top of page

The Army in Crisis

Executive Summary

Two years in, the Burundi crisis shows little sign of resolution. Following the July 2015 re-election of President Nkurunziza, whose April decision to run again sparked the troubles, and with no progress made in the mediation, the crisis has turned into a low intensity conflict. Almost 400,000 Burundians have fled the country. Since the attempted coup of May 2015, political polarisation has had violent repercussions in the army. A series of attacks have targeted numerous officers, both those favourable to the president’s political ambitions and those suspected of sympathy with the coup plotters. Assassination attempts have also taken place abroad. Following over ten years of foreign support for the army’s transformation, its reputation has suffered greatly. International training has ended, and the army’s lucrative participation in peacekeeping operations is in doubt. This divided and demoralised army is a major threat to the country’s stability. Only a real dialogue, more urgent now than ever, between the government and the opposition could offer assurances to those officers concerned at the politicisation of their institution.

Long seen as the primary achievement of the Arusha peace agreement which ended the civil war in 2000, the army today is a microcosm of the country’s crisis. Through its multi-ethnic makeup, foreign training, and its role in international peacekeeping, the Burundian army had acquired a good reputation outside the country and a privileged position at home. But fragilities remained under the surface, and the 2015 crisis easily broke the key consensus on which the stability of the regime was based: between the army and civilian power, and within the army between the former rebels, most of whom come from the ruling party, and the old guard. Ever since, the regime has tried to regain its hold on the military through purging or killing real or suspected opponents within its ranks – starting with officers from the pre-war army and Tutsi officers, but also targeting former Hutu rebels, including high ranking officers.

The current crisis, in the form of tit-for-tat assassinations of soldiers and officers, is a violent reminder of the limits of the Arusha agreement within the army, and of the efforts made over ten years to depoliticise and professionalise it. It also reveals political and ethnic tensions that have continued to undermine it despite the reforms. The crisis has led to numerous defections and has compromised its future prospects. The European Union and the UN are reluctant to increase Burundi’s participation in peacekeeping missions and have taken steps to limit it. This participation used to be a source of revenue for an otherwise impoverished army, and a way of integrating its different parts. The current challenge to it and to associated external support could eventually weaken the economic and social advantages associated with the military career, and is a further risk for the stability of the country.

Impoverished and ethnically and politically polarised, the army is reforming around a loyalist hard core and open confrontations between army factions have been avoided since May 2015. But this apparent and only relative calm is based mainly on fear and should not mislead outside observers. The army that has been built since 2004 is now in ruins, and cannot be reconstituted short of an inclusive political agreement. This appears ever further off with the continued hardening of the regime and consequent difficulties encountered by the mediation of former Tanzanian President Benjamin Mkapa. Without such a political agreement, the army faces two scenarios: a new confrontation, which could take the form of a new coup d’Etat, or a quiet but certain decline.

The relative success of army integration since 2004 has flowed from the Arusha Agreement. In this context, only guarantees concerning its continued application, or its consensual updating, could reassure officers that their future and that of their institution is secure. The UN, the African Union, the East African Community and the European Union should continue to push for an inclusive dialogue between the government and the exiled opposition, despite the government’s intransigence, which has hindered mediation attempts, and international partners who have supported the army since 2004 should not reinvest in an institution now deeply politicised as long as it remains under the control of an authoritarian and violent regime. The involvement of the Burundian army in peacekeeping operations should continue only under strict vetting conditions of the individuals taking part. The crisis in the army, reflecting that of the country, underlines the continued risk that the situation could deteriorate further.

Nairobi/Brussels, 5 April 201


The Burundian crisis, which erupted in April 2015 over disagreement about the legitimacy of President Pierre Nkurunziza’s candidacy for a third term, continues. Since his re-election in July 2015, the government and its opponents have been involved in a low-intensity armed struggle. While demonstrators protested against a third term (April-July 2015), the army made sure it stayed out of the political crisis, observing developments but not taking part in the repression. Unlike the police, the army avoided the use of force. Some soldiers even stepped in to prevent confrontation between demonstrators and police officers, which sometimes led to violence between police or intelligence service officers, and soldiers.

