WHEN, IN 2006, Joseph Kabila became the first democratically elected president of the Democratic Republic of the Congo, many Congolese and international observers hoped that stability had finally come to the country. During the previous decade, Congo had been ravaged by widespread violence, including the world's deadliest conflict since World War II--a conflict that involved three Congolese rebel movements, 14 foreign armed groups, and countless militias; killed over 3.3 million Congolese; and destabilized most of central Africa. In 2001, the United Nations dispatched to the country what was to become its largest and most expensive peacekeeping mission. A peace settlement was reached in 2003, paving the way for the 2006 elections. The entire effort was touted as an example of a successful international intervention in a collapsing state.

        Yet over two million more Congolese have died since the official end of the war. According to the International Rescue Committee, over a thousand civilians continue to die in Congo every day, mostly due to malnutrition and diseases that could be easily prevented if Congo's already weak economic and social structures had not collapsed during the conflict. In mid-2007, in the eastern province of Nord-Kivu, low-level fighting between government forces and troops of the renegade Tutsi general Laurent Nkunda escalated into a major confrontation, both playing off and exacerbating long-standing animosity between the Tutsis, the Hutus, and other groups. Since then, clashes have killed hundreds, maybe thousands, of fighters and civilians and forced half a million people to relocate. Congo is now the stage for the largest humanitarian disaster in the world--far larger than the crisis in Sudan.

        The international community has admittedly been facing a very complex situation: all the parties have legitimate grievances, but all are also responsible for massive human rights violations; the fighting involves many armed groups, and these often fragment and shift alliances. Still, the main reason that the peace-building strategy in Congo has failed is that the international community has paid too little attention to the root causes of the violence there: local disputes over land and power. If anything, international efforts to bring peace have enhanced local tensions. While it focused on organizing the presidential, legislative, and provincial elections of 2006, the international community overlooked other critical postconflict tasks, such as local peace building and overhauling the justice system. Meanwhile, the electoral process fueled ethnic hatred and marginalized ethnic minorities, making the reemergence of armed movements all the more likely.

        The international community must fundamentally revise its strategy. It must focus on local antagonisms, because they often cause or fuel broader tensions, and regional and national actors hijack local agendas to serve their own ends. Until the local grievances that are feeding the violence throughout eastern Congo are addressed, security in the entire country and the Great Lakes region overall will remain uncertain.

YOUR LAND IS MY LAND

        TENSIONS AT the levels of the individual, the family, the clan, the village, and the district are a critical source of instability and violence in Congo. Control over land, especially, has historically been a major bone of contention in rural areas because the stakes are high and the interested parties numerous. Land matters because for many people it is the key to survival and feeding one's family. For many more, it is both a primary method of gaining the social capital needed to integrate local structures and a means of securing natural resources.

        In the territories of Masisi and Walikale, in Nord-Kivu, different ethnic groups, clans, and families are fighting over competing claims. There are centuries-old antagonisms among native Congolese communities, such as the Hundes, the Nandes, and the Nyangas. But the fiercest disputes oppose them to Congolese of Rwandan descent. In the early part of the twentieth century, Belgian colonial administrators relocated over 85,000 people, both Hutu and Tutsi, from overpopulated Rwanda to the sparse Kivu provinces in Congo, and in the 1960s and 1970s various waves of Tutsis fled there to escape pogroms in Rwanda. Today, Congolese of Rwandan descent, especially the Tutsis among them, own most of the land, but the Hundes and the Nyangas continue to claim it as their own on the grounds that it was never rightfully sold or given away.

        These competing claims have gotten far more complicated since the 1990s, as the 1994 genocide in Rwanda and various wars, invasions, and refugee movements caused multiple shifts in the ownership or control of land in the Kivus. Many Tutsis in the region, in particular, whether Congolese or Rwandan, have fled prosecution several times over the past decade, abandoning their plots or selling them at a discount and then claiming them back again, sometimes by force, on their return. The provincial authorities have resolved some of these disputes since the peace deal in 2003, but land ownership is at the core of the current fighting in Nord-Kivu. Throughout eastern Congo, historical grievances of this kind also fuel battles between (and within) dozens of mini factions from different tribes, clans, and families--such as the Hemas and the Lendus in Ituri, in the eastern part of the province of Orientale, and the Bembes, the Holoholos, and the Kalangas in northern Katanga--and greatly impede the peaceful return of refugees and displaced persons.

