Term Limits and Democratic Consolidation in Sub-Saharan Africa: Lessons from Burundi

On July 12, 2015, delegates from Tanzania’s ruling party (CCM - Chama Cha Mapinduzi) elected John Pombe Magufuli, 56, to be the party’s flag bearer in the General Election due on 25 October 2015.  The selection most likely guaranteed that Magufuli will be elected Tanzania’s fifth president. Current president, Jakaya Kikwete, is due to retire after October’s election having served his second and final term. Since 1985 three different Tanzanian presidents have honored the country’s constitutional term limits. Despite CCM’s hegemonic party status and dominance in the legislature, the party has consistently been able to enforce among its members respect for the constitutional two-term limit.

The events in Tanzania contrasted sharply with those in its neighbor to the west, Burundi. There, President Pierre Nkurunziza pushed to stay in office beyond his second term. The Burundian constitution, a product of the Arusha Accords that ended the country’s 12-year civil war, limits presidents to two terms. However, president Nkurunziza was determined to stay in office beyond his second term. He therefore submitted to Parliament constitutional amendments to the same effect. But, despite his ruling party’s overwhelming majority in Parliament, Nkurunziza narrowly lost a Parliamentary vote in March 2014 that would have allowed him to seek a third term. He then took the matter to the courts, arguing that the constitution barred individuals from two elective terms -- a restriction that did not apply to him since in 2005 he was indirectly elected by Parliament and not via a popular vote. Under pressure from the president, the Constitution Court ruled in his favor, but not before its vice president, Sylvere Nimpagaritse, fled the country after receiving death threats. Nkurunziza’s bid for a third term unleashed protests that led to a coup attempt on May 13 while the president was away in Tanzania for talks over the political crisis. But divisions among the disciplined forces allowed him to return to the country and resume power. However, the violence has not abated, with more than 100 people reported dead and an estimated 158,000 forced to flee the country according to the UN.

On June 29, 2015 President Nkurunziza went ahead with legislative elections amid protests and opposition boycott. The official results showed the ruling party, CNDD-FDD, to have won 77 of the 100 seats in the National Assembly. International observers, including the African Union, boycotted the election and declared it to have failed to meet the minimum threshold for free and fair elections. And on July 21 the country held the presidential election in the face of international protests and a domestic boycott. Nkurunziza will now serve a third term as president, remaining in office until 2020.

The cases of Burundi and Tanzania are symptomatic of general trends across much of sub-Saharan Africa. Yet, in a subset of countries, the requisite political culture to facilitate respect for term limits is gaining a foothold. Countries from Kenya, to Mozambique, to Ghana have had multiple presidential successions over the last two decades. Still, in others like Uganda, Chad, and Cameroon, presidents have managed to hang on to power beyond their constitutionally-sanctioned stay in office. There are also countries like Zambia, Niger, and Senegal where attempts to extend term limits failed.

Across all of these cases, the one constant in all of the countries for which survey data is available (via Afrobarometer) is that the vast majority of Africans unambiguously support term limits (75% on average) – including 87% in Tanzania and 61% in Burundi.

The crisis in Burundi provides important lessons about political development in Africa’s fragile electoral democracies. What it shows is that constitutional engineering alone is not enough to guarantee respect for the rule of law. Along with constitutional safeguards for term limits, emerging democracies also require sufficient elite consensus on the rules of the game and bounds of acceptable political behavior. Ultimately, institutional safeguards against personalist rule are only as strong as the specific interests that populate the same institutions. In other words, the de facto balance of power among elites matters for the enforcement of constitutional provisions such as term limits. This brief essay demonstrates this fact with a survey of term limit violations across Sub-Saharan Africa since the early 1990s and with a discussion of the specific case of Burundi.

Constitutional Term Limits in Africa

Research shows that lack of regular turnover in leadership fosters the entrenchment of personalist politics, stunts the process of institutionalization of politics, and increases the risk of political instability. For these reasons, term limits are common phenomenon in many of the world’s electoral democracies. For example, a 2007 study found that 87 out of the examined 92 countries at one point had term limits (67 percent of African countries have term limit provisions in the constitution). But, while countries may have provisions for term limits in their constitutions, honoring them is a different matter altogether. The same study found that 23% of electoral authoritarian regimes had eliminated term limits. Interestingly, most of the leaders who violated term limits engaged in “soft” contraventions. Like Nkurunziza, they merely shifted the constitutional definition of a “full term” without scrapping term limits altogether. This they achieved through court rulings, parliamentary votes, or public referenda.

