Nyando MP Fred Outa (second right) flees after rowdy youths ejected him from a burial in Nyando,Kisumu County, on June 10. [Photo: Collins Oduor, Standard]
This week has been ablaze with the “usual” political intrigue, slurs and accusations that is a mainstay of election campaigning in Kenya.
Just within a week, ruling coalition Jubilee has called for the arrest of NASA candidate Raila Odinga for alleged hate speech and the Opposition coalition wants to move to court to block the printing of ballot boxes. There are accusations that the Independent Electoral and Boundaries Commission (IEBC) and/or the military is planning to rig the August poll, while internal supremacy wars in both the Jubilee and NASA camps are threatening to derail the respective campaign teams – all reported in the space of a few days.
It is enough to make one breathless.
It is common knowledge that the line between the “usual” political competition of a healthy democracy, and electoral crimes and offences is very thin in Kenya.
Violence, bribery, voter intimidation, ballot fraud, arson, looting, undue influence, even kidnappings and murder are all real risks associated with elections in Kenya.
But a recent report from the National Crime Research Centre (NCRC) reveals that not all election offences are “created equal”. The report analysed various election offences and criminal acts in 22 counties in Kenya, taken as a nationally representative sample.
Say – purely for argument’s sake – that you were a devious politician wanting to win an electoral contest by all means necessary. Would it make more sense to bribe voters, or give inflammatory speeches and rousing rhetoric against your opponents? Does vote suppression and intimidating the electorate work better for you than ballot fraud or fiddling with tallying?
The data suggests that there is “method in the madness” of electoral malpractice in Kenya. It gives hope that chaos, violence and fraud are not an inevitable, or even random, feature of politics in Kenya.
The dark choices that politicians make are informed by rational “cost-benefit analysis of malpractice”, you could argue.
And understanding how those trade-offs work can make monitoring and accountability of the electoral process more effective, because you can now anticipate the motives and likely actions of unscrupulous politicians.
The NCRC report lists a host of electoral offences that were mentioned by respondents in the respective counties, but the top five most prevalent are: bribery of voters, voter/ballot fraud, hate speech, physical fighting and voter intimidation.
The researchers define voter/ballot fraud as offences related to improper marking of ballot papers, ballot stuffing, irregular tallying and the like.
Ignoring signs of war is what causes one
Ignoring signs of war is what causes one
Ignoring signs of war is what causes one
Hate speech involves the use of derogatory statements against opponents, and inciting messages that could trigger hatred or violence between parties.
Voter intimidation is a climate of fear or anxiety created by the political contest that discourages freedom of choice in elections, or even prevents people from voting all together.
Of course, the lines between the offences are not always so clear-cut: voter intimidation is often the result of hate speech; ballot fraud can trigger physical fighting, while bribery is very often the currency that fuels all these malpractices.
But there are some insights to be drawn even from these imperfect categories. The researchers found that in the 2013 General Election, reports of bribery were highest in Narok, Garissa and Bomet counties. Voter/ballot fraud was highest in Garissa, Mombasa and Kisumu, while hate speech was highest in Narok, Kisumu and Nairobi. Physical fighting was most prevalent in Nairobi, Bomet and Siaya, while voter intimidation was most common in Garissa, Migori and Uasin Gishu, out of the 22 counties surveyed.
Parse the data further, and more trends emerge. Narok County was leading in terms of bribery and hate speech, but respondents were much less likely to say they felt intimidated or were afraid to vote.
This suggests that in Narok’s case, the political class was working to energise the electorate and encourage them to come out and vote. Narok is an ethnically mixed, swing county, where every single vote was up for grabs.
Both Jubilee and CORD (at the time) needed all the bodies they could get, so suppressing the vote by intimidating the electorate would end up disadvantaging both camps.
Ballot fraud in Narok was also relatively low, perhaps because the race was so close that each side was watching the other very closely. Stealing ballot papers or pre-marking them would be caught right away.
