by Susan Rice
Personally, I never fully got over my own frustration with Somalia; for the next two decades, I maintained a studied ambivalence to conflict resolution there. Knowing well the costs of failure, I would struggle with whether the potential benefits of reengaging in the thankless task of trying to help Somalis reconcile and rebuild were worth the enduring risks. The greatest cautionary tale for me was how essential it is for the cabinet-level officials to remain hands-on in decision making whenever U.S. forces are engaged in combat. Their deputies, however skilled and effective, need full support when crises occur. These hard-learned lessons from Somalia stuck with me through the following twenty-five years, as I rose through the ranks of national security policymaking.
Wrestling with the explosion of complex U.N. missions globally, the new Clinton administration sought to define and codify a rational and sustainable basis for U.S. support to U.N. peacekeeping operations. We believed in the value of the U.N. and wanted peacekeeping to endure as a vital tool of conflict resolution. We recognized the need to do our part but refused to become the world’s policeman and bear a disproportionate share of global costs and risks. Congress was deeply skeptical of U.N. peacekeeping and, after Somalia, any remaining public tolerance of perceived U.S. overreach had evaporated.
Early in the administration, President Clinton signed a directive mandating that we develop a new U.S. policy on U.N. peace operations—defining how and when we should decide to vote for, support, fund, and terminate them. As the NSC director responsible for U.N. affairs, I played a central role in crafting that policy along with Dick Clarke and Randy Beers. It was a complex bureaucratic “goat-rope,” a term Dick used to describe the challenge of corralling and reconciling the competing views of the various government agencies with much at stake in the outcome.
The Pentagon wanted as little to do with the U.N. as possible and viewed peacekeeping as an unworthy sissy mission. The State Department and U.S. Mission to the U.N. favored more robust U.S. support for the U.N. and saw peacekeeping as a comparatively low-cost way to share the global burden of preventing and resolving regional conflicts. My personal instincts aligned largely with State’s but, as an NSC director, my role was to be the honest broker, to try to find common ground and forge compromise.
It took months to develop the policy and months more to build interagency support for its key elements. As we did, we were constantly buffeted by unfolding world events—the U.N.’s challenges in Bosnia, Haiti, Cambodia, Somalia, and eventually Rwanda. From each, we learned different yet valuable, and sometimes competing, lessons. We worked to refine the impending peacekeeping policy, incorporating our learning in real time. The result was Presidential Decision Directive 25 (PDD-25): “U.S. Policy on Reforming Multilateral Peace Operations,” issued May 3, 1994, almost a year and a half after the policy review process began. It was a comprehensive, thoughtful, and well-conceived document, of which I remain proud. It has endured as standing U.S. policy through at least three administrations—Clinton, Bush 43, Obama—and remains, at the time of this writing, on the books.
PDD-25 recognized the strengths and limitations of U.N. peacekeeping and mandated increased U.S. support to improve the U.N.’s capacities, while placing wise and appropriate constraints on U.S. involvement. Yet in many ways, it was a conservative, cautious document, tempered in its initial enthusiasm by the weight of hard-won experience. It was meant to ensure we did not allow the U.N. or ourselves to overreach or misalign U.N. mandates with resources. In that, it was largely successful. Inadvertently, it also served to constrain our imagination and hinder our ability to react swiftly to the unforeseen. In a tale of unintended consequences, PDD-25 may have contributed to one of the Clinton administration’s most difficult and controversial decisions. Or, more precisely, lack of decision.
To this day, I am haunted.
When I stepped off the helicopter, I had no idea what I was about to see. It was December 1994, and I was among a small party of Washington officials who accompanied National Security Advisor Anthony Lake to a church and adjoining school in rural eastern Rwanda. A few yards from the landing site, the ground became thick with decomposed corpses. They were jammed so tight that it was treacherous to walk without stepping on one. Grown bodies. Children’s bodies. Babies in mothers’ arms. Shot up. Hacked up. Decaying faces. Frozen in the position they fell. All over the churchyard. In the church. Across the school grounds. Dead bodies everywhere. Hundreds, if not thousands. And the putrid smell, though ebbing, remained.