        Control over land is also a ticket to natural resources. Congo has massive reserves of gold and diamonds, most of the world's columbo-tantalite and cassiterite (essential materials for most electronic equipment), and many deposits of rare minerals. Since the end of the war, most of the local ethnic militias in northern Katanga, which are known as the Mai Mai, have regrouped around mining sites throughout the region and fought among themselves or against soldiers of the national army for their control. In 2005, in the town of Shabunda, in Sud-Kivu, soldiers pitted persons with competing claims over mineral-rich areas against one another and then disarmed them when small-scale violence broke out--only to exploit the concessions for themselves or hand them over to third parties. Provincial and national commanders were reportedly bribed into looking the other way.

        In most cases, economic tensions feed politically motivated hostilities, and vice versa. Access to resources means the ability to buy arms and reward troops, and thus to secure political power; political power, in turn, guarantees access to land and resources. Tensions between the so-called indigenous communities and people of Rwandan descent (who are often still considered immigrants even though many of them have lived in Congo for generations) also influence claims over political representation. In Nord-Kivu, the Hutus and the Nandes, the province's two largest ethnic groups, have fought each other over control of provincial politics. Factionalism and shifting alliances complicate matters further. In each village, different members of the same family or different branches of the same clan compete to be designated chief under traditional law. In 2002, Hunde and Nyanga elites fought large-scale battles for control of the town of Pinga, in Nord-Kivu. Hutus and Tutsis of Rwandan ancestry, who had combated indigenous groups together during the late 1990s and early 2000s, split apart in 2006, after a law confirmed that most of the Hutus among them were also Congolese citizens, with rights to land ownership and political representation, thus making the alliance less important to them. Since then, they have tried to partner with the Nandes, who won leadership of Nord-Kivu in the 2006 elections. As a result of this shift, the province's Tutsis have lost hope of gaining political representation and become both more marginalized and more radicalized.

THE CIRCLES OF HELL

        FOR DECADES, these local tensions have also fueled broader struggles at the regional and national levels--and, at times, the other way around. Both Congolese and foreign politicians have long manipulated local leaders and fragmented militias to enrich themselves, advance their careers, or rally support for their causes. Local actors have also recruited national allies. For example, in 1963, three years after Congo's independence, tensions over access to land and representation in local administrations in Nord-Kivu led to tremendous violence between the "indigenous" groups and the "immigrant" ones. To undermine the "immigrants"' claims over land, the "indigenous" communities contested their Congolese nationality; the "immigrants" then turned to national politics for an alternative strategy. They won the backing of then President Mobutu Sese Seko, who favored promoting ethnic minorities because they could help him govern without threatening his regime. Several people of Rwandan descent thus got top political positions, which they leveraged to help other people of Rwandan descent increase their own economic, political, and social power, notably in the Kivus. Still, in the early 1980s, the "indigenous" lobby managed to get a law passed denying "immigrants" Congolese citizenship. The measure was not implemented, but it jeopardized the political and economic status of people of Rwandan descent and strongly reinforced their fear of disenfranchisement. Throughout the 1970s and 1980s, tensions over land and power caused frequent skirmishes in Nord-Kivu.

        These problems exploded in the 1990s, this time with a regional dimension. In 1994, following the genocide in Rwanda and the Tutsis' subsequent rise to power in Kigali, one million Rwandan Hutu refugees, including many militia members, flowed into the Kivus, bringing with them raw rivalries from home. Indigenous Congolese groups of all stripes organized themselves into Mai Mai forces, and many allied themselves with the defeated Rwandan Hutus, who were thankful for any support that would help them survive in Congo's jungle and for access to mining resources and thus a means to buy arms. The interests of Paul Kagame's newly empowered Tutsi government in Rwanda converged with those of the Congolese Tutsis. Both sides originally intended merely to protect their kinsfolk, but they quickly started using their military might to seize land or capture political power.

