Teetering on the edge: The 2013 Kenyan Presidential election

Today, Keynans will vote for a new president.  The whole world is watching as well.  Why?  The last Kenyan presidential election in 2007 lead to widespread violence as supporters of Raila Odinga accused Mwai Kibaki and his supporters of stealing the election. Given that politics in Kenya is often strongly linked to tribal affiliation, much of the violence was directed by members of one tribe toward those of another. At its core, much of the violence found its origins in many long-standing economic grievances. Against this backdrop, the international community has kept a close eye on this election.

One constitutional change resulting from the violence in 2007 was that if no one candidate gets a majority of the votes, there will automatically be a second round run-off election between the top two vote getters within 30 days of the first round election-day.

In total this year, there are eight candidates running for president this year but only three primary players.   Who are these players?


First, Uhuru Kenyatta is presently a deputy prime minister and until recently was the minister of finance in Kibaki’s government. He also has considerable political pedigree, being  the son of Kenya’s first president Jomo Kenyatta and is a member of the Kikuyu tribe, which represents about 20% of the population—the largest tribal grouping in Kenya. He also ran for president in 2002 but lost that election to Kibaki, a fellow Kikuyu.

Second, Raila Odinga is the prime minister of the Kenyan parliament, as a result of a post-2007 power-sharing agreement, and was one of the main protagonists in the 2007 controversy, being on the losing end against Kibaki in that presidential election. Odinga’s party is the ODM, and he is from the Luo tribe which represents about 14% of the population. Odinga clearly represents the opposition.

Finally, Musalia Mudavadi is another deputy prime minister in the current government.  Having past political ties with both Kenyatta and Odinga, his current party is the UDF; he is from the Luhya tribe, which is Kenya’s second largest but is less cohesive than some others.  Although his prospects of winning the election are slim, Mudavadi could be a spoiler or kingmaker in a possible second round.

Here it is important to note for those novice observers like myself that Kenyan politics is less about party and much more about tribal membership and candidate personality.  As one Kenyan analyst once told me, “parties and party alliances come and go with the weather—but tribal affiliations are constant and perhaps the important driver of Kenyan politics”.  Noted!

So what are some likely outcomes?

First, there probably will be a second round if there are no wild shifts in voter turnout as no candidate has a majority of voters at this point in the polls. Indeed, Kenyatta and Odinga are running about dead even—both with numbers in the mid-40’s. Mudavadi is a distant third with polling in the single digits. Among valid votes (total declared vote – undecided), Odinga only has about 47.5 percent of votes (see table below).


So, all things considered ,  it looks like there be a second round between Kenyatta and Odinga. So what will happen in a possible second round?  Here predicting a winner in the second round is far less clear at this point.  There are a number of ways to look at possible scenarios.

First and foremost, let’s step back a second and look at elections in general.  Based on our own database of over 200 elections around the world, we know that election in which the government is viewed favorably typically favor the government candidate (’Much ado about nothing: Obama will be president, again, in 2013’).  In contrast, in an election where the government has weak polling numbers, its candidate is less likely to fair well.

Additionally, we know that incumbent candidates have almost a three-fold advantage over successor candidates. This will weigh in our assessment of second round chances.

So how does this play out the in Kenyan case?

For the sake of analysis, I treat Odinga as the opposition candidate and Kenyatta as the government candidate. I also assume an approval rating of 58%, an average of recent polls.   Here it is worth noting that the opposition-government distinction is far from clear as both Odinga and Kenyatta are ‘government’ in the sense that Odinga is the prime minister and Kenyatta the deputy prime minister.  I see Kenyatta as the ‘government’ candidate, however, because of both his closer ties to the Kibaki government and the fact that they both are Kikuyu.

Probability of Government Candidate winning election under different scenario assumptions

Approval Rating



























So what do we find?

According to our own electoral forecasting model, a government candidate who is a successor, like Kenyatta, has about a 58% chance of winning the second-round run-off–so about close to even odds for both Kenyatta  and Odinga (see table above). If Kenyatta were the incumbent candidate, he would almost certainly be assured a second-round model, underscoring the power of incumbency, but unfortunately for him he is not.

The implicit assumption of our model is that undecideds or swing-voters will typically break for the government candidate when the government is rated well and will end up supporting the opposition candidate when they are more pessimistic.

This assumption may not hold when voting is highly crystallized around affinity groups like tribes.  Voters might, rather, follow the lead of their political leaders in the second-round.

An alternative approach would thus be to assess to which candidate political leaders like Mudavadi would be likely to throw their support.  One indirect way to do this is to assess base support: if a voter’s top choice did not make it to the second round, for whom would the person vote?  Here our own Ipsos polling suggests that Odinga has roughly a 3 to 1 advantage over these swing voters (29% to 12%), though a strong plurality of Kenyan swing voters are undecided (37%) when it comes to their second round choice.  Here let’s say advantage goes to Odinga.

Ultimately, the outcome of a possible second round run-off is far from clear and most probably will be very close. Here I would suggest that the analyst pay special attention to those problems which most resonate with voters (number 1 is the price of food) and how each candidate has been messaging to this point. My experience is that second round elections tend to be much clearer in terms of the policy trade-offs with those candidates who are best able to draw such distinctions winning the day. But let’s see.