The case for term limits in Africa is clear. Peaceful regular leadership transition is the hallmark of political development and long term stability. Regular leadership succession also favors the institutionalization of politics. The cycling of elites during transitions allows for the emergence of elite coalitions that cycle in and out of power, and that can at any one time check the ruling coalition. The lack of leadership turnover weakens elites out of power, disincentivizes investments in institutions, limits economic activity to those close to the incumbent ruler, and fosters the entrenchment of personalist politics by putting a premium on unquestioning loyalty to the incumbent leader. The resultant sub-optimal state of politics is especially pernicious in countries where identity politics based on race, language group, or ethnicity predominates. The level of ethnic fractionalization in Africa and significant evidence of ethnic favoritism in public goods provision in the region puts Africa at risk of political instability when one leader monopolizes power for too long. From a normative standpoint, the political economy of Africa demands regular leadership turnover as a means of both securing stability and ensuring equal economic opportunities across ethno-linguistic lines.

Since the early 1990s, 24 presidents in Sub-Saharan Africa initiated discussions in attempts to stay in office for more than two terms when confronted with term limits. In 15 countries presidents started the process of actually amending the constitution. In 12 of these cases (Burkina Faso, Burundi, Cameroon, Chad, Gabon, Guinea, Namibia, Niger, Rwanda, Senegal, Togo, and Uganda) the presidents succeeded. In three (Malawi, Nigeria, and Zambia) they failed. At the same time, leaders in Benin, Cape Verde, Ghana, Kenya, Mali, Mozambique, Sao Tome and Principe, Tanzania, and recently Namibia have adhered to constitutional term limits without a protracted challenge. In the next few years, presidents in Angola, the Democratic Republic of Congo (DRC), Liberia, the Republic of Congo, and Sierra Leone will face constitutional term limits. Political developments in these countries suggest that only the presidents in Liberia and Sierra Leone plan to step down at the end of their second terms.

The modes of altering constitutional provisions for term limits differ depending on context. In the three cases where presidents failed, divisions within both the ruling parties and parliaments proved critical. In cases where presidents had overwhelming majorities in the legislature and could control the institutions, they were able to effect constitutional amendments via parliamentary vote. However, under conditions whereby presidents were unsure about a parliamentary acquiescence they chose to hold popular referenda, or to seek judicial interpretation of the constitution in their favor.

In Namibia, President Samuel Nujoma altered the constitution in order to run for a third term in 1999. Article 28 (2) of the constitution stipulated that the president was to be elected by direct, universal and equal suffrage. Article 29 (3) then added that a person shall hold office for president for not more than two terms. Nujoma, like Nkurunziza in Burundi, argued that because in 1989 he was indirectly elected before serving his first term the constitutional restriction did not apply to him. He prevailed in Parliament and served a third term before retiring in 2004. Similarly in Burkina Faso, President Blaise Compaore argued in 2005 that the term limit restriction in Article 37 of the constitution did not apply retroactively (he had already served two previous terms), thereby allowing him to serve two new terms until 2014. Then, when he tried to abolish term limits altogether in 2014 and stay in office indefinitely, riots broke out, resulting in a successful coup. Abdoulaye Wade of Senegal used similar tactics in 2012, arguing that the limits did not apply retroactively to his first term in office, and leaned on the courts to win the case that allowed him to run for a third term – only to lose the subsequent election. All these were “soft” contraventions, whereby presidents used perceived ambiguity in the law to get favorable decisions either in the courts on in parliament.

In other countries presidents carried out “hard” contraventions, abolishing term limits altogether. In 2001 president Lansana Conte of Guinea called for a referendum that scrapped term limits. Similarly, in 2005, President Idris Deby of Chad held a referendum to delete Article 61 (2) of the constitution which restricted presidents to two successive terms. Mamadou Tandja of Niger, having failed in both Parliament and the courts, scrapped term limits through a referendum even though Article 49 of the Nigerien constitution expressly forbade such a procedure.

