Electing the alliance of the accused : the success of the Jubilee Alliance in Kenya's Rift Valley

Journal of Eastern African Studies ISSN: 1753-1055 (Print) 1753-1063 (Online) Journal homepage: http://www.tandfonline.com/loi/rjea20 Electing the alliance of the accused : the success of the Jubilee Alliance in Kenya's Rift Valley Gabrielle Lynch To cite this article: Gabrielle Lynch (2014) Electing the alliance of the accused : the success of the Jubilee Alliance in Kenya's Rift Valley, Journal of Eastern African Studies, 8:1, 93-114, DOI: 10.1080/17531055.2013.844438 To link to this article: https://doi.org/10.1080/17531055.2013.844438 2013 The Author(s). Published by Taylor & Francis. Published online: 19 Nov 2013. Submit your article to this journal Article views: 2309 View related articles View Crossmark data Citing articles: 11 View citing articles Full Terms & Conditions of access and use can be found at http://www.tandfonline.com/action/journalinformation?journalcode=rjea20

Journal of Eastern African Studies, 2014 Vol. 8, No. 1, 93 114, http://dx.doi.org/10.1080/17531055.2013.844438 Electing the alliance of the accused : the success of the Jubilee Alliance in Kenya s Rift Valley Gabrielle Lynch* Depart of Politics and International Studies, University of Warwick, Coventry, UK (Received 5 July 2013; final version received 16 August 2013) Against a history of a divided Kalenjin/Kikuyu vote and election-related violence, and a contemporary context of high levels of inter-communal mistrust and intervention by the International Criminal Court (ICC), this article explains the Jubilee Alliance s success amongst Kalenjin and Kikuyu voters in the Rift Valley in the 2013 election. To do this, it examines the pre-election context, election results in Kalenjin- and Kikuyu-dominated areas, local political debates, and election campaigns to reveal how the Uhuruto team persuaded local residents to support this seemingly unlikely political marriage in all six elections. It is argued that the alliance used existing and emergent communal narratives of justice and competition to recast socio-economic and political debates in a way that persuaded the majority of Kalenjin and Kikuyu to support Jubilee and to vote against Raila Odinga and the Coalition for Reform and Democracy (CORD) as a way to protect and further their individual and collective interests. In making this argument, particular attention is given to relations between community members, and to popular support and investment in peace; negotiations between Uhuru and Ruto, and Kalenjin hosts and Kikuyu guests ; the reinterpretation of the ICC as a performance of injustice; and successful presentation of Uhuruto as a youthful team that could bring about peace and meaningful change as compared with an old, vengeful, incumbent Odinga Odinga. Keywords: Kenya; multi-party elections; democratization; ethnic politics; peace; justice; International Criminal Court (ICC) Introduction In the early hours of Saturday, 9 March 2013, Kenya s Independent Electoral and Boundaries Commission (IEBC) announced that Uhuru Kenyatta and William Ruto of the Jubilee Alliance had secured 50.03% of the popular vote in the presidential election clearing the new 50% plus one vote rule to win the first round by just 4099 votes. 1 As television and radio stations reported Uhuruto s victory, local residents took to the streets in towns across the Central and Rift Valley areas. 2 Many enthusiastically waved loaves of bread in celebration of how they now had the whole loaf and no longer had to share (as they argued had been the case in the coalition government, 2008 2013), while others held up dried fish and declared how Raila Odinga (from lakeside Nyanza) could keep his fish! Celebrations continued throughout the next day. Indeed, while many question the credibility of final results, it is clear that: the majority of Kikuyu and Kalenjin voted for the Jubilee Alliance, were happy with the *Email: G.Lynch@warwick.ac.uk 2013 The Author(s). Published by Taylor & Francis. This is an Open Access article distributed under the terms of the Creative Commons Attribution-NonCommercial-NoDerivatives License (http://creativecommons.org/licenses/by-nc-nd/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited, and is not altered, transformed, or built upon in any way.

