Among the many important questions about yesterday’s election, is this: Will the energetic endorsement given to Raila Odinga by outgoing President Uhuru Kenyatta be a blessing or a course in his fifth – and almost certainly, final – attempt to enter State House.
Or to put it another way: Given Odinga’s long history as the country’s most steadfast opponent of the status quo/advocate for change (of whatever nature, and however extensive), and with a phalanx of dedicated followers, much of whose loyalty was derived from that intransigent stance and often very personally injurious stance, what will be the ultimate impact of the President’s endorsement, given that their March 9, 2018 ‘handshake’ led to the eager ‘appropriation’ of the opposition space Odinga and his ODM party were seen to vacate by William Ruto, the President’s deputy over the last decade.
This situation has obvious implications for voter behaviour, not just whom they will voted for, but also whether they voted at all, especially those registered voters from ethnic groups without a (major) presidential candidate or deputy-presidential one.
Such ‘behavior’ in this context begins in the Mt. Kenya region – the President’s turf. In none of the three previous elections when Odinga was running against a Kikuyu (in 2007, Mwai Kibaki, the incumbent) and in 2013 and 2017, Kenyatta) he did not surpass the 5 percent mark.
But his campaign was highly competitive because of his wide margins in the Coast, Lower Eastern and Western. This time, however, the Ruto-Gachagua pair had apparently made substantial inroads in all three of these regions.
For example, in TIFA’s most recent polls (both conducted within the last three weeks), the Kenya Kwanza team attracted about one-fifth of all survey respondents in the former region and more than one-quarter in each of the latter.
For his part, Odinga had just over one-quarter of support in Mt. Kenya.
Assuming Ruto does as well or better than the TIFA figures indicate, it seems Odinga would need at least one-third of the Mt. Kenya vote to stay competitive – a figure way above the 10 percent figure Nyeri Town MP Wambugu Ngunjiri asserted that they will find “hard to get” in an August 3 Twitter post.
Of course, in a close election – which, by every indication, this one is likely to be – every vote counts, so the ‘margin-of-victory’ (as well as the margin to prevent a run-off contest) could come from anywhere, but the numbers from Mt. Kenya, more than elsewhere, will indicate the magnitude of the endorsement Odinga has received from the current president.
Looking at the ‘Changing Voters’ for Answers: Who Gains…and Who Loses?
Further interrogation of the data recently obtained by TIFA helps to at least partly answer this question.
To begin: In my previous piece for the Standard (of August 3: ‘Who are undecided voters and why have others changed their voting intentions?’), I suggested that the greater numbers of survey respondents claiming to be ‘undecided’ (or declining to state their intentions) regarding their presidential voting intentions has been markedly higher than in the past, even if they have continued to decline as August 9 approached.
And the reason for this, I argued, is mainly because of the ‘switching-of-positions’ between erstwhile Opposition leader Raila Odinga and Deputy President William Ruto, as a direct consequence of the March 9, 2018 ‘hand-shake’.
In this context, I considered (hitherto unreleased) data from TIFA’s late July survey that showed how many people say that had they changed their voting intentions since last year – whether from one candidate to another, from being undecided to a candidate, or from a candidate to being undecided: a total of 13 per cent, as shown below:
Among all those (13 percent), there was little difference between those supporting Odinga and Ruto (11 percent of the former and 11 percent of the latter).
However, among the few respondents who admitted to having changed from one candidate previously to another now, Odinga was clearly the main beneficiary, even if, among those who had changed from being decided previously to supporting any candidate now, Ruto has made a greater gain.
That is, while among those who have changed their voting intention from one president candidate to another Odinga had a 12 percent advantage (i.e., 40% vs. 31%):
Next, for all those who switched from one candidate to another, and to having been ‘undecided’ to supporting any candidate now, we asked what is “the main reason” why they did so, and obtained these responses:
Several points can be drawn from these responses. One is that more ‘newcomers’ in terms of deciding which of the two main presidential candidates to support had moved to Ruto than to Raila on the basis on specific campaign promises, including those detailed in the Kenya Kwanza manifesto (56% vs. 43%).
On the other hand, more of those who decided to vote for Odinga did so on the basis of his choice of a running-mate (16% vs. 10%).
Likewise, Odinga had an advantage among those who had to support another candidate because their preferred one had withdrawn from the race (13% vs. 8%), clearly a consequence of the decision not to contest by Kalonzo Musyoka and Musalia Mudavadi in particular.
Next, in this series of questions, we asked all respondents if they had voted for Odinga in the 2017 election, and split them into two groups: those who say did and did not do this. We then correlated each group with their present presidential voting intentions, producing the following figures:
They revealed that there had been a quite even ‘switching of places’ in terms of their previous vs. current voting intentions. Specifically, whereas about three-quarters of those who voted for Odinga in 2017 said they would again vote for him in 2022, the same proportion who did not vote for him then said they would be voting for Ruto this time.
However, while only 15 percent of those who voted for him in the last election have moved to support Ruto now, 23 percent of those who did not vote for him last time said they would do so in 2022 – a clear gain for him.
Finally, we asked those who voted for him in 2017 but not in this election what is the main reason they have ‘abandoned’ him:
As shown, the most frequent response was his relationship with President Kenyatta (42 percent). But other reasons also received considerable mentions: unhappiness with the ODM nominations (12 percent), and that he has “changed his policies/agenda” – also clearly another way of expressing a rejection of his ‘handshake’ relationship with the president.
At the same time, two other reasons – that “his elections are always stolen” and “his age” having nothing to do with his political ‘about-face’ since March, 2018.
So, Who Gains, Who Loses, and What Future for Kenya’s Electoral Politics?
I began this analysis by posing the question of how the nearly-total role-reversal of the former Prime Minister with the Deputy, encapsulated in his inverted relationship with President Kenyatta since the ‘handshake’, would affect the outcome of the presidential election.
While survey figures, however ‘scientific’ cannot predict such a result in a close contest and where differential voter turnout cannot be calculated in advance (or where not all respondents are prepared to reveal their true intentions), such research can, nevertheless, reveal some of the underlying dynamics at play in the decision-making choices that all voters must confront.
Of course, with all the attention on the outcome, it is understandable if certain critical, longer-term, questions have seldom been asked.
These might include the following: (1) What is the relationship of leaders to their political parties when they contemplate making such 180-degree flips in terms of their relationships with one another? Or to put it another way, does the general absence of policy-identity of the main parties invite such short-term, if not opportunistic, role-reversals?
(2) What are voters to make of the policy-content of the ballot-choices they confront when the most influential leaders restructure their relationships in such ways?
(3) How will the fall-out between the president and his deputy, and the former’s embracing of his former electoral foe affect the calculations of other leaders moving forward when achieving power is seen as contingent on jettisoning certain past positions and alliances?
And more generally, (4) Does this fluidity of election alliances in Kenyan politics support rather than undermine national unity, if former foes can become fast-friends, based on changing circumstances, so that “issue-based” politics as embodied more or less permanently in political parties – one of the governance pillars of Vision 2030 – will have to wait until Kenyan society is more concretely divided into interest groups that use competitive politics to achieve their policy ends?
Such questions may seem quite distant this week. For now, as the votes are tallied, let us see how voters have answered some of the more immediate ones that were addressed in a number of recent surveys such as TIFA’s.
Dr. Wolf is a research analyst for TIFA Research in Nairobi.