        The fighting in the Kivus quickly evolved into a full-scale regional and national war. In 1996, the growing unpopularity of the Mobutu regime among Congo's neighboring countries, as well as in the West, prompted the formation of an alliance among a Congolese rebel group with many members of Rwandan descent; the governments of Rwanda, Burundi, Angola, and Uganda; and southern Sudanese rebels. Within a year, the coalition overthrew Mobutu and replaced him with its spokesperson, Laurent Kabila. When the Rwandan army invaded Congo to support the rebellion, it had two basic objectives: hunting down Rwandan Hutu rebels in the Kivus and protecting the Congolese of Rwandan ancestry there. It soon developed a third: exploiting Congo's mineral resources.

        Once in power, Kabila quickly turned on his former allies. He fired his Rwandan advisers, ended Congo's military cooperation with Rwanda, and began inciting the population to racial hatred toward Rwandans and Congolese of Rwandan ancestry. With these groups feeling increasingly threatened, in 1998 the governments of Rwanda, Burundi, and Uganda helped engineer a new rebel movement led by Congolese Tutsis. This alliance was less successful than that of 1996 because it met with opposition from the governments of Angola, Namibia, and Zimbabwe, which sided with Kabila. The conflict quickly turned into a stalemate, with a fierce guerrilla war raging in the eastern provinces. Kabila managed to contain Rwanda and its allies for several years thanks to local proxies, the Mai Mai and Rwandan Hutu militias. In the meantime, however, people of Rwandan descent and Rwandan elites developed lucrative networks for trafficking resources. The Rwandan army officially withdrew from eastern Congo after the peace deal in 2003, but part of the Rwandan establishment has continued to unofficially provide financial, logistical, and military support to Congolese fighters of Rwandan origin there.

        Over the past few years, these long-standing local disputes in eastern Congo have also been exacerbated by political developments at the national level. For example, many experts argue that Hema and Lendu factions from Ituri have been violently asserting themselves partly in reaction to their having been excluded from the lengthy peace process that ended the last war in 2003. Similarly, the highly selective fashion in which national actors picked Mai Mai representatives to the transitional assembly that ran the country until the 2006 elections created widespread infighting among Mai Mai forces in the Kivus and northern Katanga.

        These tensions could have been managed peacefully, but the 1998-2003 war destroyed the existing institutional means to do so. Congo's justice system has collapsed, like much of the state at large. The war dislocated many communities, disrupting the operation of traditional conflict-resolution mechanisms. The government's all-around poor performance, especially its failure to reestablish the rule of law in the eastern provinces, has perpetuated a culture of impunity, which has facilitated the use of violence, and the widespread availability of small arms has made force an easily accessible option for almost anybody. The national security forces cannot be relied on to maintain stability, because the utter lack of economic development in the eastern provinces means that belonging to an armed group is one of few profitable occupations.

        Today, most of the Mai Mai in Nord-Kivu remain allied to Rwandan Hutu militias, support President Joseph Kabila (the son of and immediate successor to Laurent Kabila, who died in 2001), and continue to oppose the armed Tutsi groups--all because doing so is still the best way for them to consolidate their claims to ancestral land rights and positions of authority. The Tutsis, for their part, have recently rallied around Nkunda, who belonged to the Rwanda-backed rebel movement that fought the Congolese government during the last war. He refuses to disarm and integrate his troops into the national army in order to better protect his ethnic community, which he believes is once again threatened by various local and national Congolese groups. In keeping with Congo's history since independence, the dispute between the Mai Mai and the Tutsis has a regional dimension, too: Nkunda is said to be recruiting fighters and obtaining arms from Rwanda.