94 G. Lynch results, and regard the elections as free and fair. Thus, while the IEBC reported an impressive average voter turnout of 86%, this rises to 89.58% for the Kalenjin- and Kikuyu-dominated constituencies of the Rift Valley. Moreover, while these figures may be inflated, the majority of Kalenjin and Kikuyu did for vote for Uhuruto in the presidential election; Jubilee s constituent parties Kenyatta s The National Alliance (TNA) and Ruto s United Republican Party (URP) also winning the majority of senate, governor, national assembly, women s representative, and councillor seats in Kikuyu- and Kalenjin-dominated areas. Finally, while 72.2% of Kenyans regard the election as free and fair according to a May 2013 opinion poll, this figure rises to 95.2% and 94.1% for Kikuyu and Kalenjin respondents respectively one of the most common reasons given being that there was no violence. 3 Support for Jubilee was particularly striking in the Rift Valley where competition with Kikuyu has been central to the formation and salience of an inclusive Kalenjin identity, 4 and where inter-communal relations are strained by a history of election-related violence that dates back to the early 1990s. The area was also the epicentre of post-election violence (PEV) in 2007 2008, which led to the death of over 1000 Kenyans and displacement of almost 700,000 others, 5 and is the geographic focus of two cases at the International Criminal Court (ICC) in which Kenyatta and Ruto face separate charges of crimes against humanity for their alleged role in organizing violence against each other s support bases and communities. Moreover, while the majority of Kalenjin and Kikuyu hoped for the best in 2013, in certain areas, some had prepared for the worst. Given this context, some may be tempted to see Jubilee s success as evidence that the election was little more than an ethnic census, with leading community spokesman carrying ethnic constituencies in and out of political alliances. 6 In turn, the lack of violence could be taken as evidence of how previous periods of election-related violence resulted from elite-level incitement and organization alone. However, this paper argues that Kenyatta and Ruto did not and could not just tell their supporters how to vote or behave, but instead had to persuade them. In so doing, the analysis draws upon an understanding of ethnic communities as moral and historical communities and a widespread perception of Kenyan politics as ethnic to show how communal narratives are negotiated and can be recast in ways that help shape perceptions of individual and collective interests in an ever-changing political arena producing complex and contested groupings that enjoy greater relevance to local actors. 7 More specifically, the paper argues that the Jubilee Alliance ran a well-funded and coordinated campaign that used an array of strategies both analytical and performative to effectively reframe overarching narratives, which drew upon, and helped shape local understandings of justice, injustice, opportunity and threat. Just as in 2007, strongly ethnicized discourses were interwoven with narratives of past and potential injustice in ways that led to the vilification [of one candidate] and [presentation of the other] as the vehicle of meaningful change among members of particular ethnic groups. 8 As a result, the election was once again as much about what one might call exclusionary ethnicity and who would not get power and control the state s resources, 9 as it was about speculative loyalty, or calculation of the potential advantages of electing community spokesmen. 10 The article begins with a discussion of the obstacles to Jubilee s success amongst Kalenjin and Kikuyu voters in the Rift Valley, and strong desire for peace. It then outlines local voting patterns, and highlights how a power-sharing agreement was reached between Kenyatta and Ruto, and the Kalenjin and Kikuyu communities; before moving on to analyse how Jubilee reframed the ICC story at least in the eyes of a significant

Journal of Eastern African Studies 95 number of Kenyans as a performance of injustice, neo-colonialism, and threat to the country s sovereignty, peace and stability. Finally, it analyses the effective juxtaposition of Uhuruto as a young team that could bring about peace and meaningful change, with Odinga as the candidate of continuity rather than reform; messages which for various reasons resonated with local constituents. The analysis focuses on Kalenjin and Kikuyu in the Rift Valley, but has broader relevance for explaining support for the Jubilee Alliance in other parts of Kenya, as well as for furthering our understanding of multiparty politics, the often unintended consequences of external intervention, the political salience and role of ethnic identities, and importance of political campaigns, narratives and performance. The article draws upon extensive fieldwork in the Rift Valley in 2008, 2009, 2011, 2012 and 2013 (this includes work in Nakuru, Baringo, Kericho, Bomet, Nandi, Uasin Gishu and Trans Nzoia counties), and trips to Mt Elgon District in 2011 and 2012, and former Central Province in 2011. Overall, the author conducted more than 200 interviews with politicians, civil servants, religious leaders, civil society activists and ordinary citizens, and a number of focus group discussions. Evidence was also drawn from participant observation of Truth, Justice and Reconciliation Commission (TJRC) hearings, peace meetings and political rallies, as well as from secondary literature, government and non-government reports, media reports and social media sites. However, due to the sensitivity of the subject matter, limited use is made of direct quotes, while all interviewees are anonymized with the interview location sometimes also omitted. Finally, some of the research was conducted as part of an election-monitoring project (December 2012 June 2013). However, while the analysis has been greatly enriched by involvement in this collective endeavour, the arguments (and any mistakes) are the author s own. Our enemies over there : a political marriage of unlikely bedfellows [Between the] Kikuyu and Kalenjin there is no trust. Honestly. Yes, we can meet in a hall, talk, agree, and do things together, but there s no trust. 11 So we ve opted to come together, forget the past, and do the necessary so violence doesn t recur. 