[2023] KEHC 816 (KLR)

Ayiera v Kimwomi & 3 others (Election Petition E002 of 2022) [2023] KEHC 816 (KLR) (16 February 2023) (Judgment)

Neutral citation: [2023] KEHC 816 (KLR)

Republic of Kenya

In the High Court at Nyamira

Election Petition E002 of 2022

K Kimondo, J

February 16, 2023

Between

Dennis Omwenga Ayiera

Petitioner

and

Nyaribo Amos Kimwomi

1st Respondent

James Gesami

2nd Respondent

County Returning Officer, Nyamira

3rd Respondent

IEBC

4th Respondent

Judgment

Introduction 1. Dennis Omwenga Ayiera (hereafter the petitioner or PW1) was a registered voter at Nyakongo D.O.K. Primary School Polling Station in Manga Ward, Kitutu Masaba Constituency within Nyamira County.

2. He was also the County Elections Manager for the United Democratic Alliance Party (hereafter the UDA Party) for one of the gubernatorial candidates, Walter Osebe Nyambati (hereafter Nyambati).

3. The petitioner challenged the election of Nyaribo Amos Kimwomi and James Gesami (hereafter the 1st and 2nd respondents) as the governor and deputy governor respectively for Nyamira County in the general elections held on 9th August 2022.

4. The 3rd respondent (also referred to as DW1) was employed by the Independent Electoral and Boundaries Commission (hereafter the 4th respondent or the IEBC) as the County Returning Officer in the impugned poll. The IEBC on the other hand is a creature of Article 88 as read with Article 248 (2) (c) of the Constitution; and mandated, among other functions, to conduct and supervise elections.

5. There were 12 candidates who vied for the position of governor. On 12th August 2022, the 3rd respondent declared that the 1st respondent won the election by garnering 81,980 votes against his closest challenger, Nyambati, who polled 49,339 votes.

Pleadings. 6. The petition was filed on 7th September 2022 and well within the time stipulated by section 76 (1)(a) of the Elections Act (hereafter the Act). On 9th September 2022, and still within time, the petitioner filed an amended petition but without leave of the court.

7. On 12th October 2022 the court granted leave to file and serve the amended petition to the intent that the amended petition be deemed to have been filed with such leave and to have been served within time.

8. The amended petition was largely predicated on the affidavit of the petitioner sworn on 9th September 2022 together with several annexures. It is instructive that the original petition had been filed with several witness affidavits and other documents contained in volumes 1 to 5.

9. The 1st and 2nd respondents were served by advertisement in The Standard newspaper of 12th September 2022. However, they failed to file a response within the 7 days prescribed by Rule 11 (1) of the Elections (Parliamentary and County Elections) Petition Rules, 2017 (hereafter the Petition Rules or Rules).

10. By a considered ruling (Ruling No. 2) dated 12th October 2022, I extended time for the 1st and 2nd respondents to file and serve their response dated 23rd September 2022 together with the two annexed affidavits of Nyaribo Amos Kimwomi and Richard Okiega Nyabate which were now deemed to have been filed and served within the required time. However, the application by the 1st and 2nd respondents to extend the time for filing another lot of 23 affidavits was disallowed.

11. The 3rd and 4th respondents were also served by substituted service but equally failed to respond. On 12th October 2022 the court extended time within which they ought to have filed their response to the petition, replying affidavit and other witness affidavits “to the intent that they will be deemed to have been filed and served within the required time”.

12. The court however dismissed an additional prayer by the 3rd and 4th respondents for a further 3 days leave to formally file other witness affidavits annexed as “drafts” to the above motion.

13. I should add that the response by the 3rd and 4th respondents was accompanied by an affidavit of the 3rd respondent together with 4 volumes containing other witnesses’ affidavits. On 7th November 2022, the IEBC filed certified copies of the results declaratory forms.

Petitioners Case 14. There were 643 polling stations or thereabouts in Nyamira spread out in four constituencies. In a synopsis, the petitioner’s case is that the elections for governor were not conducted in compliance with Articles 81 and 86 of the Constitution, the Act and the Regulations. In particular, he contended that the IEBC did not conduct a transparent, impartial, neutral, efficient, accurate or verifiable poll.

