Tag Archives: power-sharing

Burundi – New constitution, new president?

Burundi adopted a new constitution on May 17, 2018 by referendum, with 73 percent voting in favor. The adoption of a new fundamental text was seen by opponents as a move by President Pierre Nkurunziza to extend himself in power by resetting the term limit clock to zero, while by the same token doing away with power sharing provisions from the 2000 Arusha Peace and Reconciliation Agreement. The campaign period was tense and, according to Human Rights Watch, at least 15 people were killed. The opposition called for the outcome to be annulled due to vote-rigging and intimidation, but was overruled by the constitutional court that validated the results.

Against all expectations, at the ceremony for the promulgation of the new constitution, on June 7, President Nkurunziza declared on national TV that he will not stand for reelection in 2020, when his current mandate ends, stating that “This constitution was not modified for Pierre Nkurunziza as the country’s enemies have been saying. It was amended for the good and better future of Burundi and the Burundian people.”

Nkurunziza, in power since 2005 at the end of the civil war that killed 300,000 people, stood for and won a highly controversial third term in 2015 [see previous blog post here]. A failed coup and crack-down against opponents followed, and it is estimated than 400,000 Burundians have since fled the country, out of a population of 10 million. Under investigation by the International Criminal Court (ICC), Burundi became the first country to leave the ICC in 2017.

So what are some of the main changes included in the new constitution? As it appears, a number of provisions for ethnic power sharing have been maintained, while some requirements for power sharing between parties have been eliminated:

  • Burundi returns to semi-presidentialism (the country was previously semi-presidential from 1992 to 1994) with a prime minister as head of government, accountable to both the president and the legislature. The candidature of the prime minister must be approved by both chambers of parliament voting separately (Art. 130), and he or she can be dismissed by a two third majority vote of members of the National Assembly (lower house) – though the president in turn can dismiss the National Assembly (Art. 208). There are no constraints with regards to the ethnicity or the party affiliation of the prime minister.
  • The length of presidential terms is extended to 7 years from five (Art. 97); also, the provision regarding term limits now states that no one can serve more than two consecutive terms – which would seemingly leave open the door for a Putin-like come-back as president after a stint as prime minister.
  • There is now only one vice-president instead of two, who assists the president in the exercise of his or her functions. The vice-president must be approved by both houses of the legislature and must belong to a different ethnic group and party/coalition than the president. Previously, the two vice-presidents had to be from different ethnic groups and parties [from each other, not necessarily from the president]. The First Vice-President was responsible for the coordination of the political and administrative domain, and the second Vice-President for the coordination of economic and social affairs. In the new constitution, the role of the vice-president is left at the discretion of the president.
  • The provision for proportional representation of parties in the cabinet having earned more than 20% of the vote has been removed. The ethnic and gender representation requirements that, overall, at most 60% of cabinet ministers can be Hutu and at most 40% can be Tutsi, and at least 30% must be women (Art. 128) remain in place. Also maintained is the requirement that the Minister in charge of National Defense is not from the same ethnic group as the Minister responsible for the National Police (Art. 135).
  • The parliamentary majority required to pass legislation is reduced from a super majority of two thirds to a simple majority (Art. 180).

While streamlining governing processes – by introducing a prime minister as head of government, removing one vice-president and eliminating the requirement for a super majority to pass legislation – the constitutional changes also eliminate some of the power-sharing provisions enshrined in the 2005 constitution, in accordance with the Arusha Accords. As noted in a previous blog, these power-sharing arrangements had been successful to the extent that: “Today, political competition in Burundi no longer coincides with ethnic cleavages. Furthermore, the dominant party CNDD-FDD, while rooted in a Hutu rebel movement, is no longer perceived as an exclusive Hutu party. In fact, most Tutsi members of parliament are members of the CNDD-FDD and many presidential advisors are Tutsi” (Vandeginste, 2009 p. 75).

However, in the new constitution the representation in government of either of the two majority ethnic groups is still capped at 60% and 40% for Hutus and Tutsis respectively; this does maintain pressure on political parties to be ethnically inclusive (Hutus account for around 85% of the population, Tutsis around 19%, and Twa people around 1%). So not all power-sharing provisions are lost.

The constitutional changes do remove some constraints on the president’s powers and clearly provide greater opportunities for extended stays in the presidential palace. So it is intriguing that Nkurunziza has stated he will not run again in 2020. According to critics, 2020 is still far away, however, leaving Nkurunziza “room to maneuver” in response to “popular pressure” for him to extend his stay in power. That would certainly not be an unexpected development.

Dan Slater – Victory vs. Reciprocity: Presidential Power-sharing and Party Cartelization in Indonesia

This is a guest post by Dan Slater, University of Michigan. It is based on his article in Journal of East Asian Studies

Democracy and opposition are supposed to go hand in hand. Opposition did not emerge as automatically as expected after Indonesia democratized in 1998, however, because presidents shared power much more widely than expected. The result has been what I call party cartelization, Indonesian-style. As I argue in my new article in the Journal of East Asian Studies, this differs significantly from canonical cases of party cartelization in Europe. Yet it exhibits the same troubling outcome for democratic accountability: the stunted development of a clearly identifiable party opposition.