        Thus, for much of the 1990s and early years of this century, local tensions in the Kivus have repeatedly prompted outbreaks of ethnic violence, with so-called indigenous groups forming alliances with Rwandan Hutu militias and, in response, the Rwandan government supporting Congolese fighters of Rwandan ancestry and intervening in the name of national security. And the situation, which shows that local troubles in eastern Congo jeopardize the entire country's stability, is consistent with recent academic research about civil wars. The Yale political scientist Stathis Kalyvas, among other scholars, has shown how in many conflict environments, land disputes, social antagonisms, professional jealousies, family feuds, and romantic rivalries become the fodder for tensions at the regional and national levels. Local leaders learn to couch their feuds in the rhetoric that dominates the national discourse--be it about ideology, ethnicity, religion, or class--in order to enlist support from government actors. Conversely, national politicians use local players to find the recruits, resources, and information they need to pursue their own objectives. Local violence may be fueled by regional and national antagonisms, but it is above all motivated by distinctively local tensions.

SEEING THE TREES FOR THE FOREST

        DISTINCTIVELY LOCAL agendas motivate a large part of the ongoing violence in Congo, yet diplomats, UN officials, and journalists have focused almost exclusively on the regional and national problems. To ease economic and security tensions between Congolese and Rwandan actors, for example, diplomats and UN officials have organized numerous dialogues and conferences in the region and elsewhere, including some with the Congolese, Rwandan, Burundian, and Ugandan governments to discuss their support for various rebel groups, the repatriation of Congolese refugees, and developing a code for the exploitation of Congo's natural resources. In times of crises, the UN leadership and African and Western states, such as South Africa, the United States, and European Union countries, have put pressure on the Rwandan government, in some cases by threatening to withdraw international aid, in order to prevent it from invading Congo again. After the 2003 peace agreement, former warlords were continuing to fight one another politically and militarily, while Congolese military leaders at all levels were diverting funds destined for the national army. African and Western diplomats from the 15 states and organizations involved in Congo's postconflict transition endeavored to convince the warlords to integrate their soldiers into the army, supervised the disbursement of soldiers' pay to prevent the diversion of funds, and trained a few integrated brigades.

        But this effort overlooked the critical fact that today local conflicts are driving the broader conflicts, not the other way around--and with counterproductive effects. Most notably, the international community's insistence on organizing elections in 2006 has ended up jeopardizing the peace. There was no outbreak of violence on the day of the polls; many Congolese were enthusiastic about voting for the first time in their lives. But the elections cemented Kabila's strongman government, which is bent on harassing the opposition and carrying on Mobutu's legacy of corruption--two destabilizing factors. The election process itself was also damaging. After the calm that immediately followed the voting, many provinces experienced renewed tensions along ethnic lines because of candidates who had propagated hatred during their campaigns in order to boost their popularity. The campaign was marred by major intimidation and fraud, which significantly tipped the balance of power at the provincial level. In Bas-Congo and Kasai-Oriental, the contest further marginalized minorities. The Tutsis of Nord-Kivu could not get any representatives into the provincial assembly because some 40,000 of them are refugees in Rwanda and cannot vote. The National Assembly, moreover, now counts many radicals bent on cleansing Congo of people of Rwandan descent. The Tutsi minority's renewed fears that an ethnic-cleansing campaign may be in the offing was a major reason for Nkunda's popularity late last year and, indirectly, for the renewed fighting in Nord-Kivu.

        Instead of focusing solely on large-scale peacekeeping and elections, the international community should have also taken on other critical postconflict tasks, such as institution building. But since 2003, diplomats and UN officials have left it up to Congolese authorities, Congolese religious leaders, and nongovernmental organizations (NGOS) to conduct bottom-up peace-building work. And with only a few exceptions, Congolese authorities and religious leaders have been unable or unwilling to conduct peace building locally--when they have not been involved in fueling the violence outright. A handful of NGOS, Congolese and international, have implemented local conflict-resolution projects, but their numbers have been too few, and they have faced too many challenges to make much of a difference.