15. There are five key complaints pleaded at paragraph 15 of the amended petition as well as paragraph 10 of the supporting affidavit-a.That there were two differing sets of results declaration forms 37C and it is doubtful which of the two was used to announce the results. A related complaint is that UDA agents or those of Nyambati did not participate or execute the form.b.That the results in forms 37C did not tally or correspond with a number of forms 37A from the polling stations.c.That the agents from UDA Party were either denied access into the polling stations, were harassed or did not participate in the counting and tallying of results. Those issues are expounded at paragraphs 125 to 131 of the supporting affidavit;d.That the results in forms 37A were altered or manipulated by officials of the 4th respondent in favour of the 1st respondent and another candidate, Timothy Bosire;e.That employees of the 3rd and 4th respondents were biased against the other candidates and favored the 1st respondent and Bosire.

16. There is an additional complaint at paragraphs 161,162 and 163 of the amended petition relating to criminal conduct and what the petitioner terms a “corrupt practice” in breach of sections 14 and 15 of the Elections Offences Act. He pleaded that the Chief Agent of the 1st respondent, one Leonard Okari Mogaru, was an employee of the County Public Service Board. A Compact Disk marked DOA1 containing four short MP4 video clips was played in court showing Mogaru at the tallying centre wearing a red and white shirt emblazoned at the back: Mogaru Supports Governor Nyaribo.

17. The petitioner also complained that votes were deducted in various polling stations in the four constituencies of Nyamira county from Nyambati and added to 1st respondent and another candidate, Timothy Bosire. Examples of such discrepancies are listed in paragraphs 28 to 143 of the amended petition and the annexed forms and table.

18. There were other categories of complaints. For instance, the petitioner alleged at paragraph 135 of his affidavit that a ballot box for Nyanchoka TBC Polling Centre “was found open at the tallying center without reasonable explanation”. I should point out that there is a photograph and a certificate of electronic record of Yabesh Mokua Ntabo in petitioner’s Volume 2. However, the witness did not take to the stand and the photograph was never admitted in evidence.

19. It was also pleaded that a ballot box for a polling station at Omonono Primary School was found in a different constituency. A witness Kinaro Ndubi (PW13) claimed that it was found at Kebabe Girls. An analogous complaint was made by Antony Miranyi (PW11).

20. I will give a more detailed summary and analysis of the testimony of the petitioner and his 13 witnesses later.

Reliefs Sought 21. The Petitioner seeks five prayers:i.Order for scrutiny of all election materials used in the elections;ii.Declaration that the 1st respondent was not validly elected as governor of Nyamira county;iii.Order for fresh elections;iv.That the 3rd respondent and officers of the 4th respondents committed electoral malpractices of criminal nature; andv.Costs of the petition.

1st and 2nd Respondent’s Case 22. The amended petition is strenuously opposed. The 1st and 2nd respondents deny all the allegations by the petitioner in toto. In a nutshell, they contend that the elections were conducted in a free, fair, accurate and verifiable manner; and, that the 1st and 2nd respondents were validly elected.

23. In the alternative, they argue that the malpractices or irregularities alleged by the petitioner did not substantially affect the results of the election. They thus prayed that the amended petition be dismissed with costs. As I will discuss later, the 1st and 2nd respondents opted not to testify or call any witness.

3rd and 4th Respondents Case 24. The response by the 3rd and 4th respondents can be summarized as follows: That the IEBC conducted the elections in accordance with the Constitution, the Elections Act and Regulations thereunder. They also challenged the authenticity of some of the results declaration forms annexed to the petition or witness affidavits.

25. In reply to the specific allegations by the petitioner, the 3rd and 4th respondents aver that there was only one valid form 37C that was used to declare the results of the elections which appears at page 141 to 150 of volume 4 of the IEBC bundle; and, that the other form 37C referred to by the petitioner or some of his witnesses was a “draft”.

26. Regarding alterations to forms 37A, the 3rd respondent (DW1) conceded that there were some changes which were necessary to rectify certain errors. He testified that some of the errors were typos or “transpositional errors” which skewed data to the wrong columns on forms 37B or 37C. He explained that the errors occurred when transferring data from forms 37A to 37B or 37C. He gave examples at pages 110, 130 and 151 of volume 1 of the IEBC bundle.

27. He downplayed the effect of the errors stating that when they are all computed, Nyambati was denied 432 votes while the 1st respondent only gained 17 votes. The votes of all the other candidates were unaffected. In a nutshell, the 3rd and 4th respondents’ case is that the errors did not affect the substance of the election.

