Burundi’s constitutional referendum: Consolidating the fait accompli in the run-up to the 2020 elections

By Dr. Stef Vandeginste, 23 January 2018
President Nkurunziza receives report of the CNDI (National Commission for Inter-Burundian Dialogue) in May 2017 (photo credit: Burundian Presidency)
President Nkurunziza receives report of the CNDI (National Commission for Inter-Burundian Dialogue) in May 2017 (photo credit: Burundian Presidency)

Burundi has embarked upon a constitutional amendment process which mirrors the country’s recent political evolution. If adopted, the revised constitution is likely to consolidate the political dominance of the presidency and of the ruling party. More than any other factor (opposition parties, civil society or international diplomatic pressure), intra-party dynamics are likely to determine the implications of the constitutional amendment for incumbent President Nkurunziza’s future after the 2020 elections. Although parliament has so far been sidelined in the ongoing process presumably leading to a popular referendum in May 2018, a legal technicality might bring the draft constitution before parliament prior to its promulgation – writes Dr. Vandeginste. 

On 12 December 2017, President Pierre Nkurunziza of Burundi officially launched the campaign in preparation of a constitutional referendum scheduled for May 2018. Although highly controversial for a variety of reasons, the referendum does not come as a surprise, neither in terms of timing and procedure, nor of the proposed substantive reforms.

Elected in 2005 by parliament and in 2010 through a popular vote, Nkurunziza had already announced a constitutional reform in his 2012 New Year’s address. Because of the strict qualified majority requirements for a constitutional revision in parliament, a first attempt at revising the current 2005 Constitution and allowing him to run for a third term failed in early 2014. A controversial Constitutional Court ruling of 4 May 2015 provided the highly contested legal basis for Nkurunziza’s third candidacy in the 2015 elections, amidst street protests, increasingly violent oppression, massive numbers of refugees leaving Burundi and, most importantly, a major internal rift within the ruling CNDD-FDD (National Council for the Defense of Democracy – Forces for the Defense of Democracy) party that culminated in a failed coup d’Etat attempt on 13 May 2015. Nkurunziza survived the crisis and won a third term in July 2015. During a carefully orchestrated national dialogue launched in October 2015 and completed in May 2017, the need for a constitutional reform was put high on the agenda by the authorities. Subsequently, the President established a technical commission in charge of drafting constitutional amendments, which submitted a first (unpublished) report in October and completed its work in November 2017.

The referendum clearly serves as a first step in his campaign towards the 2020 elections.

Rather than going through parliament again, where CNDD-FDD does not have the required qualified majority to unilaterally approve constitutional amendments, the President has initiated a constitutional reform through a referendum, for which no qualified majority requirement is laid down in the Constitution, a seemingly remarkable omission in what has rightly been referred to as the most consociational constitution on the African continent. For the President, the referendum clearly also serves as a first step in his campaign towards the 2020 elections. Also in December 2017, he extended the mandate of the Electoral Commission (CENI), the body in charge of organizing the referendum. Within days, CENI presented the new voters’ card that will be used both for the constitutional referendum and for the 2020 elections. Furthermore, the presidential decree on the CENI was amended, allowing provincial and local CENI branches to remain in office until the 2020 general elections.

Intra-party dynamics

Some three weeks before the launch of the referendum campaign, at the annual meeting in commemoration of the end of his CNDD-FDD rebel movement’s armed struggle in November 2003, Nkurunziza called for more internal cohesion and sent a strong message of zero tolerance to dissidents within CNDD-FDD. The two events were closely connected. Nkurunziza’s combined core message was that the political crisis, occasioned by his 2015 third term bid, had ended and should not reignite due to renewed intra-party dissidence. Furthermore, he urged the party and the country to support a far-reaching constitutional reform, thus allowing for the establishment, after the 2020 elections, of stronger political institutions ‘better capable of protecting the country against the threats of non-democratic forces that have destabilized Burundi for many decades’.

The draft constitution is surprisingly ambiguous on the eligibility of incumbent President Nkurunziza.

