribution affected the immediate security situation as well as the climate for implementing the peace accords. It also went to the core of UNAMIR's mandate to help make Kigali a "weapons secure area" where weapons were strictly controlled. Failure to react to flagrant and illegal arms distribution clearly would erode UNAMIR's credibility. Immediately after sending the 11 January cable detailing the interahamwe plans for killing Tutsi in Kigali, Dallaire requested permission to seek out arms caches that the informant had offered to identify. In the UN Secretariat, however, the DPKO denied permission. On three occasions in early February, Dallaire made similar requests to carry out cordon-and-search operations to seize arms. Each time the DPKO turned down the request. Similar calls from the Belgian government to permit UNAMIR to interpret its mandate more proactively - issued after the Belgian Defence Minister Leo Delcroix had visited Rwanda in mid-March - were likewise rejected.[63]
In retrospect, it is hard to fault the UN for failing to anticipate the impact of external events, notably the October massacres in Burundi. The prescient and cumulative warning signals from Rwanda itself were more compelling. They were taken seriously by many in the field: diplomats redoubled their efforts while UNAMIR tried to respond. In New York, however, the Secretariat was passive or cautious. The most critical warning from the field - the 11 January cable - was put aside in the DPKO.[64] Requests to bring the force up to strength, let alone provide reinforcements, were rejected, although the deployment of the second battalion was speeded up. UNAMIR was denied permission to interpret its mandate proactively and search for arms. When the Secretary-General requested that the Security Council renew UNAMIR's three-month authorization, he did not mention the critical, new intelligence elements in the situation. The reasons why this information apparently was not reported to the Council are unclear.
New York's response reflected a cautious and politically sensitive interpretation of UNAMIR's mandate. The latter had been tailored to peacekeeping of a classic consent-and-cooperation kind. The mission was to "contribute to the security" of the Kigali area, monitor the cease-fire, and assist local authorities in demobilizing the two Rwandese armies and investigating violations of the Accords (Res.872(1993)). It is important to recall that this mandate grew out of but differed from what was envisaged in the Arusha Accords. In the central clauses defining UNAMIR's role in providing security, in protecting civilians, and in confiscating illegal arms, the Arusha Accords were significantly broader than the terms of the final UN mandate.[65] The disjuncture between the mediation phase and implementation was again revealed: the signatories to the regionally-brokered peace accords had evidently assigned an importance to the peacekeeping force that exceeded what the UN was willing to provide.
In the fall of 1993, the UN Secretariat and the Security Council had considered Rwanda a low priority, drafted a classic and minimalist peacekeeping operation, and hoped for the best. As the situation in Rwanda progressively deteriorated, the Secretariat and the Security Council members - in particular the United States - labored under the impact of the Somalia experience. The crisis in Somalia, where US/UN troops became embroiled in a de facto war with a local faction and took casualties, had been politically costly both for the US government and for the UN Secretary-General. Not crossing "the Mogadishu-line" became a firm dictum in the Secretariat, as a high-level official in the DPKO later put it. A proactive interpretation of UNAMIR's mandate to permit confiscation of illegal arms was much too close to the Somalia syndrome for comfort.[66]
Having ruled out a proactive and coercive response in Rwanda, the Secretariat and the Security Council left themselves only one option in case of a crisis: withdrawal. The strategy was spelled out by Boutros-Ghali in the form of a bargaining tactic: unless the peace agreement was implemented, the UN would withdraw its peacekeeping force. The message of the Secretary-General was reiterated by the Security Council when it renewed UNAMIR's authorization in January and again on 5 April.
As the crisis intensified, the range of options to save the peace agreement narrowed. Nevertheless, in Kigali, UNDP proceeded with the planning of the demobilization program by bringing the RPF and Rwandese army officers together to prepare the details, including the rate of demobilization and measures to reintegrate the soldiers back into the civilian economy. The planning process - which was sustained for a full five months - was in itself a confidence-building measure. In a similar spirit, donors offered economic incentives by holding out the promise of large development funds once the new transitional government were installed. Using a different tactic, the UN Secretary-General and the Security Council threatened to withdraw UNAMIR, although this was what the Hutu extremists wanted.
In retrospect, it is possible that by early 1994, Habyarimana himself was a prisoner of the extremists. If so, putting pressure on the President would yield few results. The option of strengthening UNAMIR and applying force against the extremists - which theoretically was available to the UN - might have been effective. This option was not considered; allowing UNAMIR to show a bit of muscle was rejected. While this might not have resulted in the installation of the BBTG, constraining the extremists could have modified the violent consequences.
It is not a foregone conclusion that a stronger or more proactive UNAMIR would have altered the course of events leading up to 6 April. The consequences of weakness, however, were reasonably clear. Slowly and haltingly deployed, UNAMIR failed to project strength and decisiveness from the beginning. As the struggle over implementation intensified, the force was constrained from acting in matters pertaining to the security of Kigali and thus central to its mission. Its presence gave a false sense of security to Tutsi as well as Hutu regime opponents and moderate coalition members who, feeling themselves protected, spoke out against extremists and coup-plotters. All of them, a close observer later wrote, "knew what the Hutu extremists were capable of, but many ... were persuaded that the presence of UNAMIR would prevent things from going much further than the repression in October 1990."[67] When the situation radically changed and widespread violence erupted, UNAMIR lacked everything from sandbags to armoured personnel carriers to protect either its own troops or civilians.