28. Regarding the complaint over the agent known as Okari who was alleged to be a member of the County Public Service Board, DW1 stated that it is the candidates who appoint their agents; and, that there was nothing irregular about the agent wearing his party colors at the county tallying center.

29. Lastly, the 3rd and 4th respondents denied the allegations that Nyambati’s agents were harassed, denied access at polling stations or ejected from there. They contended that in the majority of cases, the results declaration forms are signed by agents from the UDA party.

30. I will deal in more detail with the evidence of the 7 witnesses called by the 3rd and 4th respondents below.

Detailed oral and documentary evidence 31. The petitioner (PW1) and his 13 witnesses testified in this matter. He made extensive reference to his supporting affidavit. The deposition largely mirrored the main grounds set out in the amended petition save for the annexures. One of his key complaints was that the 3rd respondent prepared two conflicting forms 37C. He contended at paragraph 22 that the second form was “filled in the absence of the UDA agent”. In his view, the results were thus not transparent, accountable or verifiable.

32. He also dwelt at length with alterations or cancellations in entries in forms 37A as well as errors in transferring data from the forms into forms 37B and eventually into the impugned form 37C. He contended that the votes for Nyambati were systematically reduced “in over 100 polling stations” while those of the 1st respondent and another candidate, Bosire, were increased. According to the petitioner, it was a well-choreographed scheme to subvert the will of the people.

33. The witness highlighted about 38 forms with “multiple alterations” in either the entries for the votes of candidates, the total number of valid votes, the registered voters or rejected ballots. He pointed out disputed returns or complaints at North Mugirango Constituency at paragraphs 22 to 62 of his affidavit; in West Mugirango Constituency at paragraphs 63 to 102; in Kitutu Masaba Constituency from paragraphs 103 to 116; and, from Borabu Constituency at paragraphs 113 to 123. A table containing his analysis was annexed.

34. He testified that UDA agents or those for candidate Nyambati were not allowed at a number of polling stations or were thrown out at some point. At paragraph 136 of the amended petition, he claimed that 138 stations, did not have UDA or Nyambati’s agents “due to discriminative restrictions imposed upon them by the 3rd and 4th respondents”. He listed 39 polling stations in Borabu; 64 in West Mugirango; and, 42 in North Mugirango.

35. He testified that some of the agents were allowed into the stations late. This was expounded by Elvis Simba Kibeni (PW10), the chief agent for Nyambati in Kitutu Masaba Constituency. The latter testified that at Nyaisa SDA Primary School Polling Station 1, Manga TBC Polling Station 2 and Mokwerero DEB Polling Station 1, the presiding officers were “so biased that at some point, some of our agents were ejected”.

36. The petitioner deposed in paragraph 135 that a ballot box of Nyanchoka TBC “was found open at the tallying station without any reasonable explanation”. Like I stated, there is a photograph and a certificate of electronic record of Yabesh Ntabo in Volume 2. However, the witness did not take to the stand and the photograph was never admitted in evidence.

37. The petitioner also claimed that in some cases, for instance at Motagara 2 of 2, agents were given differing forms 37A. In one of the forms 37A there, Nyambati has 88 votes but only 1 vote in the other. He blamed the IEBC for its casualness and opaqueness; and, that granted all the circumstances, the election was not conducted in an accurate or verifiable manner.

38. PW2 was Divinah Moraa, an agent for UDA at Kerongo DOK primary school. Her material evidence was that in form 37C, the 1st respondent is given 150 votes while in truth, he only garnered 115 votes. As I will discuss shortly, that fact was conceded by the IEBC at page 119 of the response.

39. Alfayo Moire (PW3) was an agent at Kenyamware DEB Primary School. His complaint is that that form 37C does not correspond with results in forms 37A. For instance, Timothy Bosire got 86 votes but form 37C reflects 38; John Nyariki got 2 votes but in 37C he is given 4; Arnold Nyasiga got 19 votes but in form 37C he is awarded 10; and, Nyambati got 280 votes but form 37C reflects 220.

40. Joseah Omariba (PW4) was an agent at Riabaruoti polling station 1. His key evidence is that Nyaribo garnered 105 votes as per form 37A but was added 10 votes in form 37C.

