This article is part of a debate organized by Oxford Transitional Justice Research (OTJR) in collaboration with Moi University (Eldoret) and Pambazuka News. A selection of essays based on this debate will be published in an edited volume by Fahamu Books. For PDF documents of the debate please go to www.csls.ox.ac.uk/otjr.php.

It is not often that participants in ethnic cleansing confess to it openly, but William ole Ntimama has managed it twice: in a 1996 interview, and more recently. The brazenness of the impunity is revolting: it is natural to want accountability and reform, and equally natural to think we can have both. This, unfortunately, is a bit of a farce: stable reform and calling the violent to account are incompatible. The key is to see that the main strand of political violence in multiparty Kenya is unified by a stable and clear set of aims: majimboism, understood to mean the Kenyan form of exclusive ethnic federalism which finds its most fervent advocates in Rift Valley Province’s political class. In the 1990s, the violence was driven and supported by the majimboist-controlled state; it didn’t require mass mobilisation. 2007 was a genuine departure because the extent and intensity of majimboist violence demonstrated that communal mobilisation for violence is an effective substitute for state support. The beneficiaries have no incentive to give it up, and every incentive to avoid the consequences of past violence by holding onto power. Since their participation is necessary for reform, we can have either reform or accountability but not both.

My first job is to show that despite appearances (diversity of actors) the violence was actually unified in aim. The argument is simple: Rift Valley province is the centre of political violence in multi-party Kenya. The easy metric is deaths: even in 2007, when the violence is supposed to have been much better spread, 65% (744/1133) of recorded murders happened there (Waki: 309). We’re now eighteen years into the violence: it has broken out intermittently since 1991. Prolonged violence of this sort – locally-specific, ethnically-targeted, lethal, and carried out by a number of coordinated small groups – is organized and backed by some sort of ideological structure. That follows from the fact that most unplanned violence is difficult to start or maintain, tends to be brief, and is usually non-lethal (Collins 2008: 14-16). The exceptions to the rule of brevity (for small-group violence) occur where:

either (a) the fight is highly circumscribed, so that it is not really “serious,” or it is clearly understood that there are safeguards to limit the fighting; or (b) the type of exception described by the expression “hitting a man when he is down” (although the victim may well be a woman or a child), where in effect there is no real fight but a massacre or punishment (Collins 2008: 16).

Repeated bouts of this kind of sustained lethal violence require planning and preparation; planning and preparation for violence require coordination and justification, and hence institutionalisation. The justification is fairly clear: a middle-aged man interviewed by Al-Jazeera in Kibera, and Jason Kosgei in the Christian Science Monitor, gave almost identical answers: the violence was to end state-backed Gikuyu domination, which had begun with Kenyatta and never ended. As Lynch 2008 reports (Lynch 2008: 567), a significant portion of Kalenjin backed the violence, and have fairly specific reasons for doing so. Those reasons aren’t significantly different from those reported in Multiparty Politics in Kenya: In 1992, Biwott promised that non-Kalenjin trading licences would be revoked, and Lotodo demanded that all Gikuyu leave West Pokot (Throup and Horsnby 1998: 543). Then, as now, the immediate aims of the violence — to remove non-Kalenjin from the Rift Valley, and to place the remainder, if any, in a subordinate and dependent position –were clear.

The state did outsource violence in the 1990s; much less so afterwards. Why? In the face of the state’s significantly increased capacity for repression (Branch and Cheeseman 2008: 20), why was the violence so much worse in 2007? And why was violence was much better controlled in the 1990s than it was later? Most analyses of the violence have proceeded by identifying the actors, on the reasonable assumption that pinpointing the actor is a good proxy for pinpointing the motive. Going directly to motives, however, has some explanatory advantage: it promises informative answers to each of those questions.

Susan Mueller’s The Political Economy of Kenya’s Crisis may be the most comprehensive analysis of the underlying causes of the post-election violence. Her argument is pretty much that three factors – privatized, diffused, extra-State violence; ethnic clientelist parties; and the high-stakes prize of the Imperial Presidency – conjoined (with a very close election) to blow things up in 2007. The obvious response is to ask why nothing similar happened in 1997, and why all the factors she mentions are structural: the explanation, as given, would still work if the agents were switched. Every factor she lists was present then – if anything, the Presidency was even more imperial, the ethnic clientelist parties even more intensely ethnocentric. Yet there was relatively little violence around election time in 1997: most of the violence came well before or well after polling day. In particular, the announcement of the results in 1997 – results which in several cases were known to be entirely fraudulent – passed without incident.

