Kenya has a president-elect, at least according to the official result announced on Monday. William Ruto, the current deputy president, was declared the winner of the 9 August election with 50.49% of the vote, narrowly crossing the 50% threshold required to avoid a runoff. Raila Odinga received 48.85%, leaving fewer than 250,000 votes between the two candidates in a country where more than fourteen million presidential votes were cast.
A close election naturally creates political tension. What made Monday’s announcement more extraordinary was that the institution responsible for administering the election publicly divided over its own result. Shortly before electoral commission chairman Wafula Chebukati announced Ruto’s victory, four of the seven commissioners held a separate press conference and said they could not take ownership of the final stage of the process, which they described as “opaque”. Odinga subsequently rejected the result and has said he will challenge it before the Supreme Court.
As of tonight, I do not know whether William Ruto was correctly declared the winner. That is not evasiveness. It is the appropriate conclusion while a legal challenge is being prepared and important questions about the tabulation process remain contested. Ruto may have won exactly as the declared numbers indicate. Odinga’s challenge may identify genuine legal or procedural failures. The evidence needs to be examined through the mechanisms Kenya created specifically for resolving these disputes.
The more interesting political development is that such mechanisms exist and, so far, the principal contestants are using them.
That sounds unimpressive until Kenya’s recent electoral history is remembered. The disputed 2007 presidential election produced catastrophic violence. More than a thousand people were killed and hundreds of thousands were displaced as political conflict interacted with ethnic divisions and longstanding grievances over land, inequality and state power. Kenya eventually formed a power-sharing government, but the crisis demonstrated what happens when neither political institutions nor political opponents possess enough credibility to resolve disagreement peacefully.
Fifteen years later, another extremely close presidential election has produced accusations of irregularity and a rejected result. Yet the immediate response is not equivalent to 2007. Odinga is preparing a court case.
That difference may be more important than whether Kenya has finally learned how to conduct a completely uncontested election.
Perhaps democratic consolidation should not be measured by the absence of political disputes. Elections involving enormous stakes will always generate disagreements. Candidates will challenge procedures. Officials will make mistakes. Parties will accuse one another of misconduct. Some accusations will be justified and others opportunistic.
The stronger measure may be where those disputes go after they appear.
Do they go to armed supporters?
Do they go to military commanders?
Do they go to party militias?
Or do they go to courts?
Institutions do not eliminate conflict. They domesticate it.
They take political disagreements that could otherwise become struggles over force and convert them into arguments over evidence, procedure and law. That process can be slow, frustrating and imperfect, but the alternative is deciding political truth according to which side can mobilise more coercive power.
Kenya has experienced both models within living memory.
This is why Odinga’s decision to challenge the result legally should not automatically be treated as another politician refusing to accept defeat. Democratic losers have a right to challenge elections when they possess evidence of irregularities. Accepting democracy does not mean accepting every announced result without scrutiny.
The relevant distinction is between challenging an election through the institutions established for challenges and rejecting those institutions whenever they produce an unfavourable answer.
Donald Trump’s behaviour after the 2020 American election illustrates the second problem. Legal challenges were filed and repeatedly failed, yet the broader claim that the election was stolen persisted even after courts and Republican election officials rejected it.
Kenya now faces the institutional test in two stages.
First, can Odinga present his challenge through constitutional procedures rather than treating the declared result as sufficient justification for extra-constitutional mobilisation?
Second, if the Supreme Court eventually rules against him, will he accept the ruling even if he continues believing the election was flawed?
The second is harder.
Anyone can support judicial review while expecting the court to agree with them.
Institutional commitment becomes visible when the institution does not.
Ruto faces an equivalent responsibility. If the court finds serious irregularities and orders corrective action, his supporters would need to accept that decision even though the electoral commission has already declared him president-elect.
Rules have to constrain winners too.
This is why Kenya’s Supreme Court carries an unusually heavy burden. It is not simply resolving a technical election petition. It is converting uncertainty about millions of votes into one authoritative constitutional answer.
Courts can perform that role only when enough people believe their procedures are more legitimate than settling the dispute elsewhere.
That trust cannot be created during one case.
It accumulates through institutional history.