It makes more political sense for all the competing parties, in such circumstances, to let the contest unfold in a relatively free, but acrimonious way.
It seemingly worked – voter turnout in 2013 was 89.8 per cent in Narok, higher than the national average of 86 per cent. And the presidential results were close: 50.6 per cent for Raila Odinga and 46.7 per cent for Uhuru Kenyatta.
Contrast Narok’s case with that of Mombasa, where reported bribery was quite low, in fact, the second lowest of the 22 counties sampled. However, Mombasa was second highest in ballot fraud. What’s going on?
As a political tactic, bribery is effective as a credibility signalling mechanism. It tells voters that the bribe-giver has access to resources, and gives them hope that they may receive benefits in the future.
But bribery has a limitation. There is the problem of “free-riders”, that is people who would take the bribe and not vote, or perhaps are not even registered voters. In an urban context where there are a lot of people, there is often a feeling of anonymity, and that your single vote might not even matter.
This presents a problem for a bribe-giving politician: there are so many people to bribe, and you might just be “wasting” money on people who come and cheer at rallies but will not vote for you.
The IEBC data confirms this view – voter turnout in Mombasa was just 66.6 per cent in the 2013 polls, far lower than the national turnout.
So politicians in Mombasa seemingly chose a different route. Rather than spend money on bribes, they chose to fiddle with the process on Election Day itself – the actual voting and tallying, hence Mombasa’s high score in ballot fraud.
It also helps that Mombasa was, by and large, a CORD stronghold in 2013. Unlike Narok, there were fewer “Jubilee eyes” watching the tallying in Mombasa, thus more opportunity to get away with fraud on the material voting day, and its immediate aftermath.
This data suggests that the underhand tactics politicians use to get ahead are by no means random or inevitable.
In your “strongholds”, for example, you really don’t have to bribe or intimidate voters. What you need is some “glue” that will energise and hold your base together, so hate speech would be a tactic politicians would reach for here.
It raises fear of “the other”, creating a bogeyman that must be collectively defeated. It is not a coincidence that some of the counties that had the highest reports of hate speech in 2013 were in “stronghold” areas, such as Kisumu and Kitui, as well as much of Central Kenya where the ICC cases were a rallying point in the campaigns.
Intimidation as a political tactic, however, would be attractive is some swing areas where there is monitoring and enforcement capacity, for example in the form of already existing political or criminal gangs, to keep voters at home and fearful.
It is not a coincidence that reported intimidation was highest in Garissa, a region where Al-Shabaab militants have been active since 2011. Grenade and IED attacks were a very real risk for voters, as they had experienced Al-Shabaab’s bloody ways, first hand.
Separate data from the Global Terrorism Database supports this view – between October 2011 and December 2015, the terror group was responsible for 68 per cent of all terror attacks, but accounted for 81 per cent of deaths, and 82 per cent of injuries from terrorism.
In other words, terror attacks claimed by Al-Shabaab have a higher casualty rate than those that are not, suggesting the terror group has access to more powerful weapons and better coordination and technology.
Similarly, Uasin Gishu was another place where there was a high prevalence of reported intimidation, according to the data. Eldoret and its surrounding areas had been a hotspot in the post-election violence five years prior, and the collective trauma was, in some ways, still fresh. The two opposing sides in 2007 were now on a joint ticket, but there was an element of coercion and pressure on the electorate to vote for the Kenyatta-Ruto ticket, this data suggests.
Ultimately, this data suggests that it is possible for election monitors, observers and ordinary citizens to deploy their resources in a more targeted way in the August 8 election.
Instead of trying to control and prevent everything, better understand the factors that are likely to be in play in particular counties, cities and towns – and be one step ahead of the “bad guys”.
– The writer is a writer, journalist and executive editor of Africa data explainer site Africapedia, and a 2018 Nieman fellow at Harvard University.