However, an attempted coup on 13 May 2015 highlighted dissent within the army. The Arusha Agreement of 2000, which enshrined the principle of ethnic parity in the security forces, and later agreements between the National Council for the Defence of Democracy/Forces for the Defence of Democracy (Conseil national pour la défense de la démocratie-Forces de défense de la démocratie, CNDD-FDD) and the National Liberation Forces (Forces nationales de libération, FNL), a rebel Hutu group dating from the civil war and the FDD’s great rival, provided for the integration of the rebels into the army, with the support of guarantors of the Arusha Agreement, including the UN and South Africa. In 2004, the rebel groups and an army mainly composed of and led by Tutsis merged to form the National Defence Force (Force de défense nationale, FDN). The former Burundian Armed Forces (ex-FAB) form the old guard of the army, mainly Tutsi, while the former Armed Political Parties and Movements (ex-Partis et mouvements politiques armés, ex-PMPA) are former combatants of mainly Hutu armed groups, including the FDD, which is now in power, and which were integrated into the army after the peace agreements.

The attempted overthrow of the government, when President Nkurunziza was in Tanzania for an East African Community (EAC) summit, was led by Godefroid Niyombare, former armed forces chief of staff and a very popular and historic figure in the governing party, and Cyrille Ndayirukiye, former defence minister and member of the former Burundian Armed Forces. It revealed opposition to a third term among some officers and dragged the army right into the middle of the political maelstrom. Ndayirukiye and three other generals were sentenced to life imprisonment and Niyombare went into exile. Political violence then erupted in the army: the government moved against suspects in an attempt to eradicate every pocket of resistance.

The army and its dissident factions are far from being the only perpetrators of the violence that has shaken Burundi since 2015. While the army has been reformed in line with the Arusha Agreement, the agreed quotas were not implemented in the police force, where many officers today are ex-PMPA, and were not applied strictly to the National Intelligence Service (Service national de renseignement, SNR). The government has therefore been able to place former civil war allies in these two institutions. They are now largely loyal to the government and their leadership is very politicised. The SNR, very close to the PMPA, and formerly led by a radical member of the government, Adolphe Nshimirimana, has long been the most feared institution in Burundi.

The FNL was the last armed group to join the peace talks and the government only started to integrate its combatants into the FDN in 2009. After their political party boycotted the 2010 elections, they suffered fierce repression from the government and some of them who had joined the FDN fled. Some went back to the bush led by Aloys Nzabempema in South Kivu in the Democratic Republic of Congo (DRC). The crisis has accelerated these desertions and intensified the repression of FNL militants. However, their historic leader, Agathon Rwasa, remains in Burundi and has sat in parliament since the 2015 elections.

II.From a Political to a Military Crisis

A.Purges and Reprisals

The 2015 campaign against the third term continued into 2016 within the military. A series of tit-for-tat assassinations has created a climate of paranoia and major tensions within the army.

Since August 2015, killings of soldiers have continued. The victims’ identity, often officers, indicates that these are mainly targeted killings in reprisal for either support of or opposition to a third term. The assassination of General Adolphe Nshimirimana, former head of the National Intelligence Service, on 2 August 2015 was followed on 15 August by the assassination of Colonel Jean Bikomagu, from the former Burundian Armed Forces (ex-FAB), armed forces chief of staff at the time of the coup against the Hutu President Ndadaye in 1993 and a symbol of the Tutsi military old guard and the rejection of a Hutu government.

On 22 March 2016, Lieutenant Colonel Darius Ikurakure, from the former Armed Political Parties and Movements and in charge of repression in the northern neighbourhoods of Bujumbura, and, a few hours later, Major Didier Muhimpundu (ex-FAB), were killed. The armed forces chief of staff, ex-FDD, General Prime Niyongabo, escaped an assassination attempt in September 2015, while General Athanase Kararuza (ex-FAB and military adviser to the first vice president) was killed in an ambush in front of the Saint-Esprit College in Bujumbura on 25 April 2016.

Although the ex-FAB (on active service and retired) were the first to come under suspicion from the regime, its violence has not spared the former Armed Political Parties and Movements opposed to the third term, such as Colonel Emmanuel Buzubona, former number two of military intelligence, assassinated on 20 April 2016. When armed groups do not openly claim responsibility, such as the Republican Forces of Burundi (Forces républicaines du Burundi, FOREBU) for the killing of Darius Ikurakure, the modus operandi generally bears the hallmarks of the military (weaponry, knowledge of the routines followed by victims, access to places, etc.). In addition to these tit-for-tat killings, soldiers are responsible for other acts of violence against each other, which the government tries to minimise, being unable to conceal them (one killed on 28 March 2016 at Muzinda camp, several members of the presidential guard killed on 19 June in Bujumbura).