Since the advent of direct presidential elections in 2004, Indonesian democratic competition has unsurprisingly assumed somewhat more of a government-vs.-opposition cast. But this shift has arisen more from contingent failures of elite bargaining than from any decisive change in the power-sharing game. So long as Indonesia’s presidents consider it strategically advantageous to share power with any party that declares its support, opposition will remain difficult to identify and vulnerable to being extinguished entirely in the world’s largest emerging democracy.

I reached these conclusions by asking three interrelated theoretical questions. First, how does opposition emerge as a political process in newly democratic settings? Second, how do democratically elected presidents share power and build ruling coalitions? And third, how might new political rules reshape those power-sharing practices?

Presidential power-sharing is a strategic political game. It is shaped, accordingly, by political institutions. Of particular importance are the rules governing selection of the chief executive; in Indonesia’s case, always a president. If a president is elected by parliament, as in Indonesia from 1999-2004, he or she is an agent of parliament. He can be expected to share power, roughly proportionally, with the parties resident there that selected him. If the people elect the president, he is an agent of the people, and should face less imperative to share power with parties in parliament that not only played no role in putting him there, but in many cases directly opposed his candidacy.

Yet in both instances, the same implicit assumption underpins our expectations. We assume that a president will share power with whichever parties helped put him in power, and not with those who played no role or even tried to prevent his election. This is what I call Victory: a power-sharing game predicated upon the unwritten rule that presidents will share power only with parties that supported him during his election campaign. To the extent that Victory is the power-sharing game, identifiable party opposition arises automatically. Someone must lose, so someone must go into opposition.

But what if Victory is not the game presidents play? Either in the presence or absence of direct presidential elections, a president might offer to share power with any and all parties that promise to support the presidency, even if they earlier opposed the presidential candidate. Instead of Victory, I call this power-sharing game Reciprocity. If a president prefers or is pressured to play Reciprocity, the emergence of identifiable party opposition becomes contingent rather than automatic. So long as post-electoral Reciprocity bargains can be struck with all parties, all parties can join the executive. Identifiable party opposition may thus vanish, as it did in Indonesia from 1999-2004, even in a perfectly functional and democratic electoral system. Someone must lose the election, but no one has to lose power.

This allows us to recast Indonesia’s struggle to generate an identifiable opposition in straightforward theoretical language. Party cartelization, Indonesian-style rests upon the power-sharing game of Reciprocity. Direct presidential elections will only disrupt or dismantle the cartelized party system if presidents build coalitions comprised of parties that supported him as well as nonparty allies of his own choosing, through the game of Victory.

Yet there are two critical wrinkles to consider. The first is that presidents not only make strategic choices about whom to share power with, but about how much power each partner will receive. Power-sharing games involve distributional conflict among coalition partners, not just between government insiders and outsiders. This means that presidents can strategically provide bonuses to existing supporters through a super-proportional share of cabinet seats, while relegating previous opponents to a sub-proportional share.

Hence when examining cabinet data, we must be attentive not only to whether presidents are sharing power with parties that opposed them during the election (i.e. playing a Reciprocity game), but also to deviations from the principle that cabinet seats should be distributed proportionally to coalition partners. This should indicate whether presidents have always strategically offered bonuses to electoral backers and imposed punishments on electoral opponents, and whether they are doing so more often since direct presidential elections were introduced. The more willing Indonesian presidents are to sideline erstwhile opponents, the more they shift from a Reciprocity game toward a Victory game, and the better the prospects become for identifiable opposition to emerge and strengthen in Indonesia.

The second caveat is perhaps even more important. It is that presidential coalitions are not necessarily faithful reflections of a president’s strategic preferences. Although presidents can choose to play a Victory game by fiat, Reciprocity is a resolutely two-sided game. In other words, Victory games only require a directly elected president to exclude electoral opponents from power as a unilateral strategy. Reciprocity demands that they engage those former opponents in a more complicated, multilateral bargaining process.

Whether a president seeking to play Reciprocity can actually find willing coalition partners at a price the president is ready to pay depends not on executive decree, but on hard political bargaining. Hence even when we see a coalitional outcome that seems to reflect a Victory game, we must examine whether the absence of Reciprocity arose from a president’s strategic decision to play Victory, or from his contingent failure to “seal the deal” with active negotiating partners in an ongoing Reciprocity game.

The implications of this seemingly minor distinction are quite major. If direct presidential elections have emboldened Indonesia’s presidents since 2004 to start pursuing Victory rather than playing Reciprocity, then the strategic underpinning of party cartelization is seriously weakening. This would mean that recent moves toward more identifiable opposition, as detailed in my JEAS article, are unlikely to be reversed. But if directly elected presidents are still playing Reciprocity, and simply failing to strike bargains, then the game of power-sharing remains unchanged, even as the final outcome has shifted. This implies that a return to the full party cartelization of the 1999-2004 period remains a meaningful specter, even more than a decade after direct presidential elections were introduced and the party cartel was first disrupted.

