Presidential petition: Your math is wrong, statistician tells Chebukati

The presidential election results announced by the Independent Electoral and Boundaries Commission (IEBC) chairman Wafula Chebukati are mathematically incorrect, a university lecturer claims.

According to Mr Chebukati, William Ruto garnered 7.1 million votes, translating to 50.49 per cent of the cast votes, followed by Raila Odinga with 6.9 million, which is 48.85 per cent.

The other candidates George Wajackoyah and David Mwaure received 61,969 votes (0.44 per cent) and 31,897 (0.23 per cent). However, Dr Edgar Otumba, a Maseno University statistician, claims Mr Chebukati got his presidential election math wrong.

His arguments agree with those contained in a petition by Busia Senator-elect Okiya Omtatah, Wycliffe Nyakina, Victor Okuna and John Maina, who also insist they came to the conclusion the results released were not adding up.

He is of the view neither Dr Ruto nor Mr Odinga garnered 50 per cent plus one vote as required by the law. The don bases this argument on the voter turnout and the total valid votes cast.

Dr Otumba says the IEBC tally has 140,028 missing votes, which are vital to either sway the vote for a runoff or consolidate the win for Dr Ruto.

According to him, if all the missing votes were cast in Azimio's favour, then Mr Odinga would have had 7.08 million votes, translating to 49.3477 per cent and denying UDA a win as Dr Ruto would have 7.17 million or 49.9969 per cent.

Alternatively, if the 140,028 were all given to the Deputy President, the statistician argues, Dr Ruto would have garnered 7.31 million votes (50.9725 per cent) and Mr Odinga would retain the 6.94 million votes (48.4721 per cent).

If the same were given to Mr Mwaure or Prof Wajakoya, then both Mr Odinga and Dr Ruto would not reach the 50 per cent plus one vote threshold. Dr Otumba states that by 4pm on August 9, the IEBC had indicated that the voter turnout was 56.17 per cent. At the close of voting, the commission said the number had increased to 14,466,779 voters, or 65.4 per cent.

He notes that Form 34C indicates the cast votes were 14,213, 137 and the rejected votes were 113,614, hence a total of 14,326,751.

According to him, the figures released by Mr Chebukati show the voter turnout on the Kiems kit is higher than the voter turnout recorded in Form 34C, with a difference of 140,028 votes.

"This can be discerned from subtracting the total number of valid votes cast as recorded in Form 34C from the number of voter turnout based on the Kiems kit. The mathematical calculation is 14,466,779 - 14,326,751 = 140,028.

"It was incumbent upon the second respondent to inform the public about the whereabouts of the 140,028 votes plus those who were identified using a manual register."

According to him, if it will be argued that 140,028 votes were spoilt, then the number ought to rise to 253,642. He adds that the number of those identified manually cannot also be accounted for. Dr Otumba says if the 140,028 votes are added to the tally that Mr Chebukati announced, it would translate to 0.9679 per cent of the total valid votes for the candidates and would not have summed to 100 per cent.

"None of the candidates would have crossed 50 per cent plus one mark. This is where the multiplier effect came up so as to enable the top candidate to cross the 50 per cent mark and to correct the percentage back to 100 per cent," he argues.

The don's calculation started with the known number that Kenya has 22.1 million voters. According to him, turnout brings out a different total from what Mr Chebukati declared.

He also wades into the split between Mr Chebukati and Vice-Chairperson Juliana Cherera, commissioners Francis Wanderi, Justus Nyang'aya and Irene Masit on the 0.01 per cent difference in the final tally.

According to him, the issue is not about whether 0.01 per cent is a percentage or not, but rather the number of votes represented by the factor of 0.01.

He states that if 0.01 is multiplied by the 14,214,137 it would amount to 142,131 votes. This, he says, could be thought to be the missing 140,028 votes or the ones identified manually.