Army personnel, both on active service and retired, arrested by the National Intelligence Service (SNR) sometimes disappear. It is impossible to make an exhaustive list of these arrests but some of them attest that the ex-FAB is being targeted. In 2016, some ex-FAB retired personnel were arrested or killed, soldiers arrested in September were found guilty of endangering state security, and an adjutant died on SNR premises on 14 September. After having targeted officers, repression now seems to be focusing on intermediate army levels (non-commissioned officers) as well as on retired personnel. The climate of fear is such that ex-FAB personnel serving in Somalia dread returning to their country when on leave – several of them have been arrested on their return.

The authorities deny the seriousness of the crisis. They denounce an external attempt to destabilise the country and claim “there is no unrest in the army”. However, violence carried out with a military modus operandi continued to the end of 2016 and into 2017: attempted assassination of Willy Nyamitwe on 28 November; clashes leaving several victims in the ranks of the National Defence Force in South Kivu, DRC, on 21 December; assassination of the environment minister on 1 January; attack on Mukoni military base in Muyinga province on 23 January, followed by another wave of arrests of military personnel; and a clash between a faction of the National Liberation Forces (FNL) and the army in Gatumba, in the province of Bujumbura Rural, on 7 February.

B.Desertions and Attempts to Organise Abroad

The fear of being killed has increased the number of desertions from the Burundian army since 2015, as confirmed by the UN. According to Burundian military sources, between 600 and 2,000 soldiers have deserted since the crisis began, including senior officers. These include Lieutenant Colonel Alexandre Mbazumutima, intelligence officer with the 120th brigade, Major Emmanuel Ndayikeza, second-in-command of the support battalion in the first military region, an elite unit based in Bujumbura, and Colonel Edouard Nshimirimana, communications officer, who reportedly deserted, the latter with about 40 soldiers and carrying arms, ammunition and communications equipment. Several desertions took place in the summer of 2016 in Ethiopia, Belgium and from the Higher Institute of Army Officers.

Meanwhile, several armed opposition groups have appeared. The Resistance for the Rule of Law in Burundi (Résistance pour un Etat de droit au Burundi, RED-Tabara) and the Republican Forces of Burundi (FOREBU) were formed at the end of 2015/beginning of 2016. A group of National Liberation Forces combatants commanded by Aloys Nzabampema, opposed to the historic leader Agathon Rwasa, has been active on the Congo-Burundi border for several years.

The latter group scarcely communicates but the other two groups have expressed their willingness to resort to arms against the government without, however, opposing attempts at mediation.All three probably have contacts inside the country. The RED-Tabara is the armed wing of the Movement for Solidarity and Democracy (MSD) led by Alexis Sinduhidje. FOREBU has been led by Colonel Nshimirimana since the coup leader General Godefroid Niyombare took on a secondary role, at least in terms of media presence. This is the only group to be mainly composed of soldiers who served in the National Defence Force. Its hard core is formed by soldiers involved in the coup of 13 May 2015.

As the leaders of RED-Tabara and FOREBU are in exile, like most opponents, the regime has tried to target them abroad. It uses members of the SNR and Imbonerakure to infiltrate refugee camps and opposition circles. Attempts to kill opponents (not necessarily linked to armed groups) have already occurred in Nairobi (Kenya) and Kampala (Uganda).

III.An End to the Pretence: Undermining Arusha

A.Integration without Cohesion: Army Reform

Formed in 2004, the National Defence Force is the product of a politico-military agreement. After three decades of mono-ethnic military dictatorship, the signatories of the Arusha Agreement accepted the principle of ethnic parity in the security forces (50 per cent Hutu and 50 per cent Tutsi) and included it in the third protocol. The duration of this ethnic balance remained undecided and has still not been set. In 2003, as part of peace negotiations with the National Council for the Defence of Democracy/Forces for the Defence of Democracy (CNDD-FDD), this agreement on ethnic composition was completed by an agreement on the political composition of the security forces. The Technical Forces Agreement stated that the CNDD-FDD should occupy 40 per cent of army command posts.