Presumably, ruling after 2020, which the current constitutional term limit does not permit, is his main motivation for the proposed constitutional reform, although, quite paradoxically, the draft remains surprisingly ambiguous on his eligibility, as explained below. After the Constitutional Court ruling of 4 May 2015 and in his inaugural address on 20 August 2015, Nkurunziza promised that his third term would be his final term in office, presumably in order to have regional and international actors accept the third term fait accompli and to enable those actors to urge the opposition to look ahead at 2020 rather than backwards. However, already in December 2016, while realizing that not all of his party cadres and inner circle supported his ambition of running for president in 2020, Nkurunziza stated he might decide to run again if the people wanted it.

As he has skillfully managed to do for the past 16 years since taking the leadership of the CNDD-FDD Hutu rebel movement in October 2001, Nkurunziza faces the ongoing challenge of rallying internal support by managing, and, if strategically useful, stirring, cross-cutting interest-driven lines of divide within the ruling party and its docile satellite ‘opposition’ parties, i.e. the dissident (but at the same time the only legally recognized) government-friendly wings of opposition parties. Indeed, some intra-CNDD-FDD fourth term opponents may favour a constitutional reform if that offers them the prospect of removing ethnic quotas, and, in particular, of ending the guaranteed representation of the Tutsi minority in political institutions, thus avenging decades of Hutu majority exclusion and oppression. Also, Nkurunziza’s repeated insistence on the need to revalorize clan identities rather than perpetuating colonially driven ethnic (Hutu-Tutsi-Twa) divisions may attract fourth term sceptics within his Bahanza clan and set them up against other Hutu fourth term sceptics from a different clan. Along regional fault lines, as well, the presidential ambitions of senior party officials and military generals have been carefully played out against each other, which on balance for all of them may mean that the incumbent continues to offer better prospects than a so far unknown alternative. Also, the proposed constitutional reform is likely to attract the support of leaders (including Tutsi) from CNDD-FDD satellite opposition parties to the extent that it reduces the chances of ‘real’ opposition parties or independent candidates successfully contesting the elections, which would indeed be the effect of part of the proposed amendments. Against this background, the constitutional reform – necessary for a term renewal in 2020 – is one of the strategic ‘divide and rule’ tools of the incumbent who simultaneously plays his political survival game on multiple chessboards.

The launch of the constitutional referendum is one major step in an incremental process of post-2015 re-consolidation of President Nkurunziza’s control over the ruling party.

In summary, the launch of the constitutional referendum is one major step in an incremental process of post-2015 re-consolidation of President Nkurunziza’s control over the CNDD-FDD ruling party and an attempted pre-emptive strike against any renewed palace revolution ahead of the 2020 elections. More than any other variable (international pressure, civil society and opposition-in-exile protests), his ability to curb intra-party dissidence in the run-up to the 2020 elections will determine his political future. The stakes of pursuing this fait accompli have heightened further after the Bureau of the Prosecutor of the International Criminal Court was authorized, in November 2017, to open an investigation into the human rights violations that occurred between 26 April 2015 and 26 October 2017, when Burundi’s withdrawal from the ICC Statute became effective.

Stalled international mediation

Parallel to the above-mentioned internal political dialogue, the East African Community – supported by the African Union and the European Union – launched an international mediation process to solve the political crisis in Burundi. Ironically, the mediation was led by Ugandan President Museveni who, like his presidential neighbours of Rwanda (Kagame), DRC (Kabila) and Congo (Sassou-Nguesso), has been busy removing constitutional obstacles to his life presidency. Chances for a successful EAC mediation seemingly improved when former president Benjamin Mkapa of Tanzania stepped in as facilitator in March 2016. However, after four rounds of inter-Burundi dialogue in Arusha (Tanzania), no result has been obtained. In his opening statement to the third round in February 2017, Mkapa stated that there was a general consensus among Burundi’s international partners that ‘fundamental constitutional amendments should not be advanced until the situation has stabilized’. However, in his briefing to the EAC Summit on 20 May 2017, shortly after President Nkurunziza established a commission to draft a constitutional amendment, Mkapa stated: ‘Whither the EAC-led mediation whose dialogue I am facilitating? For I fear the region will find itself before a fait accompli’, prophetic words indeed. At the time of writing of this analysis, the next steps in the EAC mediation remain highly unclear.