12 The Kalenjin make up around 14% of the total Kenyan population; 13 the signifier embracing a number of subgroups administered as separate tribes during the colonial period including the Nandi, Kipsigis, Tugen, Keiyo, Marakwet, Sabaot and Pokot. The Kalenjin constitute a majority in 40 out of 76 constituencies in the Rift Valley, 14 and also Mount Elgon constituency in Bungoma County in former Western Province. Overrepresented in the National Assembly, the Kalenjin enjoy a majority in seven of the country s 47 counties 15 administrative and political units that were formed after the inauguration of a new constitution in 2010. Historically, Kalenjin have tended to vote as a near-united bloc in multi-party elections; the majority supporting the Kenya African Democratic Union (KADU) in 1963, Kenya African National Union (KANU) in 1992, 1997 and 2002, and Orange Democratic Movement (ODM) in 2007. The Kikuyu are Kenya s largest ethnic group at around 17% of the total population. 16 The community constitutes a clear majority across former Central Province, and a majority in ten out of 76 constituencies, 17 and two out of 14 counties 18 in the Rift Valley. Up until 2013, Kikuyu consistently cast their vote for a party other than that supported by the majority of Kalenjin; namely KANU in 1963, FORD-Asili or the Democratic Party

96 G. Lynch (DP) in 1992, DP in 1997, the National Rainbow Coalition (NaRC) in 2002, and Party of National Unity (PNU) in 2007 all parties that fronted a Kikuyu presidential candidate. Against this background of divergent voting patterns, a number of factors suggested that Uhuruto would find it difficult to mobilize Kikuyu and Kalenjin voters across the Rift Valley in 2013. First, and with regards to Kalenjin voters, there are clear limits to Ruto s grip over the Rift Valley. In the run-up to the 2007 election, Ruto s late move to join Odinga in ODM was in response, at least in part, to pressure from community members most notably Nandi elders who felt that Odinga was best placed to oust President Mwai Kibaki. The importance of this directive highlights the nature of local political support, which is largely based on perceptions of who can best protect and further individual and collective interests, rather than on a blind following of ethnic spokesmen. This fact was also evident during the URP party primaries in January 2013, when a number of Ruto s close allies who he had publicly endorsed proved unsuccessful in their bid for the party ticket. This was particularly glaring in his home area of Uasin Gishu; a common conclusion being that here [people] can see [and] know what he s capable of doing and has done. 19 Defeated aspirants included Isaac Maiyo, the Constituency Development Fund (CDF) manager for Ruto s Eldoret North constituency and parliamentary campaign manager in 2007, who failed to win the parliamentary ticket for Turbo constituency; the incumbent MP for Eldoret South, Peris Simam, who failed in her bid to stand in the newly created Kesses constituency; and Dr Julius Bitok who had hoped to stand for governor of Uasin Gishu County. Significantly, all of these individuals were widely believed to have misused funds under their control in the case of Maiyo and Simam this related to their use of CDF, and for Bitok of Moi University finances their failure thus highlighted the limits of top-down endorsement in the instance of candidates who lacked legitimacy and support at the local level, and popular rejection of corruption. Second, while Ruto emerged from the 2007 election as the most prominent Kalenjin politician, and managed to mobilize an impressive level of support for a vote against the new constitution in 2010, various legal processes raised questions regarding his integrity and suitability for political office. The most glaring was an announcement in December 2010 that he was under investigation followed by the confirmation of charges against him in January 2012 by the ICC for his alleged role as an indirect co-perpetrator of the crimes against humanity of murder, forcible transfer, and persecution in parts of Uasin Gishu and Nandi counties during the PEV of 2007 2008. This is critical since the role of political incitement in fuelling inter-ethnic violence since the early 1990s is widely recognized including by Kalenjin in the Rift Valley, many of whom have suffered directly or indirectly as a result. 20 Ruto s legal battles also extended to the national courts. In October 2010 he was suspended as Minister of Higher Education after the High Court ruled that he should face trial for his alleged role in the fraudulent sale of forestland. Ruto was acquitted in April 2011 for lack of evidence, but there were rumours that he had bribed, intimidated, and perhaps even killed key witnesses. While, at the time of the election, Ruto was involved in another court case in which he was alleged to have grabbed a 100-acre plot from an internally displaced person (IDP) following the 2007 2008 PEV. Third, while Ruto parted ways with Odinga soon after the 2007 election, a number of prominent Kalenjin politicians remained in ODM. These included Henry Kosgey, MP in Nandi since 1979, ODM chairman, and outgoing industrialization minister; Sally Kosgei, a former Cabinet Secretary and MP for Aldai (2007 2013); and Franklin Bett, a former State House Comptroller and MP for Bureti (2007 2013). A number of prominent

Journal of Eastern African Studies 97 Kalenjin voices also joined the ODM camp in the run-up to the 2013 election. This included Joshua arap Sang, a Kalenjin-language radio presenter and Ruto s co-accused in Case 1 at the ICC. Moreover, Odinga still enjoyed pockets of support in the inter-election period although this waned in the face of the Jubilee campaign. Indeed, according to an Ipsos Synovate opinion poll, as late as September 2012, ODM was the single most popular party in the Rift Valley with 26%, as compared with TNA s 25% and URP s 19%. In addition, 26% of Rift Valley respondents said that they would vote for Odinga, as compared with 28% for Kenyatta and 18% for Ruto. 21 These data were supported by interviews in December 2012, in which many Kalenjin argued that if Ruto allied with Kenyatta he would garner around 60 75% of the Kalenjin vote; Odinga then expected to pick up the remainder. Sub-Kalenjin group identities, and generational and class divisions also rendered it difficult to mobilize the Kalenjin as a bloc. Odinga seemed to enjoy more support among well-educated Kalenjin who were worried about the consequences of electing ICC indictees, and older members of the Kalenjin community who were more concerned about the possibility of another Kikuyu president (after Kenyatta and Kibaki, 2002 2013). As one member of the influential Kalenjin Council of Elders explained, The elderly have experience and have seen things, [they] can see that [someone] can t wipe out animosity between Kikuyu and Kalenjin overnight because this unity is purely The Hague. 