This lack of specificity leaves the analysis less compelling than it might be; nowhere more so than her analysis of the state’s cession of its monopoly of violence. It is one thing to observe that the state outsourced violence; quite another to ignore the fact that the first Kibaki administration sought, very crudely, to re-establish the monopoly of violence. It is more accurate to attribute the cession of the state’s monopoly of violence to the Moi state – the state in the hands of the majimboist faction. That move – appeal to the motives of the faction in control of the state, rather than the state itself – explains why the state acted so differently either side of 2002, and it offers a direct explanation for the state’s choice and method of outsourcing violence. Moi’s outsourcing of violence in the 1990s is often explained as a pragmatic choice: irregular gangs and militias are untraceable; in employing them, the state got its extra-legal coercion done while minimizing its exposure. This is utterly unconvincing. A quick flick through the Akiwumi report demonstrates that civil servants openly participated in the violence. Nicholas Mberia – then the District Commissioner in Kericho – and 29 APs in his command violently evicted tenants from Buru farm on the morning of 13 December 1993. Not long after, he was promoted to Provincial Commissioner, Rift Valley Province. Several witnesses to the evictions in Enoosupukia testified that the Narok County Council wildlife ranger Johnson ole Punywa shot dead three residents. He too was later promoted. (Klopp 2001: 496). If the point of outsourcing violence was to conceal the state’s hand, then the state made a fearful mess of it. It’s likelier that the outsourcing of violence was driven, at least in part, by ideological motives – the drive to weaken and personalize the centre of the state, while strengthening the majimboist periphery.

Branch and Cheeseman account for the upsurge in violence by appeal to elite fragmentation. That’s a necessary rather than a sufficient condition. Remember that what’s wanted is an answer to why the violence crossed a certain threshold – why it escaped control of the state. Without an underlying capacity for violence, elite fragmentation need not have violent consequences, and it certainly need not have consequences so violent that the state struggles to control them. Appeal to a generalised diffusion of violence is nearer the mark, but it still underdetermines the quality of the violence in the Rift Valley: if elite fragmentation were sufficient to explain the escape of the violence from state control, then that would have happened in more than one place. It didn’t so, it isn’t. Capacity for violence matters; appeal to majimboist motives is sufficient to predict it.

After nearly 20 years or so of intermittent ethnic violence with zero consequences, with and without state support – and since much of the Kalenjin political class (and William ole Ntimama) is on board with the violence – it is difficult to avoid the conclusion that the violence has communal approval and support (Lynch 2008: 566-7; Ashforth 2009: 16). Some significant proportion of Kalenjin opinion leaders outside the political class – the rural middle classes, in particular – have been radicalised. That has been a necessity: when the violence had state support, it did not need communal mobilisation, and there was no need for the ideological backing. Absent state support, communal backing is necessary: the violence has become more ideological as it has become more popular. The balance of power is such that Kalenjin opinion leaders who support ethnic violence, and the majimbo project which justifies it, lack effective internal constraints.

The view that majimboist violence is driven by elite incitement is false: rather, majimboist aims are now widely popular outside the political class, and are captured by it (Ashforth 2009: 18-19). Majimboists willing to resort to violence are well-mobilised because they’ve had to be: without state patronage, the fervour of their cause has had to cover for the organizational goodies the state would have brought. The underlying strategy of reform-by-coalition-government in Kenya is to get the big beasts of the political jungle into government, so that they’re all bought into the new constitutional order. If they are to feel invested, they must be free to manoeuvre; for majimboist politicians, that freedom of action is directed, as it must be, to avoiding accountability for the violence. There can be no new constitutional order without majimboist involvement; since most of the violence has been in majimboist areas, accountability and reform are incompatible.

*Dr. Daniel Waweru is the Chief Editor of KenyaImagine

The above article is available as a PDF

BIBLIOGRAPHY

Adam Ashforth (2009). “Ethnic Violence and the Prospects for Democracy in the Aftermath of the 2007 Kenyan Elections.” Public Culture, 21(1): 9-19.

Shashank Bengali (2009). “One year after the massacres, Kenya’s runners reflect.” The Christian Science Monitor, February 25, 2009 <http://www.csmonitor.com/2009/0225/p25s11-woaf.html> (8 July 2009).

Bill Berkeley (1996). “An Encore for Chaos?” The Atlantic Monthly, February 1996. <http://www.theatlantic.com/issues/96feb/africa/africa.htm> (08 July 2009).

Randall Collins (2008). Violence: A Micro-sociological theory. Princeton: Princeton University Press.

Commission of Inquiry into Post Election Violence (2008). Report of the Commission of Inquiry into post-election violence (“Waki”). October 15 2008. <http://www.dialoguekenya.org/docs/PEV%20Report.pdf> (08 July 2009).

Jacqueline Klopp (2001). “Ethnic Clashes’ and Winning Elections: The Case of Kenya’s Electoral Despotism.” Canadian Journal of African Studies, 35(2): 17.

Gabrielle Lynch (2008). “Courting the Kalenjin: The failure of dynasticism and the strength of the ODM wave in Kenya’s Rift Valley province.” African Affairs, 107(429): 541-568.

NTV Kenya (2008). “William ole Ntimama War Monger or responsible minister?” 24 July 2008. <http://www.youtube.com/watch?v=dLIM9gPHq5s> (08 July 2009).

David Throup, Charles Hornby (1998). Multi-party Politics in Kenya. Oxford: James Currey.

Akiwumi Judicial Commission of Inquiry on Tribal Clashes (1999). Report of the Judicial Commission appointed to inquire into tribal clashes in Kenya: Rift Valley. Date of publication unclear. <http://www.scribd.com/doc/2204752/Akiwumi-Report-Rift-Valley-Province> (08 July 2009).