In the campaign ‘Teshwa Ute’ (‘Stop now!’) they launched on 5 January 2018, civil society actors continue to call upon regional and international actors to actively oppose the constitutional referendum. Their call is unlikely to be heeded. The launching ceremony for the referendum campaign on 12 December 2017 was attended by the US and the Chinese ambassadors, as well as by senior representatives of most EU embassies.

Far-reaching reforms

Only a selection of the far-reaching draft amendments can be introduced here. The current Constitution is largely based on the constitutional blueprint laid down in the Arusha Peace and Reconciliation Agreement (APRA) of 28 August 2000. Mediated by Nelson Mandela, and although it was not signed by CNDD-FDD, the APRA was a milestone towards ending a decade of ethnic civil war. It has rightly been praised as a successful case of conflict resolution through constitutional engineering along consociational lines.

The  reforms would consolidate a gradual de facto transition from ethnic power-sharing towards mere ethnic position-sharing in a one-party system.

Analysts feared that the ongoing constitutional reform would further renege on the APRA-based power-sharing arrangements by removing ethnic quotas from the Constitution and thus potentially reopening the wounds of Burundi’s violent ethno-political divide. However, the draft constitutional reform by and large maintains ethnic quotas. In fact, it extends their applicability to the judiciary, using the 60% Hutu / 40% Tutsi and minimum 30% women proportions that also apply to the government and to the National Assembly. In the Senate, the 50% Hutu / 50% Tutsi ethnic parity with cooptation of three Twa senators is maintained.

Yet, while seemingly endorsing the APRA spirit and architecture, other amendments strongly reduce the actual weight and impact of the quotas. The reform would consolidate a gradual de facto transition from ethnic power-sharing towards mere ethnic position-sharing in a one-party system. In the political sphere, the impact of the 40% Tutsi quota is undermined by the removal of the 2/3 qualified majority requirement for the adoption of legislation (except organic laws) in the National Assembly. Political parties obtaining 5% of the votes will no longer be automatically entitled to a ministerial position. While the APRA introduced such a political life-insurance for small Tutsi parties as part of a coalition government, ethnic quotas under the forthcoming constitution would possibly be respected within a single-party government. A fairly powerless vice-president must be from an ethnic group different from the president, but that requirement would not apply to the newly introduced and more influential prime minister, who would be appointed by, and accountable to, the president. Overall, presidential powers would increase to the detriment of parliament. For instance, laws adopted in parliament but not promulgated by the president within 30 days would become null. In the security sphere, the ethnic parity requirement (50% Hutu / 50% Tutsi) will no longer apply to the omnipresent intelligence service (SNR), contrary to the police and the army. According to the International Crisis Group, the army has already been undermined by ethnic polarization as well. A ‘soft’ sunset clause requires the senate to evaluate the continued use of ethnic quotas (in the legislative, executive, security and judicial sphere) within five years.

No debate on the far-reaching constitutional reform has thus far taken place either in the Senate or in the National Assembly.

The presidential two-term limit is not removed, but modified in two ways. First, the president will be elected for seven rather than five years. National Assembly elections will continue to be held every five years. As a result, in some five years from now, Burundi is set to embark on over a decade of quasi-permanent electoral campaigning, with elections to be held in 2025 (legislative), 2027 (presidential), 2030 (legislative), 2034 (presidential) and 2035 (legislative), no ideal recipe for political stability if the elections are somewhat competitive. For the Senate and municipalities, the constitution remains silent so the current five-year term could possibly be extended to seven years through an amendment of the electoral code. Second, the new presidential term limit will only apply to two consecutive terms. An outgoing president may run again seven years after leaving office. Remarkably, the draft constitution remains silent on the eligibility of Nkurunziza in 2020. Although all observers and actors seem to assume that Nkurunziza will be able to rule until 2034, if re-elected in 2020 and 2027, the absence of an explicit transitional provision allowing him to run for president in 2020 is noteworthy. At the very least, the pen-holders amending the term limit have – deliberately or inadvertently – created a legal loophole and a constitutional ambiguity. Burundi’s recent past shows that, depending on the intra-CNDD-FDD balance of power prevailing at the time of nominating electoral candidates, constitutional loopholes and the need for constitutional interpretation can further enhance political controversy and instability.

