The botched results from the December 27, 2007 presidential elections in Kenya sparked a wave of violence across the country that left more than 1,000 dead and 600,000 displaced. Incumbent president Mwai Kibaki, representing the ruling Party of National Unity (PNU), was declared the winner of the presidential polls over Raila Odinga, of the opposition Orange Democratic Movement (ODM). Supporters of the ODM, which had won 99 parliamentary seats against PNU’s 43 (out of 210 elected seats), charged that the election had been rigged. The chairman of the Electoral Commission of Kenya has since stated that the PNU and the ODM-K (an allied party) forced him to call the election, even with irregularities in the tallying.¹

The African Union’s Panel of Eminent Africans, which Kofi Annan led, mediated for nearly forty days. After these efforts, Kibaki and Odinga agreed to: end the violence, address the humanitarian situation, and resolve the political crisis through a National Accord and Reconciliation Act (hereafter the National Accord). Complementing the National Accord, the parties agreed to examine long-standing sources of grievances and establish an Independent Review Commission to examine the electoral process; a Truth, Justice, and Reconciliation Commission; a Commission on Inquiry on Post-Election Violence; and the Constitutional Review Commission.

The main objectives of the National Accord are to create the positions of prime minister and two deputy prime ministers and establish a governing coalition comprised of cabinet appointees from both the PNU and ODM. The coalition government can dissolve if the current parliament is dissolved, the parties agree to it in writing, or if one party withdraws from the coalition.² Even though the National Accord is but one of several significant agreements between Kibaki and Odinga, it is undoubtedly the most important. Should it fail, the full scope of the reconciliation process could not be implemented, and violence could recur. Indeed, on April 8, the announcement of the suspension of the talks over the coalition cabinet’s composition resulted in violence in Kibera, a Nairobi slum.

What are the strengths and weaknesses of the National Accord? What factors could increase the likelihood that the power-sharing coalition remains intact? What are the costs of withdrawing from the coalition government? An initial examination of internal and external sources of leverage suggests that the cost of withdrawing from the coalition must be clearer, domestic constituencies must be more actively involved, and the militia must be disarmed.

Strengths and Weaknesses of the National Accord

On March 19, the Kenyan parliament voted to amend the constitution to provide for a prime minister and two deputy prime ministers and, subsequently, enact the National Accord into law, making Odinga the prime minister. In a show of commitment, the bill was fast-tracked into parliament, reducing the bill’s review period from 14 to 6 days.³ The most important aspect of the National Accord is the political space it provides to address the deeper sources of conflict in Kenya.

By assenting to the creation of the position of prime minister, Kibaki provided Odinga with a long-sought position. A widely reported memorandum of understanding between Kibaki and Odinga during the 2002 election promised to create the position for Odinga. However, soon after taking office, Kibaki reneged. Similarly, a key component of the constitutional reform debated in 2005 included the creation of the post of a significantly powerful prime minister; Odinga supported it while the version the government put forth reduced the powers of the prime minister. The national referendum, which featured the government’s proposal, lost. The National Accord thus deals with a key opposition grievance—that the presidency has too much power. Less overtly, the National Accord’s coalition government neutralizes, to a degree, charges that non-Kikuyu have been marginalized from power under the Kibaki regime. The post-election violence took on a distinct ethnic tone, with supporters of Odinga, a Luo, charging that Kibaki, a Kikuyu, and his administration have marginalized non-Kikuyu. While there is no dominant ethnic group in Kenya, the Kikuyu are the largest group, comprising 21 percent of all Kenyans. The Luo make up 13 percent of the population. The rest of the population divides into five main ethnic groups.⁴ By addressing these triggers of violence—both at the political and the societal level—the National Accord provides the initial political space to move forward and stop the violence.

Still, power-sharing agreements are notoriously fragile. They can break down because: impasses cannot be resolved as a result of the decision-making process in place; actors outside the coalition undermine decisions; one or more members of the coalition continually feels marginalized; or one party feels that it can gain more by not cooperating with the coalition’s decisions. Indeed, we have already begun to see the strains of the coalition in Kenya. The PNU and ODM experienced great trouble in naming a cabinet. This difficulty prompted a few parliamentarians and community leaders to call for fresh elections and triggered several incidents of violence—placing the agreement’s survival in the balance.⁵

