MRC graffiti on a wall in Mombasa.

In discussing the elections at the Kenyan coast, I want to make two, perhaps contradictory, points. The first is that the coast has its own particular dynamics, which must be appreciated to understand the nature of elections there. The second is that events at the coast help us to understand why it was that the Jubilee coalition clearly won more votes in the presidential election than CORD (in spite of the opinion polls); why the precise extent of that majority is unclear; and what the consequences of the Jubilee presidential victory may be.

The distinct character of the election at the coast is mostly connected with what is called, in shorthand, “˜the MRC’, though I should stress that the use of that name suggests a degree of cohesion and organization which is really misleading. Using this name, various individuals have, over the last few years, advocated secession, arguing – wholly inaccurately – that various treaties from the colonial period provide a legal basis for this.

There are profound divisions amongst those who make these claims, and their agenda is confused in multiple ways, but – while active support for secession is limited – the idea of the MRC became a focus for a widespread sense of dissatisfaction among those who consider themselves coast people. The evocation of written documents to assert claims was a powerful and attractive idea for people who have tended to experience such documents as tools of oppression.

The incoherence of this idea was, however, very apparent in the elections. Some of those who claim to speak for the MRC called for a boycott; a spate of localized violent incidents in late 2012 were linked to MRC supporters, but also denounced by some who claimed to be leaders; some of those leaders were arrested, then freed on bail, and apparently withdrew their call for a boycott; others pursued a pointless legal challenge against the election. Some people who claimed to be local MRC officials ended up not only registering as voters but participating in the elections as activists.

And on the eve of the election, a series of organized attacks in peri-urban Mombasa and in Kilifi County led to the death of ten members of the security services and a number of other people; police and press reports linked these to “˜the MRC’, though one of the self-appointed spokesmen denied this; in the last few days alleged leaders of the MRC have been rearrested and accused of organizing violence.

It is difficult to know what to make of all this. It is hard to see how the visible leaders of the MRC would muster the financial resources or organizational ability to pursue all the activities allegedly linked to them, and there are competing conspiracy theories about what other hands may be at work here. All that can really be said is that there is an evident sense of alienation among those who call themselves “˜coast’ people, and that it is evidently not difficult to recruit young men and instigate them to acts of extreme violence.

That sense of disaffection had complex consequences. It was generally linked to support for the new constitution, and resentment against what some would see as the continuity represented by Jubilee; the popular vote on the coast generally went strongly for CORD. But disaffection also led to non-participation. There was nothing like a total boycott of the election, but registration rates in some parts of the coast were low, as were turn-out levels – relative to other parts of Kenya. Secessionist sympathies may help explain this; and the violence on the night before the election no doubt had an effect, as many polling stations opened late and some were closed early. The precise impact of these is hard to gauge – electoral turnout at the coast has always been low. But it does seem likely that the net effect of secessionist agitation was to reduce the presidential vote for CORD at the coast.

That, in turn, helps us to understand the election result as a whole, particularly why it looked quite different to the opinion polls (which predicted a very close race between CORD and Jubilee for the presidency). Registration and turnout were key to this. Put very simply, Jubilee were better at mobilizing potential voters. Jubilee ran a better campaign; they seem to have spent more money than CORD, but they also spent it to better effect; they were effective in playing a “˜nationalist’ card – in which the ICC unwittingly aided them – and also in presenting themselves as youthful, change candidates. And in their core constituencies, they turned out the vote.

CORD’s internal incoherence and disputes did not help – these were very evident in Mombasa county, where the contest for governor and senate seats, and in many cases for MPs, was between ODM and Wiper. The CORD “˜allies’ traded accusations of malpractice, and were absolutely unable to coordinate; this was NOT an election-winning approach. All in all, while Raila was well ahead in the presidential vote at the coast, he nonetheless missed out on many votes there.

The coast, therefore, helps us see why Uhuru got most votes, nationally. It also helps us to see why the precise details of the majority were open to challenge. Logistically, it should be said, a number of things went right in the election. There were dire predictions that the IEBC would not manage the physical logistics, that ballot papers would not be ready, staff would not be in place, that puzzled voters would spoil hundreds of thousands of ballots. None of that happened. In physical terms, the IEBC did a pretty good job: it got staff and (almost) all the paraphernalia to the right places at pretty much the right time; the staff were largely exemplary in their performance and despite an extremely long day, they got the voters through. Through hundreds of thousands of acts of patient explanation, they managed to get most people to vote successfully.

Some other things undeniably went very wrong. The system for electronic voter identification – the “˜Evid’ devices – failed in many places. Basically, wherever there was no mains power, the Evids only worked for a few hours, because batteries ran out; where there was power, in some places their use was intermittent because of muddles over log-ins and encryption. The paper register was not the same as the Evid register; some people (the proportion seems to have varied from place to place) had registered but were on neither system. Some people may have been unable to vote as a result; some people who were not registered may have voted; crucially, there was no reliable electronic record of who had voted and who had not.

Our ability to judge the impact of this is significantly impaired by the other great technological failure of the election: the results transmission. This failed comprehensively at the coast; only a small number of results were successfully sent by SMS from the southern coast. Here, as elsewhere, multiple and conflicting explanations were given: server problems, log-in problems, hand-set failure. And here as elsewhere, the result was that tallying and result transmission relied entirely on paper, on the chain from poll stream to constituency to county to national, with the links in that chain being impossible to verify without the services of dedicated and effective party agents working in coordination at every level. Which most parties lacked.

The consequent suspicions over the process were compounded because human error – the effect of which was significantly increased by tiredness, as the long count dragged on and on – led to multiple and pervasive inconsistencies. The CORD court case [did], of course, [seize] on those inconsistencies as evidence of malpractice, though the errors did not seem to point clearly towards systematic changing of votes. That Uhuru got significantly more votes than Raila was anyway confirmed by the Parallel Voter Tabulation process run by election observers, though by its nature that process could not confirm the exact extent of the majority.

So, what does the confirmation of Uhuru’s victory mean? From a coastal perspective – with the exception of Lamu – the vote went against Jubilee, and it is possible that the result will be to increase the long-term disaffection of those who consider themselves “˜coast’ people. The new constitution, intended to redress inequalities and devolve power, is likely to become a bone of contention. Local governors in Kwale, Mombasa, Kilifi and Taita, and in Tana River, may soon be at loggerheads with the central government over everything from budget allocation to office space to control of key resources – notably, Mombasa port. The county commissioners, created by Kibaki in the last gasp of his government, may be central government’s tool to restrict local autonomy and assert central authority.

The new constitution was always going to be problematic to implement, because of its complexity, high expectations, and low levels of understanding of detail. Kenya now has a vice-president who opposed the new constitution and a president who was – I think it is fair to say – at best ambivalent about it. The process of implementation is, therefore, likely to be challenging; and arguments over devolution will be at the heart of politics for the next few years.

Justin Willis is Professor of History at Durham University. Previously he was Director of the British Institute in Eastern Africa, based in Nairobi.