While the above presidents chose the referenda route, their counterparts that were confident of the support in the legislature chose to amend constitutions via parliamentary vote. In 2002 Gnassingbe Eyadema of Togo, after securing a parliamentary majority in an election that was boycotted by major opposition parties, successfully abolished term limits via the legislature. A similar move in Gabon in 2003 enabled President Omar Bongo to stay in office indefinitely. Yoweri Museveni of Uganda also abolished term limits through a parliamentary vote that repealed Article 105 (2) of the constitution in 2005 - but only after allegedly paying legislators $2000 each. In Cameroon Paul Biya pushed through Parliament in 2008 a repeal of Article 6 (2) of the constitution that did away with term limits. Omar Guelleh of Djibouti also went through Parliament for his extension in 2010, having been elected unopposed in 2005. And this year in Rwanda President Paul Kagame did away with term limits to allow him to stay in office beyond 2017, also on the back of the ruling party’s overwhelming parliamentary majority.

The emergent pattern is clear. Presidents that have a comfortable majority in parliament choose to amend the constitution in that institution rather than through a referendum. However, should that fail they typically go to the courts. Given that in most African countries judges are appointed by the presidents and serve at their leisure, the courts offer an easy avenue through which presidents can contravene term limits. If machinations at the institutional and elite level fail, presidents then opt for referenda. Popular votes offer presidents the chance to influence the outcome through patronage and outright rigging. For example, in Chad’s 2005 referendum the electoral commission announced that the total registered voters amounted to 5.6 million, even though the voter age population was estimated to be around 4 million.

What is also clear is that presidents have several avenues – both legal and political – that they can use to contravene term limits. However, the cases in which presidents fail, offer an important lesson: constitutional and institutional protections of term limits are a necessary but not sufficient condition for their observance. Such protections must also be accompanied by the existence of an elite consensus about the inviolability of term limits. That was the case in Malawi, Nigeria, and Zambia. It was also the case in Burkina Faso, Niger, and Senegal whereby elites supported the move to punish leaders who violated term limits either through the ballot or via coups.

Constitutional Limits Are Not Enough

The question then becomes: how can term limit observance be secured in Sub-Saharan Africa? There is no simple answer to this question. The fact of the matter is that respect for term limits evolves as part of a country’s political culture. It is also often founded on a sense of elite consensus and mutual respect. The importance of elite consensus on legitimate political behavior in the early stages of democratic consolidation cannot be gainsaid. The current political crisis in Burundi is a direct result of the lack of elite consensus on the rules of the game; and not necessarily constitutional ambiguity or weakness.

Burundi’s post-conflict constitution provides a robust array of formal checks to personal rule. Article 164 provides for a 60-40 Hutu-Tutsi split in National Assembly and 50-50 split in the Senate in order to ensure that the majority Hutu (85%) do not violate the rights of the minority Tutsi (14%). The Batwa (1%) are also guaranteed representation in Parliament through special nomination. Burundi also has a proportional representation (PR) system with a closed list that requires political parties to nominate no more than two thirds of candidates from the same ethnic group. Article 257 of the constitution reinforces the principle of ethnic balance by mandating a 50-50 split in the military. Furthermore, according to Article 300 any amendment to the constitution requires an 80% super-majority in the National Assembly and two thirds of the Senate (this is why Nkurunziza failed in an attempt to amend the constitution in early 2014). 

Yet, despite these constitutional and institutional safeguards, Nkurunziza has managed to violate term limits. He achieved this through a selected reading of the constitution and a manipulation of the constitutional court. Article 96 of the constitution says that the “President of the Republic is elected by universal direct suffrage for a mandate of five years renewable one time.” And under the transitional clause, Article 300 says that “Exceptionally, the first President of the Republic of the post-transition period is elected by the [elected] National Assembly and the elected Senate meeting in Congress, with a majority of two thirds of the members.” The first word in Article 302 highlights the unique circumstances of the first election of the president following the transition from civil war. But it does not negate the fact that a president elected under this clause is constitutionally in office for a valid term. Article 96 and 302 must therefore be read together. This was the conclusion of a leaked confidential report from the Attorneys General and Ministers of Justice within the East African Community on the legality of Nkurunziza’s third term bid.

It is telling that early last year when the plan to extend Nkurunziza’s rule was first made public, his Plan A was to go after Article 302. The proposed amendments that failed would have included the deletion of Article 302 thereby allowing Article 92 to stand on its own. This would have given the president grounds to invalidate his first indirect election in 2005 and run for a successive third term. Plan B involved the constitutional court. The court is composed of seven members appointed by the president who serve six-year non-renewable terms. When it was first presented with the case the court was equivocal. News reports suggest that at its first sitting on April 10, 2015 the court initially found that the president was ineligible to run for another term in office. However, after considerable political pressure – including death threats and the exiling of the court’s vice president – the court reversed its decision and delivered a verdict in favor of the president. The president had thus acquired the legal sanction to run for a third term.