Indeed, in Kenya’s case, the challenge is to keep both parties from withdrawing from the agreement. To some degree, the challenge is greater with the PNU than the ODM. It would be less costly for the PNU to leave; if the PNU were to pull out of the coalition, Kibaki would remain president and the PNU would have cabinet positions. In contrast, if the ODM formally left the coalition, it would also mean relinquishing the premiership and cabinet positions. While the threat of violence from the ODM party militants may deter the PNU, it is not clear how powerful that threat is: will ordinary citizens heed the call to demonstrate to the degree seen in January, after such a bloody post-election experience? Differing ODM and PNU interpretations of the National Accord, coupled with the ambiguous penalty for reneging, have underscored the agreement’s fragility.⁶ Besides dissolution of parliament, which would entail new elections, the National Accord does not indicate the steps to be taken if the coalition dissolves by other means—thus the penalties from withdrawing are unclear. While the comments by certain MPs and other community leaders suggest that elections will follow the coalition’s dissolution, this ambiguity works in favor of the withdrawing party. Rather than calling for immediate elections, the withdrawing party—especially if it is the PNU—could delay new elections until the end of the electoral term, a less costly penalty than immediate elections. Alternatively, the ODM or PNU could threaten immediate elections as a bargaining chip. A clearer understanding of the costs of reneging on the agreement could help the parties work through impasses.

While international and regional diplomatic pressure helped to bring the two sides to the negotiating table, employed in isolation, they are in a weak position to sanction and influence Kenya’s politicians over the National Accord. Even though a number of donor nations withdrew funding or threatened do so (the U.S. warned that it would not engage in "business as usual"⁷) their leverage is not absolute. Unlike many countries in Africa, only four percent of Kenya’s gross national income is derived from development assistance.⁸ Furthermore, Kenya’s high growth rate and location as a regional business hub make it an East African hegemon. Several countries depend on Kenya’s success. As the crisis unfolded, Tanzania experienced a decrease in cross-border trade, while Uganda and Rwanda experienced fuel shortages, as its imports remained blocked at the port of Mombassa. Furthermore, Uganda must now contend with 12,000 Kenyan refugees.⁹

Thus, pressure to remain in the power-sharing coalition cannot only come from regional leverage or the larger international community. The incentive to remain in the coalition must also come from the sources that pushed constitutional reform—the domestic constituencies.¹⁰ More concretely, the African Union must augment its panel with significant members of civil society organizations that can monitor and evaluate the political parties.

Notably, the Kenyan business community, having prospered under the Kibaki regime and thus holding a notable stake in maintaining peace, should be part of the domestic constituency recruited to help keep the agreement on track.¹¹ During the crisis, the business community lost $500 million a week.¹² In the first month of the crisis, nearly 50,000 jobs were lost. In fact, these losses were so significant that they propelled the business community to issue a joint statement supporting the mediation efforts.¹³ In this way, domestic pressure can be applied to help keep the power-sharing agreement together.

The armed domestic militia groups should be kept at bay. The sporadic, but continuing violence exacted by armed groups such as the Saboat Land Defense Forces (SLDF) and the Mungiki (to name two dominant groups) gives the impression that the peace dividends are not felt beyond parliament. Although these two groups formed for different reasons, their objectives are similar: to retaliate for real or perceived ethnic marginalization. The SLDF formed in response to land distribution plans that the government implemented; there are allegations that some members of government support the group. The Mungiki initially formed in response to former president Daniel arap Moi’s oppression of the Kikuyu and now claims to defend their interests in the Rift Valley. They commit notoriously gruesome acts of brutality and are known to levy taxes on public transportation operators and businesses as a means to raise funds. As with the SLDF, there are some indications that they have ties to sitting politicians.¹⁴

The government took aggressive steps to demobilize and disarm the SLDF in the weeks after the signing of the National Accord, while the Mungiki staged a 3,000-person demonstration in downtown Nairobi on March 6 to let people know they ‘still exist.’ Furthermore, on April 14, clashes involving the Mungiki resulted in four deaths.¹⁵ However, the government’s strategy in dealing with the SLDF was heavy handed—some civilian deaths were rumored to have occurred—and shrouded in mystery, as the media was barred from the area.¹⁶ It is important that the demobilization and disarmament processes must be transparent and not lend themselves to be manipulated as ethnic targeting. The government must also take steps to address the threat of Mungiki violence. Concurrently, the government, in partnership with civil society, must begin to tackle the underlying factors that give rise to these militia groups.

Conclusion

Rightly, the National Accord points to the need to address long-festering issues to engender peace. As such, durable peace will entail reintegrating the displaced, addressing land tenure concerns and confronting the human rights abuses that have taken place at the hands of government. But these important long-term steps cannot be adequately fulfilled if the agreement, which mandates cooperation and collaboration, is not strengthened. In the short term, the power-sharing agreement must be prevented from dissolving to ensure that the other important commissions will be successful. To hold the power-sharing arrangement together, the involvement of the African Union should be bolstered by reaching out to civil society organizations, the militia groups must cease operations and penalties for dissolving the coalition must be clearly established.