Andrew Linke- University of Colorado
Kenyan election violence is not new, as academic and policy-oriented observers recognize. The nuances of each bout of electoral conflict reveal – and are informed by – important social trends in the country. Political contexts at the national level, and among localities within regions of the country, determine the realities of conflict that accompanies electoral contests.
The conflict surrounding Kenya’s last election caught many by surprise. Almost immediately following the December 27th 2007 poll, a wave of violence swept across Kenya against Kikuyu, the perceived supporters of President Kibaki’s Party of National Unity. Violence erupted initially in Kalenjin areas of the Rift Valley, where Eldoret North MP William Ruto, who campaigned on the Orange Democratic Movement (ODM) ticket, is now accused of inciting unrest. The attacks spread quickly to Kisumu, and other Luo-dominated areas where Raila Odinga carries the majority of the population’s support. After an initial episode of attacks targeting the Kikuyu community, Nakuru and Naivasha regions witnessed brutal attacks by Kikuyu militias against Luos and members of other opposition (ODM) communities. By February 22nd Odinga and Kibaki had reached a power-sharing agreement, under the supervision of mediator Kofi Annan. Approximately 1,300 people had been killed as the conflict finally diffused, and nearly 500,000 were displaced from their homes.
Following the 2007 election, the level of insecurity was so severe that it led to a new constitution. This was approved by referendum in August 2010. Although prior efforts to reform the constitution (e.g. the Bomas draft in 2005) had failed, many Kenyans believed that fundamental institutional change was necessary to avoid future conflict. It is in the context of these new institutions, and new partially autonomous counties, that Kenyans vote on March 4th.
Similar to other developing contexts, political alliances and local institutional legacies influence the rate, mode and geography of electoral conflict. Understanding the current political climate in context avoids the fatalistic view that conflict is inevitable. Looking into Kenya’s past, the details of political dynamics are pivotal in explaining the results and contest. A noteworthy example of this is from the peaceful 2002 elections where the divisions between ethnic communities did not line up in a volatile manner. In that case, the broad alliance within Mwai Kibaki’s National Alliance of Rainbow Coalition (NARC), opposed Kenya African National Union (KANU) and President Daniel Arap Moi. In to the 2002 experience, the political climate leading into the 2007 election campaign was highly polarized between Kikuyu and non-Kikuyu ethnic communities. In some areas, these were based on historical land grievances (as in the territories surrounding Eldoret where ODM candidate William Ruto fomented animosities against PNU). In other regions, the mechanism for elite manipulation of supporters was routed in long term exclusion from power (as for Luo, who have been marginalized from influence in government since the early-independence dispute between ODM candidate Raila Odinga’s father, Jaramogi Odinga Odinga, and Jomo Kenyatta).
Informed conclusions about the role of either national or local level influences depends on having detailed information about the individual events that constitute a broader conflict episode. The timing of an event, the exact location where it took place, and identity of actors involved, are all crucial pieces of information for making conclusions about conflict in Kenya and other countries. For example, at the local level, one common explanation for electoral conflict lies in the insecure land tenure regimes that exist for territories of the former “white highlands” in the central-northern areas of Kenya’s fertile Rift Valley. In areas of Uasin Gishu, Tranz-Nzoia, and Koibatek, for example, attacks after the 2007 poll erupted immediately, and continued for weeks. This violence is closely tied to the messy politics of settlement schemes in the area during the early years of Kenya’s independence. Ignoring those institutional legacies is a serious mistake.
Finally, the differences between bouts of violence are as important as similarities. For example, the state played a varying role between the election violence that took place in 1992, 1997, and 2007. Rightfully, observers latch on to the discourse of eviction (mainly of Kikuyu) that emerges from narratives of conflict in the central-northern Rift Valley. But a simple interpretation ignores the fact that violence during the 1990s was state led and sponsored (by Moi-KANU), and in 2007 it was largely private individuals or political militias who perpetrated violence. This is one simple example of an important difference that has clear implications for scholars’ understanding public security provision in clientelistic political regimes.
Andrew Linke- University of Colorado