41. PW5 Hyline Isoe was an agent at Tinga DOK polling station. She testified that Nyambati received 106 votes as per form 37A but in form 37C grants him 1 vote. This is admitted by DW1 at page 191 as a “trans positional error”.

42. Andrew Ombogo (PW6) was an agent for UDA at Nyamonge Primary School Polling Station 1. He stated that Nyambati got 36 votes as per form 37A but zero in 37C. The form is annexed to his affidavit at page 433.

43. PW7 Evans Moruri, was an agent at Nyasumi DOK Polling Station 1 where Nyambati garnered 102 votes. However, in form 37C he was given only 2.

44. Chrisantus Nyakundi (PW8) claimed that the Presiding Officer at Nyaisa SDA Primary School Polling station 1 “was assisting some voters alone without involving the agents” and that when he complained “she confiscated [his] badge and threw [him] out”. PW10 intervened on his behalf and he was allowed back. PW8 testified that he was later ejected and was not allowed back until 23:00 hours by which time the count had ended.

45. Wesley Kerina (PW9) was an agent at Motagara DEB Polling Station 2. He testified that Nyambati got 88 votes. He said he was present at the count and signed form 37A. However, the candidate is only assigned 1 vote in form 37C. I observed that the image of the form annexed at page 399 of his affidavit is too faint and illegible.

46. Elvis Kibeni (PW10) was Nyambati’s Chief Agent in Kitutu Masaba Constituency. He testified that on the election day he went around to ensure that all the agents had reported. At paragraph 9 and 10 of his deposition, he avers that Nyambati’s agents were chased away from polling stations; and, that he had to intervene for them to be admitted back. However, he conceded that he did not attach a list of those witnesses. He also conceded that he did not specify the names of agents who were denied access.

47. In his oral testimony he confirmed that he visited Nyaisa SDA Polling Station and intervened to have PW8 to be re-admitted. He found PW8 outside. PW10 talked to the Presiding Officer called Happiness who agreed to readmit PW8. PW10 also intervened at the nearby Manga Tea Collection Centre to have Nyambati’s agents allowed into the station. He claimed that at this station, a polling clerk was telling people to vote for Nyaribo. I note that he did not name or identify the offending polling clerk.

48. His other material evidence was on discrepancies in results from a number of polling stations between forms 37A and form 37C. For instance, at Nyangorora DEB Polling Station 1, Nyambati had 131 votes in form 37A but zero in form 37C. In Bocharia PAG Primary School Polling Station 1, Candidate Bosire had 170 votes in 37A but was given 174 in form 37C; Nyaribo had 18 votes in form 37A but awarded 47 in form 37C; Nyambati had 142 votes but they were reduced to 109 in form 37C.

49. Regarding the neighbouring stream 2 at Bocharia, he testified that Bosire had 184 votes, the 1st respondent had 25 while Nyambati had 121. But in form 37C Bosire was given 193 votes, the 1st respondent 48 and Osebe 89.

50. The witness said that at Nyaibasa DOK Polling Station 1 of 2, Bosire had 40 votes, Nyaribo 0 and Nyambati 136 but in form 37C Bosire was granted 90, Nyaribo 33 and Osebe 86. In station 2 of 2, Bosire had 59 votes, Nyaribo 40 and Nyambati 98. However, in form 37C Bosire was given 79 votes, Nyaribo 41 and Nyambati 77.

51. He said that at Kiomakondo DOK Primary School 1 of 2, Bosire obtained 40 votes, Nyaribo 45 and Nyambati 135. But in form 37C the score for Bosire was increased to 59 votes, Nyaribo’s to 59 while Nyambati’s was reduced to 102 votes.

52. He referred to a similar trend at Kiamokondo station 2 of 2, where Bosire had 30 votes, Nyaribo 41 and Nyambati 121 votes. But in form 37C Bosire was given 54 votes, Nyaribo 56 and Nyambati 82. In Kiabiraa DEB Primary School, Bosire was added 4 votes in form 37C, Nyaribo an extra 13 while Nyaribo’s were reduced by 89 votes. In Kiabiraa DEB station 2, Bosire was added another 4 votes, those of Nyaribo were increased by 23 and while Nyaribo’s were reduced by 46.

53. The witness testified that at Mobamba TBC Polling Station 1, Bosire had 131 votes which ballooned to 160 in form 37C. Nyambati’s were reduced from 124 to 95. In stream 2, Bosire’s score was 109 but was enhanced in form 37C to 147 while Nyambati’s was reduced from 132 to 94.