22 It was not just pockets of elderly Kalenjin who expressed a strong sense of unease when Ruto entered an alliance with and then stood as the running mate to Kenyatta. Opposition drew upon the fact that Kenyatta was Kikuyu and the son of President Jomo Kenyatta. Since the formation of an inclusive Kalenjin alliance in the late colonial period, the Kikuyu have been the most significant other against which the Kalenjin have positioned themselves. Many have feared the Kikuyu and resented the perceived encroachment of Kikuyu onto land that many Kalenjin believe they were displaced from during the colonial period, and never adequately compensated for. This sense of historical injustice was reinforced by a perception that Kikuyu benefited disproportionately from settlement schemes in the 1960s and 1970s under President Kenyatta s tutelage 23 with Kenyatta s son cast as unlikely to provide redress. These narratives of competition had gained new layers with the return of multi-party politics from 1992, when political opposition was cast as Kikuyu-dominated and committed to ousting President Moi (a Kalenjin from the Tugen subgroup). There was organized violence against Kikuyu in parts of the Rift Valley in the 1990s as politicians fostered Kalenjin fears of a loss of power and future marginalization, and played upon the opportunity for Kalenjin to access high quality agricultural land in areas that were regarded as part of the community s ancestral homelands. 24 Anti-Kikuyuism reached a climax during and immediately after the 2007 election campaigns, when a strongly ethnicized discourse of persecution and bias [against the Kalenjin by a Kikuyu government ] was interwoven with narratives of past and potential injustice. 25 Then, when violence broke out, many Kalenjin demonstrated against perceived electoral malpractice, but also engaged in targeted attacks against Kikuyu neighbours to protest against Kikuyu support for Kibaki and thus a distant political hegemony, and the failure of local Kikuyu to act like guests ; 26 but also to displace outsiders from land that they felt was rightfully theirs. The latter evidenced by local narratives, the locations of violence, 27 and widespread attacks on Kisii residents who were not seen as part of the political hegemony in the same way, and had divided their vote between PNU and ODM,

98 G. Lynch but who lived on prime agricultural land in disputed settlement schemes and border areas. 28 The 2007 election and its aftermath reinforced a sense of competition with Kikuyu who were cast as incapable of voting for non-kikuyu. Targeted attacks by Kalenjin against Kikuyu in the Rift Valley together with police shootings of Kalenjin and targeted attacks by Kikuyu against Kalenjin damaged inter-personal and intercommunal relations, and fuelled a strong sense of hatred and mistrust at the local level. Moreover, it was clear in 2013 that, while a new constitution and significant institutional reform had brought substantive change, many of the underlying issues that had fuelled the violence (most notably the land issue) had not been addressed. 29 Little progress had been made in security sector reform meaning that the very same policing structures blamed by many for serious human rights violations during the PEV remained in place. 30 As a result, the risk of political violence was widely deemed to be unacceptably high. 31 This context made selling support for another Kikuyu president in 2013 a difficult task. One problem was that the Kikuyu again seemed unwilling to relinquish power and instead seemed to want to use the Kalenjin vote; support that could later be betrayed. Thus, when Sang moved from URP to ODM the week before the election, he cautioned Kalenjin that: some of our people want us to enter into [a memorandum of understanding (MoU)] again through Jubilee. They [the Kikuyu] never respected the MoU signed [with Odinga] in 2002 and the power sharing deal agreed on in 2007. Why should we expect that the one that our community has signed in Jubilee will be respected? 32 An Njoro businessman cautioned that the Kalenjin know that they are supposed to be used by the Kikuyu and, if these people get in, they ll be left out, and know they ll feel short changed. 33 New grievances also emerged, including a common perception of significantly preferential treatment for IDPs from Kibaki s Kikuyu community, 34 which reinforced narratives of favouritism. 35 Finally, there was opposition to the particular candidacy of Kenyatta as the privileged son of the former president who was unlikely to address key issues such as land. As one interviewee summarized, The majority [of Kalenjin] don t like Uhuru they want change Most of the people [ask: Kenyatta was] born in State House, schooled in State House and still wants to stay in State House? 36 Similarly, there are a number of reasons why Kikuyu in the Rift Valley might have been reluctant to vote for Uhuruto. First, there were a number of other Kikuyu presidential candidates including Peter Kenneth and Martha Karua who were seen as development-conscious candidates by many ordinary people. Second, a number of interviewees expressed a dissatisfaction with previous Kikuyu presidents from Central Province who they felt had used Kikuyu in the Rift Valley to obtain power, but then ignored them. Thus, one interviewee explained how: Kikuyu in Rift Valley feel a bit aggrieved by Kikuyu in Central as we ve never had a [Permanent Secretary, Provincial Commissioner], or anyone senior from Kikuyu in Rift Valley Problem is we vote for them, but they don t consider us although we suffered. 37 This ties to a related issue; namely, the ongoing, and highly visible plight of Kikuyu IDPs many of whom despite claims of preferential treatment noted above remained in old and battered tents on small scraps of land awaiting resettlement five years after their initial displacement.