This was one of countless massacre sites across this verdant, hilly, densely populated Central African country the size of Vermont. Ground zero of the Rwandan genocide. Up to one million killed in a preplanned slaughter of ethnic Tutsis and moderate Hutus at the hands of their fellow citizens—extremist Hutus and those they forced into violence. The primary killing machine was the simple but lethal machete. They murdered each other by hand with long knives at close range. It was unspeakable cruelty that left me without words or even tears for many, many hours after we left the church site.
This brutal question still loops in my brain: How did we, the U.S. and the international community, let the genocide happen?
I remember the news flash, April 6, 1994, just one week after the U.S. military withdrawal from Somalia, another report from the White House Situation Room: the plane carrying the presidents of Rwanda and Burundi was shot down near the airport in Kigali, Rwanda. As the U.N. director on the NSC staff, I knew this was ominous. Rwanda was already on a knife’s edge, and neighboring Burundi had endured its latest spasm of Hutu-Tutsi violence just months earlier in October 1993, when an estimated fifty thousand Tutsis were killed after Tutsi soldiers had murdered Burundi’s Hutu president.
Still, few predicted precisely what happened next.
Immediately after the plane went down, the Hutu-dominated Rwandan army carefully targeted their first victims. The next day, April 7, the rampant killing began. Radio Mille Collines—a hate radio outlet with government links—bombarded the airwaves, inciting Hutu to massacre Tutsi (“exterminate the cockroaches”), even identifying locations where those hiding could be found. The army went house to house, slaughtering. Hutu extremists murdered Rwanda’s acting prime minister in cold blood and then mutilated ten Belgian U.N. peacekeepers who were assigned to protect her. Kigali was a massive killing field. Within days, the entire country became a bloodbath.
The U.S. government’s first priority in such a crisis is always to evacuate American embassy personnel and other U.S. citizens. The French and Belgian militaries flew in to rescue their nationals. The U.S. military, which had pulled its last personnel out of Somalia barely a week earlier, was nowhere near in position to mount a swift airlift or NEO, noncombatant evacuation operation. U.S. Marines would have to come from ships in the Indian Ocean and could not arrive in time to make an immediate airlift feasible. Our embassy personnel would have to try to escape overland by convoy to Burundi, without dependable security.
Given the killing that engulfed the countryside, this was an extremely dangerous evacuation. Back in Washington, we were all gripped by the fear that we would lose our colleagues. The State Department, which had the lead in ensuring the safe departure of its personnel, set up a round-the-clock crisis task force within its Operations Center. Working long hours from the low-ceiling, bottom section of our double-decker office in the Old Executive Office Building, which had been vertically subdivided by a previous occupant (the notorious, if industrious, Oliver North), I pestered the task force and the Africa Bureau frequently for the latest updates. President Clinton drove over from the White House to State to show solidarity and buck up the duty officers on vigil. When, after two long days, all American embassy personnel were safely evacuated, the collective sigh of relief we emitted in Washington was deep but short lived. Unfortunately, U.S. personnel were not able to evacuate our Rwandan embassy employees, and we later learned many of our local staff were killed in the ongoing massacres.
With no offi
cial Americans on the ground, we lost our primary eyes and ears. There was little real-time, formal information on what was happening, just anecdotal reporting mainly from nongovernmental organizations, the NGOs, and limited flows from the U.N. force (the United Nations Assistance Mission for Rwanda, or UNAMIR), which was hunkered down in defensive positions after the Belgian peacekeepers were killed. Within a week, countries began withdrawing their contingents from the U.N. force, greatly reducing its size and efficacy and turning it into a bystander as the killing raged. It was not until a couple of weeks later, when the international press corps arrived, that American officials gained a fuller appreciation of the extent of the horror. As the nightly news carried nauseating footage of masses of bodies rushing down choked Rwandan rivers, it became increasingly clear that we were in the midst of a full-blown genocide, and neither the U.S., Rwanda’s neighbors, the U.N., nor its member states were prepared to halt it.