To achieve this dual ethnic and political objective, the National Defence Force went through a phase of demobilisation and restructuring. From 2004 to 2008, 41,000 former Burundian armed forces (ex-FAB) members and 15,500 former armed political parties and movements (ex-PMPA) members were demobilised. With the prospect of a 25,00o-strong army, half of which were to be drawn from the Hutu ethnic group, Tutsi soldiers were the most affected. Supported by foreign partners, the operation to demobilise ex-FAB and integrate ex-rebels took place smoothly even though it was considered to be the major challenge facing the new government and the main risk of destabilising the transition. The political pact on power sharing between yesterday’s enemies played a major role in the creation of a new army over a four-year period, considered to be a success by both the Burundian military and foreign partners.

However, the process was not without problems. Demobilised combatants kept their arms, which contributed to the rise in crime. As the demobilisation payment for young ex-FAB officers was €300 and three months’ wages, reintegration in civilian life was difficult for many of them.

Evaluations of the army reform generally confuse integration and cohesion. Integration, which at the start of the 2000s took the form of disarmament, and later of an integrated armed forces staff and international deployments, allowed former enemies to get to know each other. This could, moreover, explain the alliances between former armed political parties and movements and former Burundian armed forces behind the attempted coup of May 2015.

However, the lack of social contact between ex-FAB and ex-FDD indicated the lack of cohesion. For example, the officers’ mess in Bujumbura was not used by ex-FDD officers, who had their own places to socialise, notably General Nishimirimana’s bars. Attempts at fraternisation between the two groups of officers were not viewed kindly in some quarters. Prejudices remained and interaction was limited to daily tasks. Some used pejorative terms to describe each other: the ex-FAB described the PMPA as “bushmen”, while the latter described the ex-FAB as mujeris (little dogs).

Moreover, the ex-FAB silently resented the glass ceiling imposed by the political-military readjustment of the officer corps and the ultra-quick promotion of ex-PMPA officers who had neither their military experience nor level of education. Members of the National Liberation Forces who joined the army in small numbers after the 2008 agreement do not seem to have found their place in the institution. Some have even felt discouraged enough to leave it.

In addition, unification of command was more theoretical than real. Conscious of being incorporated into an institution they did not control, unlike the police force and the intelligence service, the ex-FDD created a parallel hierarchy. The official military chain of command was short-circuited by their own network, which led up to Adolphe Nshimirimana and the presidency. This posed problems in terms of discipline, promotion and personnel management. Some soldiers that formed part of the parallel system treated the official hierarchy and discipline with disdain, knowing they were covered in high places and confident of their promotion. This situation led to a lack of transparency in management of grades and even to denials of access to training. In addition, it subjected the army to a hierarchy that was unofficial, partisan but known to all.

This parallel command structure further politicised officers and harmed team spirit. Before he was killed, General Adolphe Nshimirimana occupied a strategic position at the crossroads of the presidency, the security sector, trafficking networks, the Imbonerakure and veterans of the CNDD/FDD gathered in the Nonoka Association. He was therefore a mainstay of the parallel command system that is at the heart of the government’s current repressive apparatus and that affects all the security forces: the army, the police force and the intelligence service.Repression is currently carried out by special units within the security services, whose chains of command are short and parallel to the official hierarchy.

Although integration allowed the formation of a new army, the latter has always suffered from a lack of cohesion and rampant politicisation. The success of integration was wrongly interpreted as a guarantee of reconciliation, unification and fraternisation. Peaceful coexistence of yesterday’s enemies was not synonymous with unification and cohesion of the institution, notably at the command level.

B.Clientelism as a Mode of Governance

Since 2005, the dual attitude displayed by the government, anxious to continue the integration and reform policies that allowed the army to participate in peacekeeping missions, while also promoting its former colleagues from the guerrilla war and encouraging criminal behaviour, has strongly disrupted the internal functioning of the army.

In addition to resorting to traditional clientelism toward the officer corps (promotions, postings abroad, etc.), the government bribes them and plays on their regional divisions. Authorities take care to respect ethnic balance in the military high command as well as within the government, while at the same time ensuring the loyalty of Tutsi appointees, who are considered to be “bit players”. By integrating part of the command into its clientelist network, the presidency has bypassed the power-sharing framework in the army. Several former Burundian armed forces (ex-FAB) officers, driven by regionalism, opportunism or greed, play the government’s game. This politicisation of part of the high command goes against the principle of political neutrality set out in the National Defence Force’s founding document.