Journal of Eastern African Studies 99 Third, while many Kikuyu with whom the author spoke did not think that Kenyatta should be facing charges at The Hague for his alleged role as an indirect co-perpetrator of crimes against humanity against ODM supporters in Nakuru and Naivasha towns in 2008, many did think that Ruto was responsible for inciting and organizing attacks against Kikuyu. As one interviewee queried: Kikuyu regard Ruto as the mastermind of violence in the Rift Valley, and if Kenyatta instigated in response, how can they become friends when it s Ruto who caused Uhuru to go to The Hague. So question why [are they] together, what [do they] want, what [do they have] in common? 38 The significance of such questions was heightened by the fact that, while some steps had been taken to establish a special division of the High Court to deal with international crimes committed during the PEV, there had been little progress in the investigation and prosecution of any perpetrators by 4 March 2013. 39 This meant that many known perpetrators were still walking around scot free, while most stolen livestock and household belongings had not been returned, and no apology had usually been offered. In this vein, one peace activist relayed a conversation with a Kikuyu woman who complained that: [my] neighbour still has my cow that he stole three years ago. He burnt my property and killed my husband. What peace are you talking about? In the peace forum we re together and he doesn t confess! 40 Together with a history of violence and ethnic stereotypes, e.g. of the Kalenjin as lazy and Kikuyu as hardworking, 41 this lack of justice and daily injustice (as perpetrators offered no apology and continued to benefit from the spoils of violence) helped fuel a strong sense of anger and mistrust between Kikuyu and Kalenjin at the local level. This feeling was heightened by the ineffectiveness of extensive peace-building work following ethnic clashes of 1991 to 1993, and by the scale, scope, and seemingly organized nature of violence in 2007 2008, which often involved neighbours, friends, and even family members. These memories reduced people s faith in contemporary peace-building activities, and fuelled a perception of the Kalenjin as secretive and sly who can agree [one thing] and do another. 42 As one man cautioned: Something you won t know the Kalenjin are hypocrites you can be with them and they can be planning something. 43 In turn, while the majority of Rift Valley Kikuyu hoped for the best in March 2013, some prepared for the worst by moving themselves, their families, and/or household belongings some even forming village or estate-level community defence units in anticipation of new election-related violence. For example, in Njoro constituency, Nakuru County, in January 2013, residents spoke of how many displaced from the Kikuyu/ Kalenjin cut line, which separated the Kikuyu-majority Ndeffo area from the Kalenjindominated Mauche area, had returned, but continued to sleep in local trading centres and only went to their plots during the day to farm. They also reported that some of those who had moved back were beginning to move their families and/or belongings away from the border area in fear of further election-related violence the area having suffered violence in 1991, 1992, 1997, 2002, 2005, 2007 and 2008. In another area, Kikuyu elders spoke openly about how they had vowed [not to] flee again and were going to protect themselves at all costs by organizing themselves into structures like village defence units. 44 Such local-level discussions were particularly concerning given a reported proliferation of small arms in such former hotspots. 45 In many parts of the Rift Valley, relatively small issues often strained tense intercommunal relations to near breaking point. Examples include Chemusian Farm in Burnt Forest, Uasin Gishu County, where local Kalenjin protested against the resettlement of Kikuyu IDPs in 2011 until it was agreed that some local squatters and Mau Forest

100 G. Lynch evictees would also be resettled. Or Banita settlement scheme in Rongai, Nakuru County, where disputes over land made it impossible for Kikuyu to bury their dead without a strong (and armed) provincial administrative presence from early 2012, and which then required graves to be concreted over lest the bodies be dug up by neighbouring Kalenjin. 46 In this context, there was a fear prior to the election among both Kalenjin and Kikuyu that if the Jubilee Alliance collapsed this would reinforce local narratives of difference and trigger even worse violence than that experienced in 2007 2008. This reality of surface, 47 cosmetic 48 or negative peace 49 is of critical importance since it was to a large extent a lack of trust and widespread desire to avoid further violence (with some areas having experienced violence in every multi-party election to date) that led many Kalenjin and Kikuyu in the Rift Valley to support this seemingly unlikely political alliance. As two interviewees from different parts of the Rift Valley noted: In Nakuru, Kikuyus have a very hard time thinking that they can work closely with Ruto. Common people don t understand the alliance, but many say that, if it s for us not to fight, then let it be as the truth will always come out. Main thing is to see to a peaceful election. 50 Kikuyu in Uasin Gishu [are] fearing a reoccurrence of violence, so if [there s] a way to come together with Kalenjin then the better what minding is peace in the next election. Whatever happened in past five years, not that can t recall losses but tend to forget because the much they remember the more they don t heal. 51 In this context, Jubilee offered exactly what many people wanted a way for the Kikuyu and Kalenjin to come together in the interests of peace. From orange to red: the strength of the Jubilee wave across the Rift Valley We don t want to see a situation where we will go for a second round. That is why you must vote for Uhuru Kenyatta and his entire line up from top to bottom Nandi is the nerve centre of the Kalenjin politics. You should be careful not to embarrass me when you cast your vote. 52 The Jubilee Alliance and its constituent parties recorded impressive results across the Kalenjin- and Kikuyu-dominated areas of the Rift Valley. Of 41 constituencies in which the Kalenjin constitute a majority, Kenyatta officially secured 90% of the vote or more in 20, 80 89% in eight, 70 79% in nine, 60 69% in three, and 50 59% in one. 53 Of the ten constituencies in which the Kikuyu constitute a majority, Kenyatta officially secured over 90% of the vote in five, 80 89% in three, and 70 79% in another three. 54 Some of these figures may be artificially inflated due to local-level malpractice, but it is clear that of the Kalenjin and Kikuyu in the Rift Valley who voted, the majority cast their vote for Kenyatta. Much of the non-jubilee vote can be explained by the presence of other communities. The exception was the four West Pokot constituencies, where the majority of constituents hail from the Kalenjin Pokot sub-group and where Kenyatta secured 73.33% of the vote as compared with Odinga s 22.95% an anomaly to which this article will return. The Jubilee wave was also evident across the other five elections for governor, senator, national assembly representative (MP), women s representative, and member of the county assembly (MCA). URP secured the women s representative in all six Kalenjin-majority counties, the governor in five, and senator in four. Moreover, in the three instances where

