Did Wafula Chebukati cook figures for Ruto? I have found tallying discrepancies in 21 counties that if rectified could lead to a run-off
Yesterday Wafula Chebukati declared William Ruto as the President-Elect, a declaration that was disputed by Azimio led by their presidential agent Saitabao Ole Kanchory. Kanchory’s major contention was that he wasn’t given access to form 34C, the statutory document that ought to be used in the declaration of the President-Elect. Four commissioners of the electoral commission also distanced themselves from the results citing opaqueness towards the end of the tallying of the presidential votes, though they never gave the details of the opaqueness. They have since stated that they will issue a comprehensive statement as to the nature of the opaqueness. These disputes may make one wonder whether William Ruto was rigged in by Wafula Chebukati.
Although Kanchory and the IEBC commissioners who distanced themselves from the results didn’t give specifics as to the nature of their contentions, it is easy to realise that the contention is about the process of tallying the presidential votes. According to the briefing that was given by IEBC shortly before the start of the tallying process, the IEBC chair Wafula Chebukati stated that the Constituency Returning Officers were required to present forms 34As from all the polling stations from their respective constituencies together with form 34B that aggregates the votes from that constituency. Once the forms are presented, the IEBC would take the forms through a verification process that involved checking whether the forms are legit forms that were originally issued by the IEBC, whether those forms match those that were uploaded at the IEBC forms portal, and whether figures in those forms match and tally as expected.
The process of verification ended up with a finding that results from a particular constituency are valid. Once such a finding was made, one of the IEBC commissioners was required to read the results out loud in the presence of the media. The media houses picked these figures upon announcement and tweeted them. We at KachTech Media were following the announcements together with the tweets and tallied the results per constituency. We were able to tally the results from those announcements for 249 constituencies. By the time we were through with doing our tally, Nation Media had managed to tally official results from 253 constituencies, but Francis Gachuri of Citizen while at Bomas reported that IEBC had verified and announced results from 259 constituencies. By the time Gachuri was making that report, IEBC had finalised the verification process, where we expected the IEBC to announce the results of the remaining 32 constituencies, but the announcements were never made.
That’s one cause of alarm. If transparency involves announcing the results of the verified constituencies, why didn’t IEBC announce the results of the last 32 constituencies that were verified? For if such announcements were made and the media houses continued with their tally of official results, we could have had a way of verifying whether the total figures tallied by the media houses matched with the total figures Wafula Chebukati pronounced for each of the four candidates, and thus whether indeed William Smoei Ruto was validly declared as Kenya’s President-Elect following 2022 Presidential election.
The steps of figuring out the discrepancies in the announced vs declared results
Now that we did not get the results from the 32 constituencies, I took it upon myself to verify results from the constituencies that were announced and we tallied as KachTech Media, and compared them against the results that Chebukati had used to declare William Ruto President-Elect. To do this comparison, I first determined the counties that had all the results from their respective constituencies verified. Through this process, I found out that 21 counties had all the results from their constituencies verified. The counties are presented in Table 1. Table 1 also contains votes announced in favour of Raila Odinga and William Ruto, and the totals they obtained from the 21 counties. The last row in the table has the ratio of the total which is basically the ratio of votes for each candidate in the 21 counties compared to the actual total that Wafula Chebukati declared for each candidate when he announced that William Ruto is the President-Elect.
The next step in my process was to find the number of votes Wafula Chebukati gave Raila Odinga and William Ruto in each of the 21 counties shown in Table 1. This was an easy step given that I had already written about the Presidential Results that you can read here. Table 2 provides the votes for Raila Odinga and William Ruto for the 21 counties as declared by Wafula Chebukati.
The next step was to determine if the votes given to Raila Odinga by Wafula Chebukati matched the votes for both candidates as announced by commissioners at Bomas. To check this, I simply subtracted the votes in Table 1 for each candidate from their respective votes in Table 2. If the two vote counts match, then the result of the subtraction is zero. A positive result indicates that Chebukati gave a candidate more votes compared to what the commissioners announced, whereas a negative result indicates that Chebukati subtracted some votes from a candidate’s votes. The results of this operation are presented in Table 3.
The total section in Table 3 provides the total votes added or removed by Chebukati for each candidate in the 21 counties that I have analyzed. As you can see, Chebukati added Raila a net13,216 votes. A significant addition for Raila happened in Nyandarua county. According to tallying done using figures announced by the commissioners, Raila had 33,869 votes in Nyandarua county, but Chebukati gave him 49,228 votes. The other major impact for Raila can be seen in Kiambu and Migori counties, where Chebukati substracted 8,954 votes in Kiambu and then added him 5,345 votes in Migori.
For William Ruto, Chebukati subtracted his 9,700 votes in Kirinyaga county, added to him 2,617 votes in Nyamira and added to him a further 61,966 votes in Nyandarua. The net result is that William Ruto got an additional 57,987 votes in the 21 counties.
The next step was to question whether these extra votes both for Raila and Ruto could have an impact on the outcome of the election, given our constitutional requirement that a candidate must obtain 50% plus 1 vote in order to be declared President-Elect. To do this, I asked myself if prorating the votes to the rest of the counties could yield a significant number of votes that could affect the outcome of the election. The best way to do this was to assume that each vote added was impacting a per cent of votes that each of the two candidates had obtained. To estimate this percentage, I divided the number of declared votes (6,942,930 for Raila and 7,176,141 votes for Ruto) by the number of total votes the candidates got in Table 1 votes (2,061,819 votes for Raila and 3,642,1687 votes for Ruto) to obtain the ratio of 0.297 and 0.5075 for Raila and Ruto respectively. I then added 1 to these ratios (100%) to obtain 1.297 and 1.5075 respectively and then multiplied the added votes with the new figures. That’s how I obtained the prorated votes of 17,141 and 87,415 potential votes that were added to Raila and Ruto respectively.
Assuming that the rate at which Chebukati added votes for Raila and Ruto remained the same in the other 27 constituencies, then we should expect that the final figures for Raila (6,942,930 votes) have an extra 17,141 votes whereas the final figures for Ruto (7,176,141 votes) have an extra 87,415 votes. If we remove these votes from both candidates, then they would remain with 6,925,789 votes in favour of Raila and 7,088,726 in favour of Ruto. These votes against the total valid votes of 14,213,137 work out to be 48.73% for Raila and 49.87% for Ruto; hence Ruto did not reach the 50% plus 1 vote constitutional requirement.
The question of form 34As found at the IEBC portal
One can argue that the reason we have discrepancies between the votes read by the commissioners and those declared by Chebukati is that the commissioners read their figures from verified form 34Bs, whereas Chebukati used form 34As at the portal to do his tally. Now one must be cognizant of the fact that form 34As found in the IEBC portal are not the legal statutory documents against which official results ought to be derived. The official statutory form recognised by law is the physical form 34As that the returning officers were required to present for verification at Bomas. Once these forms were verified and were found to tally with form 34Bs, then only those verified figures were expected to be used to generate form 34C (by the way no one has seen form 34C as of writing this article). Since some of the uploaded form 34As are unreadable, and given that no one can verify their authenticities (you can’t detect any of the security features that the physical forms are expected to have by downloading them), then any results generated by the use of the digitally uploaded form 34As cannot be the official legally binding results.
My conclusion is that the discrepancies I have found by analysing results from 21 counties are significant enough to warrant challenging the results at the Supreme Court. By the way once we get the tallies that were announced at Bomas for all the constituencies, we will update this article with the figures.