The failed coup in May 2015 highlighted cases of duplicity and treason within the high command. The pressure of the intelligence service, the latent divisions among soldiers and the use of some as informers led to collective paranoia in military circles.

For example, although all Burundians in the African Union Mission in Somalia (AMISOM) follow the news in their country on social media, self-censorship is required and morale is at a low ebb. The presence of informers in the battalions, confirmed by Burundian blue helmets, provokes fear and suspicion. This results in rumours about senior officers, which confirm or raise questions about their loyalty to the government.

The clientelist relations between the high command and the government allows the latter to maintain the appearance of Arusha in the army. The ethnic balance is observed in the AMISOM command and the army general staff. After a parliamentary study on the composition of the army in 2008, the parliament again examined the subject and visited military units from November 2015 to February 2016. There has been no official communication about its conclusions.

But power sharing in the army is openly questioned at several levels. The prevailing tacit consensus at the summit of the military hierarchy regarding sharing the command between a minister and chief of staff from different camps was broken by the appointment of the current defence minister after the failed coup in 2015. Although he is a Tutsi, he is from the CNDD-FDD rather than the ranks of the ex-FAB and is a civilian (judge). The government has abandoned the principle of balanced appointments (one ex-FAB and one ex-PMPA) at the highest level of the army.

Moreover, the demographic situation works against the ex-FAB, therefore raising questions about the durability of the agreement in the army. Older than the integrated rebels, they are not replaced by other Tutsis when they retire. The ex-FAB are very concerned about this process and have complained about the emergence of an army command that no longer corresponds to the Arusha Agreement and will not be able to correspond to it in the future. The climate of terror and ethnic polarisation that reigns in the security services means that finding Tutsi recruits might become a challenge. Meanwhile, the government decided to replace the 50/50 quota with a 40 (Tutsi) /60 (Hutu) quota for army recruits as from 2016.

This historic politicisation of the army (it has always been that way in Burundi), combined with the attempted coup of May 2015, which revealed the soldiers’ loyalties, left a hard core of officers loyal to President Nkurunziza in strategic positions. These old CNDD-FDD colleagues from the armed struggle are now fighting their former, now dissident, colleagues from the bush.

C.The Burundian Army’s International Policy in Jeopardy

The current crisis has gradually raised questions about the status and advantages acquired by the Burundian army since 2004. Subjected to a massive reduction in manpower at the end of the civil war, the National Defence Force has certainly been pampered by the government but its involvement in peacekeeping missions has been its real lifeline and has turned it into a privileged institution in this poor country. However, the government’s hard-line policies are potentially undermining this stabilising effect.

Making the army the showcase of the peace process allowed the government to develop a self-promoting discourse for its international partners, praise the success of Burundi’s approach to consolidating peace and reduce the UN’s presence in the country, perceived as too intrusive.By becoming the best example of the Arusha Agreement, the army improved its image and status and obtained major advantages as a result. But it was especially its participation in peacekeeping missions that brought new advantages: professionalisation, new financial resources for both the military and the government, posts in international agencies, etc. With its participation in AMISOM starting in 2007, the government discovered a strategy to reduce poverty within the army.

As with previous governments, those who came to power in 2005 were aware of the need to ensure soldiers’ welfare. Army members received a pay increase in 2006, social benefits (mortgages at preferential rates, installation of basic infrastructure at new properties, free health care, etc.) and generous promotions (inflation of the officer corps) all of which played a key role in the success of integration. However, the government’s lack of financial resources made it impossible to avoid social tension within the armed forces (for example, in 2009, on the issue of housing allocations).

From 2007, participation in AMISOM eased the shock of demobilisation/restructuring and reduced social discontent in a poor army. While a new wave of demobilisation was in the offing, the government became involved in Somalia and suspended staff cuts to the armed forces. The Burundian army currently comprises about 25,000 men, including 5,000 (a fifth) deployed in Somalia. This involvement has allowed a substantial increase in wages ($800 per month in Somalia compared to $40 in Burundi). The duration of the mission means that nearly all members of the Burundian army have completed a tour of duty in Somalia (some are even in Somalia for a second time). Moreover, the death in service benefit is $50,000 to the nominated beneficiary. In Gitega, Burundi’s second largest city, AMISOM veterans have built a new residential district.