In Washington, after the very proximate trauma of Somalia, neither senior U.S. officials, nor I, ever contemplated sending U.S. forces back to an even more remote and unknown part of Africa for another humanitarian intervention. No one in the Clinton administration argued for direct U.S. military involvement. No one in Congress did. No major U.S. editorial board did either. This was a case when the most obvious option was beyond the realm of the conceivable. In addition, the utility of the greatest military on earth in stopping thousands of people from killing their neighbors with knives was questionable. We had just failed to stop a warlord equipped with machine guns mounted on SUVs—unsophisticated “technicals,” as they were called—from killing U.S. servicemen in Somalia.
In hindsight, what still strikes me most is the lack of discussion or debate. In those early weeks, we were reeling and did not have the customary policy meetings to decide what to do next. I, we, never proposed what seemed unthinkable—U.S. military intervention. It wasn’t that President Clinton decided not to intervene. As the genocide unfolded, the president never asked for, nor did his senior advisors present him with any decision to make on the matter.
Four years later, on his first visit to Rwanda, in 1998, President Clinton apologized for the international community’s failure to respond swiftly enough to the genocide. In subsequent years, Clinton called Rwanda his “personal failure” and his “biggest regret” as president. In 2013, on one of many subsequent visits to Rwanda, where his foundation is active, President Clinton mused that, if he had deployed even ten thousand U.S. troops, the U.S. might have “saved at least a third of the lives that were lost.” I don’t know on what basis he came to this assessment, but perhaps with a sizable U.S. military intervention joined by an even greater complement of American-led international forces, we could have made a difference, even if at great risk and cost.
Yet, in reality, so soon after eighteen American service members were killed in Somalia while assisting U.N. peacekeepers, Congress would likely have refused to support another U.S. military intervention in an African country that most Americans had never heard of. Had President Clinton actually deployed U.S. forces into Rwanda, Congress might have terminated funding for the U.S. intervention and perhaps all U.S. financial support for U.N. peacekeeping.
In the moment, as events unfolded, many of the policy choices that the U.S. government did actually make seemed logical. But in hindsight, they were unacceptable in the aggregate.
The first major choice came roughly two weeks after the genocide began. Once their peacekeepers were slaughtered, the Belgians decided to withdraw their battalion from the 2,500-person U.N. force. They appealed to the U.S. as a NATO ally to back their decision at the U.N. by giving their withdrawal some diplomatic cover. This meant agreeing that the U.N. mission, with its limited size and mandate, was incapable of addressing the killing and should withdraw for its own safety and credibility. The U.N. secretary-general, Boutros Boutros-Ghali, gave the Security Council three options: a) substantially augment the U.N. force and strengthen its mandate; b) reduce the force to 270 and limit its mandate to brokering a cease-fire and assisting as it could with humanitarian relief, while trying still to provide some meager measure of protection to civilians who had taken refuge with the U.N.; or c) withdraw UNAMIR entirely, recognizing that it was unsuited to the dramatically worsened context.
In deference to the Belgians, and consistent with the rigorous criteria of PDD-25, Washington instructed the U.S. Mission to the U.N. (USUN) to go with Option C: insist that UNAMIR withdraw. Meanwhile, African members at the U.N. and others argued to increase the force dramatically but offered no capable or ready troops to do so. In addition to the Belgians, the Bangladeshi and Ghanaian battalions were also withdrawing, leaving a minimal presence. Thus, Option A was a pipe dream in practice and, after the Somalia debacle, unimaginable to many even in principle.
The Security Council coalesced around the middle option of reducing the force, leaving the veto-wielding U.S. alone in supporting full withdrawal. I recall listening on the line when Ambassador Albright called Dick Clarke. She was as livid as I have ever heard her. Albright told Dick, in effect: “I have just stepped out of the Security Council. We are totally isolated. No one else agrees we should completely withdraw the force. Plus, my instructions from Washington are just plain wrong. The U.S. can’t be responsible for completely abandoning Rwanda. I need new instructions. Now.” As a result of her willingness to fight a bruising internal battle, Ambassador Albright ultimately received authority to join the consensus in the Council on the U.N. drawing down to 270 troops. Human rights groups and many others criticized this action as cowardly and deadly, given that the small, remaining U.N. presence was powerless to stop the killing or even to protect those whose had taken refuge among U.N. personnel. But the only option that could have made a difference—augmenting the force—was not feasible, because no country had both the will and the capacity to provide the necessary troops swiftly enough.