        The UN-led peace process also did almost nothing to promote good governance or reinforce Kinshasa's administrative hold on the eastern provinces. This was a major flaw, because the reestablishment of the rule of law could have deterred some human rights abuses, assuaged resentment over past communal violence, and brought to all Congolese a level of personal and material security that might have lessened their dependence on armed groups. Instead, rivalries were left to fester.

        The result, besides a return to major violence, has been the worsening of the underlying problems. The conflicts have become increasingly decentralized, and the parties have fragmented--meaning that the basic issues have become even more localized than before. Journalists and policymakers often talk of the Rwandan Hutu militias, Tutsi dissidents, and the Mai Mai as if these were coherent groups, but none has a unified command structure. In the past several years, the Rwandan Hutu militias have increasingly fractured; now, factions fight one another over the spoils of looting, leadership antagonisms, and whether to return to Rwanda. Subgroups among the Tutsis in the Kivus have distinct and sometimes inconsistent agendas. Although the Tutsis in Nord-Kivu are currently aligned with Nkunda, relations between them can be tense. In Sud-Kivu, the Tutsis are divided between rich and poor clans, with the rich reportedly supporting the local dissidents sometimes called the Group of 47 and the poor backing the Kabila government. Meanwhile, there is no hierarchy controlling the Mai Mai, not nationally and sometimes not even within a single city. Some Mai Mai groups are allied with government troops (especially in Nord-Kivu), but others are fighting against them and among themselves (especially in Sud-Kivu and northern Katanga). The factions are so subdivided that many brigade commanders do not control their own battalion commanders. Even the national army cannot rein in its soldiers; both officers and members of the rank and file regularly loot, rape, and commit other human rights violations or strike deals with the militias they are ostensibly fighting in order to gain access to resources.

THINKING LOCAL, ACTING LOCAL

        GIVEN THE recent clashes, it is clear that more work is urgently needed to deal with the violence at the regional and national levels. Western and African governments must intensify diplomatic pressure on the Kabila government and on Nkunda in order to stop the fighting immediately. Disbursing more humanitarian and development aid would help prevent many deaths by providing much-needed medical and nutritional assistance, which the Congolese health system cannot do. The UN Security Council should request that the UN use its peacekeeping troops to protect those populations in immediate danger rather than focusing on protecting UN buildings and equipment. And the U.S. government must drastically change its Rwanda policy, threatening to sanction Kigali unless it prevents cross-border activities in support of Nkunda.

        But far more important, international actors must radically rethink their peace-building strategy if they want to accomplish more than yet another temporary cease-fire. Since 2003, most diplomats and UN staff members have been held back from getting involved at the local level by four widespread assumptions: they have treated Congo as a post-conflict situation, they have assumed that violence is pervasive throughout the country, they have relegated intervention to the national and international realms, and they have acted as though holding elections is an effective tool of peace and state building. In fact, Congo today is in the midst of a civil war, violence is not a normal feature of life there, local peace building is a legitimate task for international actors, and elections do little to stabilize countries or build institutions, and they sometimes hurt. Treating only the consequences of the ongoing conflict without addressing its underlying causes is absurd; the situation in Congo must be approached from the bottom up.

        The very first priority must be resolving land disputes in eastern Congo. For starters, the Congolese government must enact new land legislation that upholds the rights of vulnerable people (such as women, minorities, and returnees) and clarifies exactly when and how legal or traditional ownership rights apply. Throughout the country, but especially in the eastern provinces, the new legislation must mandate a review of all land property deeds. Local NGOS and judicial employees must be sent to rural areas to explain property law to the population there, which generally knows little of its rights. The new law must also include a special provision for resource-rich lands. Mining contracts for Katanga and Kasai-Oriental, among other places, are currently being reviewed; the process must be extended to all of Congo, especially to the Kivus and Ituri, where control over resources is an especially volatile issue.

