The fallout from the 2017 General Election is only now starting to become clear.
At the time of writing, debate about the quality of the polls continues to rage, and National Super Alliance (Nasa) presidential candidate Raila Odinga has confirmed that he will take his party’s complaints to the Supreme Court.
As is usually the case in Kenyan elections, the controversy has focussed on who won the presidential poll.
But what of the Independent Electoral and Boundaries Commission (IEBC): has it been a winner, or a loser?
The August 8 election seemed to start well, with the vast majority of the Kenya Integrated Election Management System (Kiems) kits working.
Talking to voters in Nairobi’s Kibera and elsewhere, it was clear that the technology boosted public confidence in the process, at least in the short term.
The situation began to look better still when the kits began to transmit data, generating a live stream of the results.
However, after this promising start a number of poor decisions and communication failures called into question the IEBCs credibility.
Most notably, the failure to provide copies of the 34A forms that confirm the result at the polling station level quickly changed the narrative from one of success to one of suspicion.
It may be that we subsequently find out there was nothing much wrong with these forms, but even in this case the IEBC has a case to answer because the way in which it handled the situation contributed to opposition leaders’ belief that the electoral process had been undermined.
In the run up to polling day, Jubilee and Nasa — the two frontrunners — competed to sway the hearts and minds of voters.
At the same time, the opposition and the electoral commission engaged in a struggle of their own.
On a number of issues, including whether polling station results should be considered final, the IEBC contested the position of the opposition, often in a court of law.
This created the unfortunate impression that the IEBC was anti-reform – especially when rumours emerged that the Commission’s decision to contest the pre-eminence of polling station results was prompted by advice from close allies of the President.
Against this backdrop, the sad death of senior IEBC information technology official Chris Msando further undermined the confidence of the opposition.
Whatever the motivations for his murder, it is clear that he was one of the more capable and independent minded figures within the Commission.
As a result, Msando’s absence raised serious questions about whether the IEBC had the capacity and determination to ensure that the technology would work.
Along with the perception that he had been killed to allow the election to be “hacked”, the opposition confidence in the broader electoral process was undermined.
Given Msando’s death just days before the polls, and the fact that the commissioners were only appointed earlier this year, the IEBC did remarkably well when it comes to the basic logistics.
Polling stations were generally able to open in time and although voting went very late in some parts of the country, in the vast majority of areas it was completed in a timely manner.
Moreover, in stark contrast to 2013, the technology initially appeared to be working well.
The Elections Observation Group of domestic monitors estimate that Kiems kits were present in 99.3 per cent of polling stations and worked in over 92 per cent.
In conversations with a network of around 20 people visiting polling stations across the country, I only heard of kits failing to verify voters in a couple of cases, and in both instances the machines were either fixed or replaced.
Despite a disappointing “test” of the vote transmission system before the election, the results transmission system also seemed to function well at first.
As a result, my early evaluation, like many of the international election monitors, was largely favourable.
In particular, I was impressed by the way in which the IEBC’s website allowed interested parties to break down the results by county, constituency and polling station.
Having spent the last two Kenyan elections desperately trying – and failing – to find the official set of polling stations results, this was a welcome improvement.
Indeed, it is worth noting that without this website, we would not have much of the data that I use in this column.
Whatever else is changed about the electoral process as the country moves towards 2022, this is one innovation worth keeping.
Sadly, the sense that this was a particularly well-organised election did not last long.
As the hours ticked by, the absence of thousands of the form 34A meant that it was not possible to verify the data that was being shown online.
Although the IEBC continually pledged to rectify this, analysis of its own portal demonstrated that little progress was being made.
It was therefore surprising – and deeply unhelpful – when the commission decided to declare that Jubilee’s Uhuru Kenyatta had won the election before putting its house in order.
The announcement of the final results suggested that one of two things had happened.
Either results were aggregated at the constituency in the absence of the 34A forms, in which case the 34B forms that were tallied at the constituency to generate the presidential result rested on shaky foundations, or the IEBC had the forms in its possession and had failed to make them available.
Either option is a significant cause for concern.
The former implies that the IEBC announced the result before it could be sure that it was correct.
The latter implies that the IEBC was reluctant to make available one of the documents in its possession that was causing the greatest controversy, and in so doing failed to take all possible steps to reduce the tension and confusion around the election.
The limitations of the IEBC’s actions were compounded by its poor communication.
In part this reflects a lack of preparation – the winning bidder for the job of providing public relations and media services to the commission, Scangroup, is reported to have only been officially contracted days before the election. But it also reflects a failure to identify and respond to the most pressing issues relating to the credibility of the polls.
IEBC commissioners are not appointed for their communication skills, of course, and it is not surprising that they are not always the most fluent or effective public speakers.
What is more worrying is the failure of the commission to explain basic aspects of the electoral process and to provide convincing explanations for its shortcomings.
The IEBC’s statements on the missing 34A forms is a case in point.
At a number of points during the counting process, it was intimated that almost all the 34A forms had been received and were being put online only for them to fail to materialise.
This was not just relayed via social media, but also in personal communications to key stakeholders.
Yet at the time of writing – the evening of August 17 – thousands of forms are still missing from the IEBC website.
With the best will in the world, it is hard to avoid the conclusion that we were deliberately misled.
The IEBC’s explanation for the absence of the forms was also lacking.
At various points, signal strength, the courts, and the fact that some files were sent as PDF files rather than as JPEGs were blamed for the delays, but none of these explanations really stacks up.
Signal strength may have prevented images from being sent directly from polling stations, but this is no explanation for why it was not possible to bring a hard copy of the form – or the kits themselves – to Nairobi.
Similarly, while it is true that the courts directed the IEBC to consider results announced at polling stations and constituencies to be final, they said nothing about the need to keep all copies of the forms at this level once they had been aggregated.
And finally, receiving an image in the wrong file type may represent a small glitch, but this should not happen with the Kiems kits and even if it does it should be straightforward to convert images from one format to another – I had a go just before writing this article and it took me five minutes to do this 20 times, and I wasn’t rushing.
The lack of a convincing technical explanation for the absence of 34A forms is problematic, because it raises serious questions about the willingness of the IEBC to run a transparent election.
Partly as a result, the opposition now faces the challenge of putting together an election petition without all of the data necessary to do so.
In this way, the commission’s failings also have significant implications for the credibility of the Supreme Court process, and the likelihood that the opposition will accept the verdict if its case is unsuccessful.
Of course, this conclusion is by necessity a preliminary one.
On the one hand, the forms may soon be provided, and they may have minimal errors.
On the other, the election petition may raise further issues, pushing the issue of the 34A forms into the background.
But as things stand, this seems to have been an election in which the electoral commission snatched defeat from the jaws of victory.
Nic Cheeseman (@fromagehomme) is Professor of Democracy at the University of Birmingham