Another dilemma was whether to employ U.S. military assets to jam Radio Mille Collines. National Security Advisor Tony Lake rightly pressed the Pentagon to develop this option, but in a May 5 memo from Undersecretary of Defense for Policy Frank Wisner to Deputy NSA Sandy Berger, DOD concluded that, in the Rwandan context, “jamming is an ineffective and expensive mechanism that will not accomplish the objective the NSC Advisor seeks.” That memo effectively ended internal discussion over the utility of shutting down the hate radio broadcasts.
The horror of the killing sparked another active debate in May and early June—whether to label what was transpiring in Rwanda a “genocide.” For many weeks, major press outlets, including The New York Times, referred to events generically as a “civil war,” which in part it was. Along with the press and many analysts outside the human rights community, U.S. policymakers were slow to recognize the organized, systematic, ethnic-based nature of the killing. The U.S. government crept up to the designation, as State Department officials were eventually allowed to acknowledge gingerly that “acts of genocide may have occurred.” Typically, the secretary of state is responsible for making formal genocide determinations. The issue ultimately fell to the elder statesman Warren Christopher, a venerable and exceedingly cautious lawyer. He and some of his staff worried about the implications of the genocide designation, fearing that saying the word would create political pressure on the administration, if not a legal obligation, to intervene.
National Security Advisor Lake repeatedly characterized events in Rwanda as “genocide” in meetings with human rights groups, a judgment with which I agreed, given the scale and intent of the killing. Still, it was not until June 10 that Secretary Christopher publicly and reluctantly conceded: “If there’s any particular magic in calling it genocide, I’ve no hesitancy in saying that.”
As the extraordinary scope of the killing in Rwanda grew more obvious by early May 1994, pressure mounted at the U.N. and in Washington to do more—and fast. The U.N. secretary-general proposed to assemble an all-African force to fly into Kigali and move outward into
the countryside to protect civilians. U.S. military planners thought this concept was madness, given the dangers of inserting any sizable force by air into a hot war zone, much less a U.N. force of limited capacity. Along with others in the administration, I supported a counterproposal for a different concept of operations wherein U.N. personnel would enter Rwanda from the neighboring countries and establish safe zones along the border where civilians could be protected. For weeks, U.S. and U.N. planners argued over the right course. Eventually, lacking support for our concept, the U.S. relented, and Boutros-Ghali won agreement gradually to establish UNAMIR II, which would base itself in Kigali and work its way out of the capital.
Several African governments committed to participate, but their contingents were very slow to arrive, in part because the U.S. insisted on a phased approach tied to conditions on the ground, and in part because they were unable to deploy without external assistance and lacked basic equipment from radios to armored personnel carriers (APCs). To try to expedite their deployment, Dick Clarke and I pressed the Pentagon to lease forty-six APCs to the U.N. to enable the Ghanaian battalion to become operational. As one of the most capable contingents and a friendly nation, Ghana was a prime candidate for U.S. assistance. We pleaded with colleagues on the Joint Staff, “You can spare less than fifty APCs. It’s not going to affect military readiness.”
Still, the Pentagon hated the idea of relinquishing its vehicles and transporting them to Rwanda; above all, strangely, the Department of Defense balked at painting their APCs white, as was necessary to identify them as neutral U.N. vehicles. Dick and I spent countless hours battling with DOD, insisting, “You can easily repaint the APCs back to Army green when the lease is over.” Eventually we wore them down, and they agreed to send the APCs. Some months later, my NSC colleagues and I were deeply chagrined (but also mildly amused) to learn that at least some of our American APCs were being used as chicken coops!