Bypassing parliament?

So far, parliament has been completely sidelined. No debate on the far-reaching constitutional reform has thus far taken place either in the Senate or in the National Assembly, where some critical voices led by another former Hutu rebel movement leader, Agathon Rwasa of PALIPEHUTU-FNL, might oppose both the procedure and the substance of the proposed constitutional reform. However, there might be a technical legal reason why the matter needs to be brought to the attention of parliament before the revised constitution can enter into force. After the referendum, presumably with a majority voting in favour, the Constitutional Court will announce the results. The revised 2005 Constitution would enter into force after its promulgation by the President. A legal case could be made that, under the current constitution and in accordance with the electoral code (article 203), the President can only promulgate a legislative act, not a Constitutional Court judgment announcing the result of the referendum or any other text. Under the current constitution, a legislative act means either a presidential decree (which clearly cannot be used here) or a law adopted in parliament, in this case in accordance with the strict qualified majority requirements laid down in article 300 (a four-fifths majority in the Assembly, a two-thirds majority in the Senate). The potential political implications of this legal issue are obviously considerable. However, it is unclear when, by whom and before which body, this technical legal concern might be raised.


For the opposition-in-exile movement CNARED, the announcement of the constitutional referendum crosses a red line. It has called for a boycott, which is unlikely to happen in a context where authorities have already announced penalties against anyone obstructing the referendum, and for international sanctions. For the first time, CNARED publicly stated that an armed insurgency is a legitimate means ‘to prevent dictatorship’ in Burundi. For the time being, however, the process leading towards the 2018 referendum and, next, towards the 2020 elections is likely to be determined, above all, by internal CNDD-FDD developments.

The constitutional amendments are likely to consolidate the fait accompli of CNDD-FDD hegemony, but not yet and not necessarily Nkurunziza’s life presidency.

Seen from this angle, Nkurunziza is likely to take the 2018 referendum hurdle without much difficulty. Indeed, even internal fourth term opponents may well support the constitutional reform for a variety of reasons and, above all, because the new constitutional framework is likely to ‘formalize’ CNDD-FDD control over the state institutions and to concentrate more power in the hands of the future president, most probably Nkurunziza or the CNDD-FDD nominee, in the foreseeable future. The new constitution may thus benefit both Nkurunziza… or another CNDD-FDD presidential candidate. In fact, although the incumbent clearly seeks to create a continuum between both events, the constitutional amendment does not exclude a next leadership battle in the run-up to the 2020 elections. Time will tell whether Nkurunziza will also be able to take that hurdle. In other words, a 2018 constitutional amendment is likely to consolidate the fait accompli of CNDD-FDD hegemony, but not yet and not necessarily Nkurunziza’s life presidency.

While unlikely to publicly applaud the constitutional reform, international partners keen on preserving short-term stability rather than on promoting democracy, and all of them seem to have given up the latter ambition, may well silently prefer this timely two-step consolidation process over a last-minute pre-2020 chaos reigniting the 2015 turbulence. At the same time, they should realize that Burundi’s history since independence (1962) has been marked by several constitutional ruptures and periods of constitutional void. Never before has such a far-reaching constitutional reform taken place in an orderly fashion, without being either preceded or followed by a palace revolution or a civil war.

Dr. Stef Vandeginste is a lecturer at the Institute of Development Policy, University of Antwerp. The constitutional history of Burundi, with all of the original texts of the constitution since independence, can be consulted on his website ‘Law, Power and Peace in Burundi’, available here: www.uantwerpen.be/burundi, section Constitution.

Disclaimer: The views expressed in Voices from the Field contributions are the author's own and do not necessarily reflect International IDEA’s positions.


Post new comment