        Land reform must also establish formal mechanisms for resolving disputes through the local courts, to be staffed with both judicial employees and representatives of the affected communities, or through ad hoc arrangements. Whenever necessary to ensure fairness or prevent creating new resentments, people whose property is being taken away should be compensated with money or in kind. For example, the beneficiaries of redistribution could be required to help the former owners build another house or to share their harvest with them. All adjudications should be handled free of charge so that the most disenfranchised people have a chance to claim what is theirs.

        In areas where many families, clans, or ethnic groups are deprived of the land they need to survive (such as in Masisi, in Nord-Kivu, or Kabare, in Sud-Kivu), the new legislation must also create provincial commissions to design a fair redistribution policy. These should include representatives from every local community and social group, Congolese experts on land issues, and neutral observers. They should focus on redressing injustices and on finding sustainable solutions. As the International Crisis Group suggests, for example, in the territories of Masisi and Walikale, in Nord-Kivu, such a commission should cancel all the title deeds for estates and ranches issued since Congo's independence. It should also compensate the former owners of expropriated land and assign some of it to landless families (notably among the Hundes, the Hutus, the Nyangas, and the Tutsis, who are the main groups living in Masisi and Walikale) for individual or collective use based on whether it is fit for agriculture or animal grazing. Broad land reforms such as these would prevent new disputes, improve intercommunal relations, and help extend state authority to the mining sites in the region. It would also go some way toward ensuring that the return of Tutsi refugees to the Kivus does not trigger another major crisis.

        It is important that these efforts target all the communities in the Kivus, not just the population with Rwandan ancestry and its traditional enemies. Even more broadly, it is also important that all local actors have a chance to air and resolve their grievances, be they about land, sharing traditional and administrative power, or anything else. To ensure a lasting peace, NGOS should help recreate social links between communities in conflict. The most effective strategy is to create enterprises, health centers, markets, and schools in whose success all the parties have a stake. A similar approach has worked in parts of Bosnia and Herzegovina, Cambodia, and Tajikistan. Combined with land reform, such a broad reconciliation program could help stem violence, address most of the grievances that gave rise to the Mai Mai, shrink the pool of local recruits for regional and national warlords, reintegrate refugees and displaced persons, and start rebuilding state institutions.

THE SUPPORT GROUP

        IDEALLY, THE Congolese would lead these initiatives. But the government in Kinshasa is weak and corrupt, and Congolese NGOS and civil-society representatives often lack the funding, logistical means, and technical capacity to implement effective peace-building programs. International actors can help, but only if they make resolving local conflicts a top priority instead of concentrating only on humanitarian programs or macro issues such as elections (as most groups currently based in the eastern provinces are doing). Diplomats and UN staffers have little experience developing and implementing comprehensive programs addressing local violence. They should urgently build up their capacity by hiring experts on Congo and Rwanda and local conflict resolution, sharing those specialists' knowledge with all existing staff, and creating specialized offices or departments in these areas.

        Since last year, international actors have taken tiny steps in the right direction. The United States and the United Kingdom have opened consulates in Goma, the capital of Nord-Kivu. The UN peacekeeping mission in Congo (known by its French acronym MONUC) redeployed troops to the eastern provinces, mostly to Nord-Kivu, and is setting up buffer zones to separate the main combatant groups. The few existing NGOS that focused on local conflict resolution in the region are more active than ever. The NGO Initiative pour un Leadership Cohésif en RDC, for example, has organized several workshops with local and national elites in order to help them work out their differences, and the Life and Peace Institute has intensified the funding, as well as the teaching and logistical support, it gives to those Congolese NGOS that do the best work promoting conflict resolution in the Kivus.

        But this is not enough. Furthermore, even well-intentioned initiatives are often ill conceived. In January 2008, for example, the Congolese government, with strong diplomatic and UN support, organized a peace conference in Goma to find a solution to the specific problems of the Kivus. Participants did have a chance to discuss their grievances over local political power, land expropriation, and mining resources, but these topics were not a priority. The conference focused instead on neutralizing the most prominent warlords, such as Nkunda and the major Mai Mai chiefs. A cease-fire agreement was signed. But the gathering's main accomplishment, a nonbinding "act of engagement," proposed no concrete solutions for local antagonisms. And the fighting never stopped, not even during the conference.