In addition to wage increases for the armed forces, participation in peacekeeping missions opened up new career prospects at the international level. Other benefits have included military training and contributions to the government budget. The Burundian armed forces have received pre-deployment training, mainly, although not only, through the American program African Contingency Operations Training and Assistance (ACOTA); the government gets around $200 from the wages of each soldier involved in the mission to Somalia, in principle to pay for equipment and other military expenditure, a normal procedure for peacekeeping missions.

To maximise the return on investment, both political and financial, of participation in peacekeeping, the government offered its services for several missions (Mali, Central African Republic). The Burundian security forces have taken part in the African Union Mission in the Central African Republic (MISCA) and the UN mission that followed (MINUSCA). Participation in peacekeeping operations has become an official policy, as part of the National Defence Force’s missions.

The current crisis has led to a withdrawal of international assistance and to strong questioning of the Burundian army’s participation in peacekeeping missions. The emergence of a low-intensity conflict in Burundi contradicts the army’s role as a “peacekeeping force” in other conflicts.

From the beginning of the crisis, the main suppliers of military cooperation retired. American authorities suspended the ACOTA training program, which had already been reduced after a significant number of troops and officers benefited from it. Dutch authorities suspended most of their security sector development program, which was the largest bilateral military cooperation program from a financial point of view. While civil society organisations have launched a campaign on social media for the return of Burundian soldiers from peacekeeping missions (bringbackoursoldiers), the UN and the European Union (EU) are contesting Burundi’s participation.

The EU funds the wages of AMISOM soldiers, with the AU acting as intermediary. Since January 2016, the EU has paid $800 per month per soldier (previously $1,028), of which the governments of the countries contributing troops have decided the amount paid to their soldiers. The Burundian government paid its troops $800 before January 2016, retaining $200 to cover general costs. The participation of the Burundian army in the mission brought $52 million per year to the armed forces and $13 million to the government budget. In March 2016, the EU decided that Burundi’s failure to comply with the Cotonou agreement, which sets out the principles and modalities for part of European aid, forbad these payments and asked the AU to find a way of paying Burundian soldiers deployed under AMISOM directly without going through the government. After several months of bitter discussions, in a letter to the AU Commission on 8 December, Burundi threatened to withdraw its troops if a satisfactory means of payment could not be found.

In January 2017, it seemed that a compromise had been found according to which payments would be paid to soldiers through the intermediary of a commercial bank and not through the government. Burundi’s threats clearly worked, probably because of AMISOM’s importance to the AU and the EU. If the new arrangement is confirmed, it will bring some fresh air to both the Burundian army and government because the latter will still be in a position to deduct a proportion of the wages paid by the EU to AMISOM troops.

However, Burundi’s participation in peacekeeping missions is subjected to greater vigilance and thorough examination. Some Burundian personnel have seen their candidacies to posts with AMISOM and MINUSCA rejected and others, already deployed, have been repatriated. This policy was intensified when the entire contingent of Burundian police officers on duty with MINUSCA (280 men) was sent home.

IV.The Fate of the Army Dependent on the Future of the Country

Before 2015, the Burundian army appeared to symbolise the country’s reconstruction after the civil war and constituted an essential element in the Western powers’ plan to train and fund African armies for peacekeeping operations on the continent. Since April 2015, a largely predictable political crisis has led to desertions, exposed historic divisions within the army and provoked opposition to its participation in international peacekeeping missions. Training on peacekeeping has slightly professionalised an army that includes an old guard and former rebels and its neutrality during the riots in 2015 showed the willingness of some officers to remain outside the political battles. However, this policy of transforming the armed forces has been frustrated by a lack of cohesion, parallel command structures and the clientelism practised by a kleptocratic and violent regime.

The resolution of the crisis in the armed forces requires a political settlement to which loyalist and rebel officers should be associated. But so far, all attempts to mediate have failed and there is no dialogue between the opposition and the government, which is hostile to the very idea of negotiations. Therefore, despite several attempts at mediation by Benjamin Mkapa in 2016 and 2017, the government still refuses to meet and negotiate with members of the opposition for whom national arrest warrants have been issued. As the prospects of dialogue between the opposition and government diminish, given the persistence of the political crisis, there is no clear solution to the crisis in the army. Consequently, the question has to be posed as to what will be the long-term impact of this situation on the military institution: could the neutralisation of the government’s opponents in the army destabilise it by provoking a violent reaction or succeed and end in the “quiet decline” of the National Defence Force?