Kenya’s 2017 presidential election matters enormously here. Odinga challenged that result too, and the Supreme Court did something extraordinary: it annulled President Uhuru Kenyatta’s victory and ordered a new election because of irregularities in the electoral process.
That decision demonstrated that judicial review was not merely ceremonial. A court was willing to invalidate the election of an incumbent president.
Whatever one thinks about the legal reasoning, the institutional signal was powerful.
Courts matter when they can rule against powerful people.
A judiciary that only confirms what the executive wants may possess buildings and judges without possessing meaningful independence.
The 2017 decision therefore provides the Supreme Court with credibility that may become valuable now.
But independence does not mean ruling against the government every time.
That would simply create another form of political bias.
The court’s job is to evaluate evidence according to law even if the result disappoints people who celebrated its independence five years ago.
Institutional consistency requires accepting that yesterday’s heroic court may become today’s frustrating one.
This is the difficulty of depersonalising institutions.
Citizens naturally interpret outcomes politically.
If Odinga wins his case, his supporters may celebrate judicial independence while Ruto’s supporters accuse judges of overriding voters.
If Ruto’s victory is upheld, the interpretations may reverse.
The institution succeeds only if enough people can distinguish disliking the judgement from believing the court had no right to make it.
That is a high standard, particularly in a closely divided country.
The electoral commission’s internal dispute makes the challenge harder because institutions normally generate legitimacy through a unified procedure. Four commissioners publicly rejecting the final stage of the process inevitably creates uncertainty. If almost sixty per cent of the commission’s members refuse to endorse the announced result, citizens reasonably want to know why.
At the same time, the commission’s disagreement does not automatically prove that the numerical result is wrong. Institutions can contain internal disagreements over procedure, authority and interpretation while the underlying count remains accurate.
Again, evidence rather than symbolism has to determine the answer.
This is one reason transparency becomes crucial in close elections. When the margin is large, small procedural errors are unlikely to alter the winner. When the margin is narrow, every disputed step acquires greater importance because the probability that an irregularity could matter increases.
Electoral legitimacy therefore depends partly on verifiability.
Citizens do not need personally to count millions of ballots. They need enough access to results, forms, observers and legal procedures that independent actors can verify the aggregation.
Trust is useful.
Verification is stronger.
The best electoral systems do not require citizens to believe officials are virtuous. They construct procedures through which dishonesty becomes difficult to conceal.
This is the same principle underlying good financial institutions, scientific replication and constitutional checks.
Systems become robust when correctness can be tested rather than merely asserted.
Kenya’s electoral commission published polling-station result forms online during the counting process, allowing parties, media organisations and citizens to conduct parallel tallies. That level of transparency can reduce opportunities for central manipulation because the final national total can be compared against thousands of local records.
Yet transparency also creates new political dynamics. Different organisations can produce competing tallies before official aggregation is complete. Social media spreads partial results rapidly. Supporters interpret incomplete information as proof their candidate is winning.
More information does not automatically create more agreement.
It can produce more competing interpretations.
This is another feature of modern democracy that I underestimated previously. Transparency is essential, but information requires institutions capable of establishing which information is authoritative.
Otherwise everyone possesses a spreadsheet and nobody possesses a result.
The ethnic dimension of Kenyan politics also needs careful treatment. Presidential coalitions have historically drawn support unevenly across ethnic communities, and political violence has sometimes followed those divisions. But describing Kenyan elections simply as ethnic censuses would be too reductive.
Ruto’s campaign attempted to organise politics partly around class through his “hustler” narrative, presenting himself as representing ordinary Kenyans struggling economically against established political “dynasties”. That framing is particularly interesting because Ruto himself has been deputy president since 2013 and has spent decades inside national politics.
Like Sadr in Iraq and Khan in Pakistan, he demonstrates how politicians can be simultaneously insiders and anti-establishment figures.
Political outsider status is increasingly less about biography than narrative.
A politician can occupy high office and still define the establishment as somebody else: wealthy families, bureaucratic elites, metropolitan professionals, foreign institutions or political dynasties.
Ruto’s principal opponent made that contrast unusually convenient. Raila Odinga is the son of Jaramogi Oginga Odinga, Kenya’s first vice-president, and was supported in this election by outgoing President Uhuru Kenyatta, himself the son of Kenya’s first president, Jomo Kenyatta.
