Could the Kericho by-election have been rigged electronically?

Tuesday March 29 2016

More by this Author

Now that the emotions around the Kericho senatorial elections have subsided, it may be a good time to interrogate the question of whether the elections could have been rigged using electronic means.

Kanu, which presented a formidable challenge to the ruling Jubilee coalition, made a startling claim. The Independent Electoral and Boundaries Commission (IEBC), they said, had rigged the elections through some "geometric progression" formula.

Kanu claimed that IEBC had tweaked the Results Transmission System (RTS) in favour of the winning candidate, using a formula which multiplied each of the KANU vote by a factor of 1.6 in favour of the Jubilee candidate.

This, it was claimed, was a breakthrough discovery, cracked by an elite team of Kanu statisticians whose services had blown open another secret plot designed for use in the upcoming 2017 elections. Or so the story went.

Social media went wild with these “revelations”.  The usual suspects cum digital robots from both sides of the political divide defended or trashed the KANU revelations, without too much interrogation.


Those who attempted to take a deeper look at the Kanu allegations were quickly labelled "sellouts" - particularly if their second name dictated they were supposed to be blindly singing from the ‘opposition’ song book.

It was clear from the arguments that the role of ICTs in the election process was not understood, and many critics could have benefited from reading on this blog about ICT in elections.

Kanu’s ‘geometric progression’ argument has a fundamental flaw. It blatantly ignores the link between what the RTS displays and what the political party agents physically sign off on at polling stations.

At each polling station, there is a tallying process in which the votes are counted and agents from the competing parties sign against the documented tally (Form35).

This signed tally is what each presiding officer sends to the results transmission system (RTS) which consolidates and displays it in real time. Clearly, even if the geometric progression "bug" was real, the signed Form35 would quickly pick the discrepancy and arrest the situation. 

Kanu, of course, knows this, which is why they were quick to promise their supporters that they would not be going to court to pursue their geometric theory to its logical conclusion.

To rig elections electronically, one must find more sophisticated approaches – the geometric progression formula is simply not up to the mark.

A simple, yet very sophisticated approach would be possible if a suitable time lag existed between one display of the RTS and what can be done in between to influence the next display.


In very tight elections, one can always boost their tally if a particular polling station has a bunch of “cooperative” or “willing” agents and presiding officers.  How?

Once counting has been done, each polling station knows the balance of voters who registered but did not actually show up to vote. This pool of absentee voters can easily be “made to vote” for one candidate or the other, and subsequently incorporated and signed into Form35.

This “spiked” data is then sent to the RTS for display. In other words, one can still beat the electronic RTS. This approach ensures that the RTS and the Form35 are synchronised and avoids the possibility of a candidate’s victory being flagged as fraudulent. 

However, there are other scientific exercises that can still be applied to test the statistical reliability of such a legally valid, but spiked outcome.

Benford’s Law, which measures how closely a series of outcomes or tallies conform to naturally occurring random numbers, is one of them.

To conclude, the Kericho by-elections were definitely not rigged according to the sweet-sounding geometric progression formula. That is not the same as saying there was no rigging, however.

Opportunities and vulnerabilities for rigging elections still exist but each party is seemingly trying its best to exploit, rather than eliminate the same. This remains an attractive, but very dangerous, dance with destiny as we move into 2017.

Mr Walubengo is a lecturer at the Multimedia University of Kenya, Faculty of Computing and IT. Email: [email protected], Twitter: @jwalu