        Donors would do better to expand the funding available for local conflict resolution by increasing their aid budgets or shifting their assistance priorities away from elections. They should focus on helping the Congolese government and representatives from all the eastern communities work on land reform and the review of mining contracts by providing independent experts on land and judicial matters. Donors should also fund the training of local Congolese NGOS and justice officials so that they can be deployed as observers to the land-redistribution commissions or sent to villages to educate the rural population. And they should provide the NGOS with the funds to compensate the parties who will lose land. To ensure that any additional money goes to efficient programs, donors should ask the experts on local conflict resolution and the specialists on Congo and Rwanda in their consulates to identify reliable local peace builders in the eastern provinces. They should offer financial support to the Congolese NGOS that organize peace talks and reconciliation programs, such as Plate-forme des Associations de Développement de Bunyakiri, which brings together military, political, business, and ethnic elites of the territory of Bunyakiri, in Sud-Kivu, and Arche d'Alliance, which helps victims of human rights violations in Sud-Kivu and promotes the reform of existing human rights legislation.

        MONUC has an important supporting role to play. Although some of its troops have been involved in resource trafficking, sexual violence, and some brutal joint operations with Congolese army personnel, the force's presence has had a positive impact overall. If nothing else, it has so far prevented the conflict in Nord-Kivu from escalating into a regional or national war. Going forward, MONUC should start working on resolving local conflicts and distributing its resources differently than it does now. (New directives from the UN's Department of Peacekeeping Operations and MONUC's leadership would allow for this, but a Security Council resolution emphasizing the dangers of local tensions and MONUC's responsibility in local peace building is preferable, as it would help overcome any resistance by UN staffers on the ground.) In the eastern provinces, MONUC should deploy more military police and special operations forces and fewer traditional troops, because the former are better trained for action at the local level, especially in logistically difficult environments. In their daily work, military and civilian UN staffers should help provincial authorities develop the capacity to oversee the exploitation of mining sites. In addition, MONUC should recruit well-trained local peace building officials for deployment in the eastern provinces, downsizing its staff in Kinshasa if necessary. MONUC should also send civilian staffers with the authority to draw on military, diplomatic, or development resources to monitor local tensions and suggest how best to broker peace. The existing Congolese NGOS are ill equipped to address the local tensions caused by military antagonisms or manipulated by regional and national actors, and so international donors and UN agencies should step in to assist them. Such interventions would help address the broader dimensions of the violence by both deterring local warlords and offering them the possibility of development assistance.

        In the long term, local peace will be sustainable only if the Congolese state is stable and its institutions are built up at all levels. To that end, the Congolese government must develop ways to integrate all the armed groups, including Nkunda's troops and the Mai Mai, into the national army; rebuild its justice system (an essential step toward ending impunity and thus deterring violence, assuaging communal resentment, and promoting good governance); and solve the security problem posed by the Rwandan Hutu militias (by resettling those Rwandan Hutus who are not guilty of war crimes and launching a campaign with MONUC to capture any perpetrators of atrocities on the Congolese population and the few Rwandans guilty of genocide still present in the Kivus). These would be extremely difficult tasks anywhere, and Congo, with its weak state, fragmented political arena, refugee flows, and poor infrastructure, is a particularly challenging environment. But with over a thousand people still dying there every day and the Kivus in the midst of a conflict that could easily engulf the Great Lakes region again, something must be done. The best approach is to make a priority of treating core problems at the local level, especially long-standing land disputes, rather than focusing exclusively on managing their broader consequences. When it comes to Congo, international actors should work, quite literally, from the ground up.

MAP: DEMOCRATIC REPUBLIC OF THE CONGO

PHOTO (BLACK & WHITE): The wretched of the earth: refugees at a camp in Goma, Congo, 1994