In the context of the failed coup of May 2015 and failed mediation, several scenarios are possible for Burundi: a new attempt at destabilisation leading to a relatively rapid change in government, or the disintegration of command structures and a civil war; or the stagnation and deterioration of the armed forces in the same way as the country as a whole.

If the idea spreads among the military that Pierre Nkurunziza’s continuance in power and the government’s hard-line attitude are prejudicial to their interests, some of them could be tempted into a new coup. The deterioration of their economic and social situation combined with the fear of physical elimination could push them into taking the plunge, as in May 2015. Some former Armed Political Parties and Movements officers share these concerns and feel that current policies compromise their future. They might also therefore participate in a new coup attempt.

The weakening of the political agreement in the security sector is leading to the re-emergence of divisions and resentments that had been put to one side but had certainly not disappeared. Consequently, a mutiny following on from further arrests of ex-Burundian armed force members, another murder of a senior army officer or a conflict with intelligence service agents or police officers cannot be excluded. The spontaneity of such a mutiny, which could only be partial because of the political and ethnic divisions within the army, would pave the way to many scenarios – from surprise victory to defeat due to a lack of a critical mass.

The policy of neutralising opponents in the army might also succeed. The combination of targeted eliminations and buying off some officers could reduce opponents to a minority that would have no other choice but to keep quiet or go into exile, like civil society and political opponents of the third term. The wave of defections that is underway could become stronger and the army could be completely purged of its anti-third term elements. The latter could join the networks of resistance abroad but a united and strong force on the borders of the country is currently not on the cards because of the divisions between opposition factions and the lack of external support.

The army would then cease to be a pocket of resistance to the power of President Nkurunziza and his circle. The National Defence Force (FDN) would not disintegrate but would enter a phase of decline because of the flight of its most experienced members, the suppression of foreign support, continued politicisation, the reduction of internationalisation and the government’s budgetary constraints. Poor and weak, it would no longer be a danger to the government, which would concentrate the remaining military resources in a few trusty units such as the Special Brigade for Protection of Institutions. The strategic objectives of the FDN and the plan to make it a “loyal, professional, prosperous, modern and republican army” would be no more than a memory.Such an outcome was known to be a risk when national defence policy was being decided and has occurred elsewhere under many authoritarian African governments.

Despite the government’s intransigence, political dialogue between the government and the opposition remains indispensable. The army’s divisions make these recommendations urgent, all the more so because the loss of external financial support threatens to accelerate its deterioration.

The multilateral organisations that are looking for a solution to the Burundian crisis, including the new president of the African Union Commission and Burundi’s donors must clearly tell the Burundian authorities that their intransigence, repression and violence are unacceptable. As far as possible, it is necessary to encourage the exiled armed groups to continue on the political path and work to unify the opposition.

The participation of Burundian forces in peacekeeping missions will probably continue, following the arrangement reached with the AU regarding AMISOM. This entails dangers that are already evident; the risk of including members of militias in the units deployed, and the risk that an army that is increasingly under the government’s control will be involved in the future in human rights abuses, the same way as the police and the intelligence service. All those who support and fund this participation must be vigilant and increase their vetting efforts. They must be ready to replace the Burundian army contingent in international missions if its behaviour continues to deteriorate.

All international training should henceforth be conditional on the government’s willingness to democratise and start a dialogue with the opposition in exile. The risk of weakening the army in the long term because of the withdrawal of external support is much less of a problem for the country than the reorganisation of the army in accordance with the government’s authoritarian plans while international partners close their eyes.


The crisis of the Burundian regime has quickly become a crisis of the Burundian army. This fact alone raises questions about the success of the “new Republican Army” project much promoted by the Burundian authorities and praised by their partners. It highlights the limits of training in changing the way institutions operate and in installing democratic governance in the security sector. While the Burundian army changed following the Arusha Agreement of 2000, pernicious links between the government and the army remained.


(c) 2017 International Crisis Group

Follow Genocide Watch for more updates:

  • Grey Facebook Icon
  • Grey Twitter Icon
  • Grey YouTube Icon
bottom of page