Journal of Eastern African Studies 101 URP was not successful, county-level positions were won by KANU candidates. In the parliamentary elections, URP secured 35 of 41 Kalenjin-majority constituencies, KANU four, and independent candidates two. In turn, TNA candidates secured the women s representative, governor, and senator seats in both Kikuyu-majority counties (Nakuru and Laikipia) and MP in nine out of ten Kikuyu-majority constituencies the exception being the election of Raymond Moi (the son of former President Moi) on a KANU ticket in Rongai. In contrast, prominent Kalenjin politicians who vied on an ODM ticket all lost to URP candidates. This included Henry Kosgey who was beaten in the Nandi senator race by a young lawyer, Stephen Sang; Professor Margaret Kamar who lost her bid to become Uasin Gishu governor to another new candidate, Jackson Mandago; Sally Kosgei who was also ousted by a political entrant, Cornelly Serem, in Aldai; and Margerer Langat who was overtaken by URP s Eric Rop in Kipkelion West. Franklin Bett, the then MP for Bureti, pulled out of competitive politics before the party primaries even took place, opting instead to chair the ODM National Election Board. With regards to anomalies in the Kalenjin-majority areas, the most striking is West Pokot where KANU secured the senator (John Lonyangapuo), governor (Simon Kielei), and MP for Kapenguria (Samuel Moroto). In all three instances, candidates benefited from the popularity and wealth of Lonyangapuo who was credited with initiating various development projects and who refrained from openly campaigning for any particular presidential candidate, instead advising constituents to observe and see which party is going to form the next government. 55 However, they also benefited from the poor performance of incumbent politicians, and peripheral position of Pokot to the Kalenjin core. 56 Moreover, while Odinga mobilized a larger Kalenjin vote here than in any other area, Uhuruto still enjoyed a majority with 73.33%. In the Baringo County senator race, Gideon Moi (the son of the former president and KANU chairman) benefited from his substantial resources and family connections, but also from the charges of corruption laid against his main competitor Bishop Jackson Kosgey of URP. 57 Kosgey s defeat given his close relations with Ruto who he had acted as a witness for during the ICC confirmation of charges hearings, and who had openly campaigned for him underscores, once again, how Ruto could not simply direct voting decisions at the local level. However, since Gideon Moi and several of his successful KANU co-members namely Hellen Sambilli in Mogotio, Eric Keter in Belgut, and Moses Kipkemboi in Kuresoi North constituencies openly campaigned for Uhuruto, it meant that a vote for KANU was not necessarily a vote against Jubilee. Kenyatta gained 87.93% of the announced presidential vote in Baringo County, 92.62% in Mogotio, 90.73% in Belgut, and 89.41% in Kuresoi North. This underscores KANU s image as a Jubilee-friendly party, and the importance of performance records and politics of persuasion, rather than mere direction. In short, all four candidates beat relatively unpopular URP candidates; the Kuresoi North URP candidate, for example, allegedly given a direct ticket through rigged nominations. 58 In Cherangany, an independent candidate, Wesley Korir, beat Joshua Kutuny, another of Ruto s close allies. However, Kutuny enjoyed little support among the sizeable non- Kalenjin community (mainly Luhya) and had gained a reputation among the slight Kalenjin majority for arrogance and disrespect. 59 However, once again, the majority of local Kalenjin likely voted for Uhuruto ; Jubilee securing 54.52% of the vote in the local presidential election.

102 G. Lynch In Mt Elgon, the former MP and independent candidate, John Serut, beat the incumbent MP and URP candidate, Fred Kapondi. From interviews it is clear that Kapondi was disliked by those who had suffered at the hands of the Sabaot Land Defence Force (SLDF), which had terrorized local residents between 2005 and 2008, as the person widely believed to have organized the SLDF. He was also disliked by former SLDF members who felt that he had abandoned them once elected. It was generally felt that he had stolen the URP ticket. In contrast, Serut was talked about as more development-conscious and peaceful. Again, Serut openly campaigned for Kenyatta and against Odinga who gained 66.77% and 18.14% of the vote respectively with much of the non-jubilee vote potentially explained by the presence of non-kalenjin (mainly Luhya and Teso). With regards to the non-tna candidate who won in a Kikuyu-majority constituency Raymond Moi in Rongai this can be seen as an indirect testament to the strength of the Jubilee Alliance. It was decided before the election that Rongai where the Kikuyu enjoy a slight majority of around 51% would be considered a Kalenjin constituency for the purposes of maintaining Kalenjin/Kikuyu relations and peace. However, the local URP candidate and incumbent MP, Luka Kigen, was unpopular and had allegedly incited local Kalenjin against Kikuyu, for example, in Banita settlement scheme (mentioned above). In this context, Raymond Moi asked people to vote for individuals rather than parties, and to shun leaders interested in manipulating the ethnic composition of the constituency in favour of one tribe. 