As my article’s data and narrative show, presidential power-sharing in Indonesia has gradually drifted, but not definitively shifted, from a Reciprocity game toward a Victory game. In raw quantitative terms, the figure below unmistakably shows that parties have increasingly positioned themselves outside of government since 2004. Yet the numbers obscure much of what my qualitative assessment reveals. The lingering importance of Reciprocity can still be seen in vigorous efforts by both President Susilo Bambang Yudhoyono (2004-14) and President Joko Widodo (or Jokowi, 2014-present) to forge alliances across the full range of Indonesian parties.

In my earlier collaborative research, I call this promiscuous power-sharing: “an especially flexible coalition-building practice in which parties express or reveal a willingness to share executive power with any and all other significant parties after an election takes place, even across a country’s most important political cleavages” (Slater and Simmons, 2013). From this perspective, promiscuous bargaining has continued almost unabated since 2004, but it has not always been consummated in power-sharing bargains. In sum, promiscuous power-sharing primarily arose from 1999-2004 because parliamentary parties had the power to demand it; it has persisted since 2004, even while evolving and abating, because even directly elected presidents have had a strategic interest in maintaining it.

Continued attempts at promiscuous power-sharing strongly suggest that Reciprocity remains the dominant game. Party cartelization has abated in Indonesia, but not vanished. And it could still easily come back in its most extreme form. Even if it does not, the public willingness of all parties to consider power-sharing alliances with all other parties means that Indonesia’s voters can never be confident that a vote for one party means a vote against any other. Under conditions of promiscuous power-sharing, objectionable and unpopular parties and individuals can only be removed from office by elites, not by the voters. Indonesia shows that direct presidential elections make party cartelization harder, but far from impossible.

In conclusion, the most intriguing implication of Indonesia’s experience with democratic power-sharing may be this: Presidents may sometimes see broad coalitions as a source instead of a drain on their power and resources. Oversized coalitions are typically seen as being more expensive to maintain. But this may not be how presidents see things at all, at least under certain conditions. Oversized coalitions may be better conceived as ways for presidents to spread the same amount of resources across more claimants, thus ensuring that no single partner can become too strong as a rival. If nothing else, the persistence and evolution of party cartelization, Indonesian-style suggests that power-sharing should not be seen as occasions for presidents simply to give. Political scientists should look more carefully to see what presidents may sometimes be taking away.

Burundi backsliding as 2015 presidential poll approaches

Previously hailed for its success in bringing peace and stability to the country, Burundi’s power-sharing arrangement is rapidly unraveling. The 2005 constitution that brought an end to the 1993 civil war between Hutus and Tutsis, instituted a complex set of power-sharing provisions, including a stipulation that the president of the republic shall be assisted by two vice-presidents belonging to different ethnic groups and parties (art. 124). The constitution also mandates that parties must not be established exclusively on ethnic or regional bases.

By some accounts, these power-sharing arrangements have been successful to the extent that: “Today, political competition in Burundi no longer coincides with ethnic cleavages. Furthermore, the dominant party CNDD-FDD, while rooted in a Hutu rebel movement, is no longer perceived as an exclusive Hutu party. In fact, most Tutsi members of parliament are members of the CNDD-FDD and many presidential advisors are Tutsi” (Vandeginste, 2009 p. 75).

The power-sharing provisions have not succeeded, however, in creating a fertile ground for consolidating democratic practices. The opposition boycott of the 2010 legislative and presidential elections following allegations of fraud in local elections earlier in the year, left the political system dominated by a single party, the CNDD-FDD. Leaders of Frodebu (a Hutu-led party) and FNL (ex Tutsi-rebel movement) went into exile as politically-motivated violence mounted.

Uprona (a Tutsi-led party) participated until recently in government, occupying the position of first vice-president. However,  the power-sharing arrangement enshrined in the 2000 Arusha Peace Accords collapsed as Uprona joined the opposition in February of this year, protesting the apparent intention of President Pierre Nkurunziza to run for a third term. Nkurunziza sought to maintain constitutional appearances by appointing a new vice-president and three ministers from a smaller wing within Uprona – a move immediately rejected by Uprona leadership.

The argument advanced by the CNDD-FDD justifying a third term for Nkurunziza is that “The constitution provides for two terms by universal suffrage. The first term was not a direct vote [Nkurunziza was appointed by parliament in 2005]. It’s a matter of interpretation of the constitution.” 

In addition to the tensions around another term for Nkurunziza, the opposition and Burundian civil society organizations worry about the government’s unilateral push for constitutional amendments that would do away with many of the power-sharing provisions of the 2005 constitution, including: substituting a simple majority vote for the current two thirds majority requirement in parliament; replacing the two vice-presidents with a largely ceremonial vice-president and a powerful prime minister that does not have to be from a different ethnic group than the president; and raising the threshold for party representation in parliament from 2% to 5% of the votes cast.

With the June 2015 presidential poll looming large on the horizon as a potential trigger for further violence, the UN Security Council recently voted to extend the UN political mission in Burundi, despite the government’s initial opposition.

So while power-sharing may have succeeded at least partially in blurring ethnic divisions in Burundi, as seen by the FNL and Frodebu joining forces in opposition, it clearly has not effectively served to strengthen democratic institutions and practices.