The Role of the International Community

As the account above shows, the main lesson from Burundi is that constitutional term limits alone are not enough to guarantee peaceful leadership turnover. Elite consensus matters as well. Because key elites within CNDD-FFD and the military did not want to relinquish power, they ensured that the party did not offer a credible alternative to Nkurunziza. This led to an escalation of conflict with the opposition that further reinforced the need for CNDD-FFD to band together and defend their hold on power. The current crisis has, perversely, strengthened the hand of this hardline section of the Burundian elite class.  

The international community also has a role to play in fostering respect for term limits. In Burundi, the first opportunity was in March 2014 when the president initially failed to amend the constitution. The failure was an opportunity for a robust engagement with the CNDD-FFD with the aim of cultivating alternative candidates to Nkurunziza. Given the party’s overwhelming dominance, it would have been prudent for the international community to accept the fact that any future president after Nkurunziza would most likely have come from within the party. Dominant parties in Africa, from Tanzania to Mozambique, have been able to successfully preside over intra-party leadership transitions and could have provided viable models for the CNDD-FFD. The unsuccessful parliamentary vote also provided an opportunity for the international community to come out unequivocally against a third term.

Yet, geopolitical concerns prevented this from happening. In addition, East Africa as a region is not particularly exemplary on this score. For example, of the five countries in the East African Community (EAC), only two (Kenya and Tanzania) had historically respected term limits. Uganda’s Yoweri Museveni (in power since 1986) scrapped term limits in 2005; and Rwanda is in the middle of doing so to allow Paul Kagame to stay in office beyond 2017. The EAC was therefore prevented from coming out strongly against Nkurunziza’s third term. Tanzania’s president, Jakaya Kikwete, was the only regional leader who opposed the move, only to backtrack later. The equivocal regional response to his machinations emboldened Nkurunziza to go ahead with his plan to violate the spirit of the Burundian constitution. Wider regional concerns about Rwanda’s support for Burundian rebels (and its security implications for the wider Great Lakes region) also aggravated the situation, and forced the international community to prioritize stability over pushing for democratic consolidation. Burundi is also a significant contributor of troops to the AMISOM mission in Somalia, a fact that gave Bujumbura leverage over the wider international community despite the fact that it depends on donors for nearly half of its budget.

Implications for Democratic Consolidation in Africa

The inherent political problem in attempts to secure constitutional term limits is clear. Depending on the context, courts, parliaments, and popular referenda offer opportunities that can be exploited by presidents. The twelve countries in Sub-Saharan Africa that have witnessed successful term limit extensions demonstrate that the absence of an elite consensus on the inviolability of term limits renders constitutional and institutional safeguards futile. This means that structural conditions that result in a de facto balance of power among elites in and out of power are crucial for democratic consolidation. Such a balance of power also results in a willingness among elites – both within and outside of the ruling coalition – to protect term limits. Unfortunately, in Sub-Saharan Africa, government-opposition balance of power is yet to be achieved in all electoral democracies. Presidents command overwhelming majorities in legislatures, control the appointment of judges, and can commandeer enough patronage resources to buy off legislators or secure victories in referenda.

Attempts at constitutional design to secure term limits and to foster democratic consolidation in the region must therefore aim at engineering a de facto balance of power among elites. Given that no constitution is a complete contract, the goal ought to be to increase the cost of term limit extensions. Focus should be on constitutional provisions that limit legislative seat malapportionment that favor ruling parties; curtailment of presidential control over the legislative agenda and calendar; a strengthening of judicial independence; and support for democracies that observe term limits with a view of reaching a critical mass of term limit observers. If enough countries uphold term limits, observance will become a norm and will have a powerful demonstration effect to countries that have or are contemplating the scrapping of term limits. In the main, over the last two decades the institutionalization of political power in Africa has seen an upward trend. Fewer leaders enter or exit power via extra-constitutional means than did in the past. Elections are the norm, and the African Union sanctions leaders who enter power via coups. Term limits are therefore the next frontier in the process of democratic consolidation in the region. However, just like with the prevailing strong norm against coups, success will hinge on the achievement of a critical mass of states that observe term limits in the region.

Ken Opalo is an Assistant Professor at Georgetown University. His research interests include institutional development in emerging democracies and the political economy of development. Ken is currently working on a book project titled, Institutions and Political Change in Africa: The Case of Legislatures. Ken’s work has been published in the Journal of Democracy and the Journal of Eastern African Studies.