60 In turn, it seems that many local Kikuyu and likely some local Kalenjin voted for peace through the rejection of a man accused of incitement. In turn, Uhuruto won 77.15% of the presidential vote; the non-jubilee vote once again likely resulting, at least in part, from the constituency s cosmopolitan make-up. The limits to the Jubilee wave did thus not reflect high levels of support for CORD and its allied parties, which failed to win any of the big seats. However, this is not to say that Odinga did not enjoy any support. Indeed, according to one exit poll, 11% and 4% of Kalenjin and Kikuyu respectively, reported voting for Odinga nationally, as compared with 74% and 83% for Kenyatta, and 4% and 3% for other candidates with 12% and 10% refusing to answer. 61 Nevertheless, none of the non-jubilee candidates elected as senators, governors, and MPs in the Rift Valley enjoyed close relations with Odinga or CORD, and all either called upon their constituents to vote for Uhuruto, or left the presidential vote open to individual choice. The success of some non-jubilee candidates instead reflects local dynamics, which KANU and some independent candidates proved well placed to exploit, and the need to persuade, rather than simply tell, people how to vote. Finally, in contrast to the Coast where It might be argued that, for those who were drawn into voting by the patronage of an aspiring MP or governor, the presidential choice was actually a secondary choice, 62 it seems that support for URP and TNA in the Rift Valley followed, in large part, from the primary choice of Uhuruto. Kenyatta and his supporters learnt an important lesson from 2007, when Kibaki won the presidential election but PNU fared relatively badly in the parliamentary elections, 63 while Ruto explicitly called for six-piece voting to ensure that he would not be embarrassed and would instead be respected as the number two in a future Jubilee government. Negotiating an alliance between locals and guests The Kalenjin have a saying ngiwe kap kipamchii iam chii If you go to a home that eats flesh or human, eat with them. Is not to say that eat flesh, but is to emphasise the need to do

Journal of Eastern African Studies 103 as the visitor does for harmony Ordinary Kalenjin will always say we like visitors, but if people are proud or arrogant even fellow Kalenjin they will be ostracised. 64 When [the church was] doing peace-building the group identified two causes of violence in the region: politics and land. So decided to address on politics Kikuyu said we won t propose Kikuyu for MP in this area, you Kalenjin propose, and we vote. 65 As Sarah Jenkins has argued, Immigrants who are seen to be on the right side of the political divide [in the Rift Valley] are [generally] recognized as welcome guests and their participation in politics is largely accepted. 66 Local residents are keenly aware of this dynamic and through a series of intra and inter-community peace meetings facilitated by faith-based organizations an idea was born to zone the Rift Valley into areas where the Kalenjin or Kikuyu would take the lead in electing political officers. Through these discussions, and the secondary involvement of politicians, it was agreed at least at a mid to upper level that TNA would not field candidates throughout the Kalenjinmajority counties of Bomet, Kericho, Baringo, Elgeyo-Marakwet, Nandi, Uasin Gishu, and West Pokot. Through further discussions, it was agreed in the interests of Kalenjin and Kikuyu political unity and thus peace that Rongai and Kuresoi South constituencies in Nakuru County, where the Kikuyu enjoy a slight majority, would be regarded as Kalenjin constituencies. The basic rationale was that, otherwise, Kalenjin would only be able to elect one representative across the whole county (beyond MCAs) namely an MP in Kuresoi North although they regard Nakuru as part of their sphere of influence ; an outcome, which many felt, could severely test Kalenjin/Kikuyu relations. This local-level agreement was then taken to, and accepted by, Kenyatta and Ruto and their respective parties. However, it initially faced opposition from many local-level politicians whom it effectively disbarred from standing 67 and ordinary citizens, who felt that the decision had been made in the big hotels in towns with insufficient popular consultation. 68 Consequently, this agreement had to be sold to the voting public. This was done through the mobilization of voters against the ICC and effective juxtaposition of Uhuruto with Odinga (discussed below), but also through reference to the mutual benefits of mass Kikuyu and Kalenjin support for Jubilee. Thus, TNA pointed to how Ruto had given up his presidential ambitions and through so doing had made a Jubilee victory a feasible outcome and rendered violence unlikely; while Ruto and his allies emphasized how they had got the best possible deal. More specifically, URP underlined how Ruto would be number two, how the party would get half of all the posts, and how the Rift would remain peaceful and thus be able to prosper. One interviewee described how, at a rally in Baringo, Ruto had talked about his negotiations with Kenyatta in which he had been very crafty and negotiated half/half, and of how he would be a second president, which the gentleman felt had been effective in persuading Kalenjin that the place to be is in Jubilee. The same interviewee went on to note how, for Kikuyus, [they ve] suffered so much so a welcome relief, while they also see that Ruto has discarded his ambitions and that ready to support so Kikuyus see an opportunity to get back into leadership. 69 Significantly, such arguments as with those regarding the ICC, and the antithetical presentation of Uhuruto and Odinga discussed below were made through an array of campaign methods, which began long before the official campaign period. This included old tropes and performative displays of branded T-shirts and hats, paid-up adverts, political rallies, and mobile loud speakers. But it also expanded to include new, or at least extended tactics that rendered the campaigns more pervasive and far less tangible. In