Ruto could therefore present the contest as the son of ordinary circumstances against two of Kenya’s most famous political families, despite being the sitting deputy president.
That narrative connected with genuine economic frustration. Kenya has experienced significant long-term economic growth, but inequality, youth unemployment, public debt and rising living costs remain important concerns. Global food and energy inflation has added further pressure this year.
Again, aggregate growth and political satisfaction are not identical.
Citizens evaluate whether economic development creates opportunities available to them personally.
A growing economy can still produce anti-establishment politics when upward mobility appears unequally distributed.
This is where Ruto’s “hustler” message becomes politically significant beyond campaign branding. It attempts to shift Kenyan politics from horizontal divisions between ethnic communities towards a vertical division between ordinary citizens and elites.
Whether that actually reduces ethnic politics remains uncertain.
Populist class narratives can coexist with ethnic coalition-building rather than replacing it.
But the attempt itself demonstrates that political cleavages are not permanent.
Politicians can reorganise them.
That is one reason elections matter even when institutions remain imperfect. Political competition creates incentives for candidates to search for new coalitions rather than assuming inherited identities determine every vote.
Kenya’s constitutional reforms after the 2007 crisis also changed those incentives. The 2010 Constitution introduced significant devolution, strengthened institutions and created forty-seven county governments.
Devolution matters because it distributes political stakes.
If almost all meaningful political power exists in the presidency, losing the presidential election becomes catastrophic for political groups.
If counties possess real authority and resources, losing nationally does not mean losing access to every meaningful institution.
This reduces the winner-takes-all character of national politics.
The same logic appears repeatedly across divided societies: distributing power can reduce the cost of losing.
Federalism does this.
Coalition government does this.
Independent courts do this.
Strong local government does this.
Power-sharing does this.
Each mechanism ensures that one electoral defeat does not give the winner control over everything.
That can make government less efficient.
It can make democracy more survivable.
Kenya’s peaceful handling of the current dispute would therefore represent something larger than one successful election. It would suggest that institutional reforms adopted after previous crises are gradually changing political incentives.
This is how institutional development often occurs.
Not through one dramatic constitutional moment after which democracy is suddenly consolidated, but through repetition.
Hold an election.
Dispute it.
Use the court.
Accept the judgement.
Hold another election.
Repeat.
Eventually the extraordinary becomes routine.
Political actors begin anticipating institutional resolution rather than violence because institutional resolution has worked before.
Expectations become self-reinforcing.
This is the opposite of the vicious cycles I have written about in authoritarian systems.
There, distrust produces repression, repression produces resistance and resistance justifies more repression.
Democratic institutions can create virtuous cycles when successful use increases the probability of future use.
The first peaceful transfer is difficult.
The tenth becomes expected.
Expectations are political infrastructure.
They are invisible until they disappear.
Kenya is not guaranteed to succeed. The electoral commission is divided. Odinga has rejected the result. Supporters on both sides care intensely about the outcome. The Supreme Court still has to hear evidence.
There is therefore no reason for complacency.
But neither should uncertainty itself be interpreted as democratic failure.
A close election producing a legal dispute is not evidence democracy has broken.
The relevant question is whether institutions can process the dispute without the political system breaking around them.
That is a much more demanding definition of democratic resilience.
It also changes how we think about political stability.
A stable democracy is not one where everyone believes every election was perfect.
No election involving millions of ballots will be perfect.
It is one where imperfection has somewhere legitimate to go.
A candidate identifies an irregularity.
Lawyers present evidence.
Judges examine it.
A ruling emerges.
The losing side continues participating.
That sequence is much less dramatic than post-election protests or political violence.
Its boringness is the achievement.
Fifteen years ago, a disputed Kenyan election helped produce violence that killed more than a thousand people.
Today, another disputed election is producing legal briefs.
We do not yet know whether the electoral commission got the result right.
We do not yet know what the Supreme Court will decide.
But perhaps democratic development can be measured partly through the fact that the most important question this week is increasingly being asked inside a courtroom rather than answered in the streets.
Institutions do not eliminate political conflict.
Sometimes their greatest achievement is simply changing its location.
