Lessons from IEBC's handling of 2017 poll

IEBC chair Wafula Chebukati (left) hands President Uhuru Kenyatta his certificate after declaring him the winner of the 2017 poll. [File, Standard]

There will be profound changes in 2020 in the management of elections in Kenya.

It will be a time to be candid and brutal about the pain points in our electoral process. We will not make any meaningful changes if we shy away from the problems and solutions.

Undeniably, the biggest headache with elections in Kenya has been the management of presidential results. The historic nullification of the presidential results in 2017 was the Independent Electoral and Boundaries Commission's (IEBC) waterloo. Since then, it has been the motivation for reforms.

More thought needs to go into the legal requirement for presidential results to be transmitted electronically. The unintended consequence of this requirement is the shifting of focus and definition of results from the paper ballots to electronic copies. A ballot slip bears the critical mark and decision of the voter. Poll officials thereafter count, scan and transmit such decisions to result portals, better known as IEBC servers. When this procedure is done, the significance of ballot papers wanes.

Missing forms

Could the substituting of the primary source of decisions by the voter with decisions mediated by technology be our undoing? The predilection for what technology offers should not obscure the importance and authenticity of the marked ballot paper. The veracity of election results can only be known if there are ballot papers and result forms that confirm the electronic results.

Had the IEBC’s returning officers delivered to the National Tallying Centre at Bomas of Kenya the hard copy result forms 34As and Bs, which they had scanned and sent, the issue of missing forms during the August 2017 elections would not have come up. That electronic copies were missing was certainly interpreted to mean that there were no results at all.

And we all know why they did not take the hard copy results to Bomas. The ruling of the petition number 207 of 2016 (Maina Kiai & others Vs IEBC) dissuaded IEBC from taking hard copy results to the Bomas of Kenya. The court had reinforced the finality of results at polling stations and even questioned “the spectacle of all the 290 returning officers from each constituency and 47 county returning officers trooping to Nairobi by whatever means of transport, carrying in hard copies of the presidential results which they had announced at their respective constituency tallying centres”.

The election law says the commission must ensure that the technology it deploys is simple, accurate, verifiable, secure, accountable and transparent. But how will such desirable attributes be attained without comparisons with primary documents? Not many people understand technical jargons of servers, logs, algorithms etc. They should be afforded the opportunity of understanding the process from simplicity and familiarity of physical documents.

Translate results

The IEBC should transmit and submit results. This is not to suggest a parallel results paths. It is to provide safeguards that strengthen credibility. A poll official bent on some mischief will think twice on the realisation that people will share on smartphones photos of the results posted at the doors of counting centres. If the enduring fear is that of manipulation of the outcomes, then availability of source documents is the cure.

Another critical measure is to undertake a postmortem of the ballot papers that are kept, undisturbed, in boxes for at least three years, unless a court rules for a recount during a petition. Advanced democracies undertake a study popularly known as Diagnostic Count Replication (DCR). A replication study may confirm original findings or uncover potential problems, but it cannot be used to nullify results nor reinforce credibility. It is simply a post-election audit of results intended to inform policy.

The commission could and should engage an independent researcher to undertake a DCR research on votes cast. This will, for example, help in the understanding of the big problem of spoilt votes. Technology mediates human functions. It does not originate or substitute our values and actions. With the rise of tech-manipulation of computer systems, aren’t we safer with old paper trails? 

Mr Limo is a former IEBC employee and researcher on computer mediated communication