104 G. Lynch short, more journalists as well as facebookers and tweeters were brought onto the waged employment of leading politicians. Politicians and their political allies and loyalists also made greater use of televised debates, opinion pieces in the leading dailies, radio talk shows, online blogs, twitter feeds, and Facebook pages to mobilize support for particular candidates, but also to engage in direct and indirect attacks against leading opponents and associated enemies. The ICC factor: a performance of justice or injustice? Luos the one who started the clashes here but not touched, so if started and not on list. So people are very bitter people see a hidden agenda, see someone that pushing that agenda. 70 Personally annoyed with Ocampo, people listed are very innocent; when this thing began it was a reaction against the government If it was Kibaki and Raila fine, but don t go and arrest foot soldiers Feeling that US government through Obama influenced Ocampo not to arrest anyone from that [Luo] community Everyone was expecting Anyang 71 or Raila and then go and pick someone like Sang! 72 During the PEV, the Kenya National Commission of Human Rights (KNCHR) began to investigate the scale, form, and causes of violence, a task that was later taken up by the Commission of Inquiry into Post-Election Violence (or Waki Commission). In the context of a long history of commissions of inquiry whose recommendations have been ignored and after deciding that much of the violence was organized and funded, and that Kenya suffered from a culture of impunity the Waki Commission recommended the establishment of a Special Tribunal to investigate and prosecute those most responsible. Failing that, the Commission recommended that information be forwarded to the ICC, and that the Office of the Prosecutor (OTP) be requested to analyze the seriousness of the information received with a view to proceeding with an investigation and prosecuting such suspected persons. 73 The Kenyan parliament failed to establish a Special Tribunal and after various delays to allow the Kenyan government more time to do so, and some further investigations in December 2010, the then ICC prosecutor, Luis Moreno Ocampo, announced the names of six Kenyans under investigation. The Ocampo Six were Ruto, Sang, and Henry Kosgey in Case 1, and Kenyatta, Francis Muthaura (the head of the civil service), and Mohammed Hussein Ali (former police chief) in Case 2. 74 Confirmation of charges hearings were held in September 2011 with charges confirmed against Ruto, Sang, Kenyatta, and Muthaura in January 2012. Then in March 2013 and a few days after Kenyatta and Ruto were announced president and vice-president elect charges against Muthaura were dropped, the OTP citing the death, bribery, and intimidation of key witnesses. 75 To understand how the Jubilee Alliance turned the heavy burden of ICC charges into part of a winning campaign strategy, this article argues that one must recognize how, among other things, legal proceedings are meant to constitute a performance of power, righteousness, and justice. Writing of the Nuremberg trials, Lawrence Douglas notes how: The trial was understood as an exercise in the reconstitution of the law, an act staged not simply to punish extreme cases but to demonstrate visibly the power of the law to submit the most horrific outrages to its sober ministrations. In this regard, the trial was to serve as a spectacle of legality, making visible both the crimes of the Germans and the sweeping neutral authority of the rule of law. 76

Journal of Eastern African Studies 105 Ocampo clearly saw the ICC s role in a similar fashion. Thus, in the context of a history of election-related violence, and concurrent post-election crises in Zimbabwe and Cote d Ivoire, Ocampo emphasized the need to tackle a culture of impunity and to make an example of Kenya to show how seriously the international community took such outbreaks of violence. However, instead of providing a clear example of international righteousness and justice, Jubilee managed to reframe the ICC story at least in the eyes of a significant number of Kenyans into a performance of injustice, neo-colonialism, and threat to the country s sovereignty, peace, and stability. Jubilee argued that the cases were marred by: poor investigations; a failure to confront those most responsible because of political influence, bias, vested interests and a misunderstanding of what had occurred; and an over-emphasis on punitive justice to the neglect of peace and reconciliation. Jubilee loyalists focused on the ICC s fairly limited investigations, which particularly in the initial phases were heavily reliant on the work of the KNCHR and Waki Commission. These earlier investigations were then presented as limited and biased; a perception that found particular resonance amongst Kalenjin respondents who frequently complained of how the Waki Commission had just passed along the highway 77 and been hijacked by a clique of Nairobi-based human rights activists. 78 The basic argument was that, the ICC [has] not done any justice through Ocampo because he uses evidence collected by the Waki Commission and the KNCHR Ocampo didn t come here and [his] people didn t collect our views Ocampo himself doesn t know Kenya! 79 Questions were raised about whether the ICC had focused on those most responsible, given that the 2007 presidential election pitted Kibaki against Odinga. The argument was that it was Odinga who had rejected the results and called upon people to demonstrate, while Kalenjin youth [had] fought in defence of Raila s stolen vote. 80 According to this logic, Ruto was a mere foot soldier, 81 while Sang was a little fellow from the radio station in a context where the other vernacular radio stations were [just] as bad. 82 Similarly, it was argued that, Kenyatta was not contesting the presidency in 2007, while it was PNU against ODM, not KANU (which Kenyatta then headed) against ODM. 83 In contrast, Kibaki had been the head of state, and if meetings were held at State House in January 2008 to organize revenge attacks as Ocampo claimed the president must have known about them. Uhuruto s troubles at the ICC were also cast as political an argument that had multiple strands. At one level, the ICC was cast as a political court that wanted to use the Kenyan cases to make itself legitimate as a meaningful global institution. 84 In this vein, much was made of the Court s limited achievements and prior criticisms faced; of how the cases took ODM/PNU in equal number, which looked very political and calculated ; 85 and the undeniable fact that all of the ICC s cases are in Africa, which was interpreted locally to mean that the ICC was a neo-imperial body that sought to prosecute Africans but not whites. 86 The last point was used, fairly successfully, to counter-shame the ICC as a neo-colonial weapon wielded by the West to punish Africa as a whole. 87 Moreover, while Odinga had called for a Special Tribunal, the fact that charges were then laid against his main political opponents, together with his pronounced support for the ICC process, facilitated a presentation of him as: desperate to remove both Uhuru and Ruto from the [2013] presidential race [and as someone who] had always prayed for the success of the ICC process; not for justice to prevail, but as a means of sorting out his political opponents. 88