54. The witness made similar observations at Kiamogiti Primary School stream 1 of 2: Bosire’s score was inflated by 1 in form 37C, that of the Nyaribo by 40 but Nyambati’s were reduced by 41. In stream 2, Bosire was added 18 votes, Nyaribo by 27 while those of Nyambati were reduced by 45. He saw a pattern to systematically deduct the votes from Nyambati and credit them to Bosire and Nyaribo.

55. The witness opined that there must have been a hidden results declaration form or another form 37C because he had a list of all agents and their contacts. He was referred to page 345 of volume 5 petitioner’s bundle. Nyambati is shown to have garnered 131 votes. At page 110 of the IEBC’s response, he has the same score of 131. But in the form at page 142 (Form 37C) he has a 0.

56. I should add that there were discrepancies between the identity card annexed to PW10’s affidavit and his ID number. His appointment letter as chief agent was not properly marked; and, although his oath of secrecy was commissioned, it was not properly marked as an annexture.

57. Antony Miranyi (PW11) was Nyambati’s chief agent for North Mugirango. Upon noticing some discrepancies between form 37A and 37C, he authored a letter to the Constituency Returning Officer dated 10th August 2022 displayed at page 448 of Volume 5. Although the letter was typed, the date was entered by hand, a matter raised in cross examination.

58. His complaints related to Nyairanga Primary School Polling Station 2 of 2 where Nyambati obtained 21 votes. However, in form 37C he was given 2. The witness became suspicious because in stream 1 the candidate also had 21. Like I stated earlier, he also complained about the presence of a ballot box of Omonono polling station which was from a different constituency of Borabu.

59. He also complained that the request for a recount was declined and that the Returning Officer did not respond to any of his complaints. He however conceded that he was not there during the counting of the votes. Upon cross examination he admitted that UDA had an agent at Nyairanga but “he was harassed”. He could not remember the name of the agent but he was “a presidential UDA agent”.

60. Walter Maake (PW12) was a ward manager for UDA in charge of Nyamaiya ward. He visited Bondeka Primary School 1, Gekemoni Primary School 1, Egesieri Primary School 1 of 2, Omokonge Primary School 1 and Masoso Primary School 2. He managed to have some agents admitted although some were allowed access as late as 3:00 p.m. However, his efforts did not bear fruit at Tonga Primary School 2 and Bugo Primary School 1 of 1.

61. Upon cross examination he admitted that at Egesieri 2 of 2, the UDA agent, Geoffrey Okola did not swear an affidavit to that effect. At Egesieri 3 of 3 he averred that no agent signed the declaratory forms. He said that at Gekomoni, their agent was let in at 2:00 p.m. and did not sign form 37A which was the same case at Bondeka, Gekomoni and Egesieri 2 of 3.

62. Like I stated earlier, PW13 was Kinaro Kimaiga Ndubi, the chief agent for UDA and Nyambati. He made a written complaint to the Constituency Retuning Officer for North Mugirango (page 457 of Volume 5). Its tenor is that there was manipulation of results at Nyairanga; and, that a ballot box for Omonono which is in Borabu Constituency was found at Kebabe Girls, North Mugirango. He largely gathered that information from PW11 but he said he saw the ballot box for Omonono. A photograph of the box taken by a witness was not produced at the trial.

63. Although counsel for the respondents challenged the authenticity of the letter, I noted that the Constituency Returning Officer admitted that he received the letter and made hand-written comments on it. The objection was thus unfounded.

64. PW13 claimed he was denied entry into the county tallying center but upon cross examination he admitted that he was denied entry by security officers outside the premises and despite showing them his badge and credentials. He conceded that he did not get in touch with the County Returning Officer (DW1). As I will discuss later, DWI confirmed that PW13 was not present when both 37C forms were signed or when the final announcement was made.

65. The witness said that on 14th August 2022, DW1 sent via WhatsApp a copy of the first impugned form 37C to Nyambati. The screenshot appears at page 458 of Volume 5. On 22 August 2022, DW1 sent a copy of the second form 37C. According to the form, the voter turnout was 179,719 and the total valid votes cast were 176,929. He opined that in the subsequent form 37C, the votes cast were 207,096 which burst the voter turnout of 179, 719 and valid votes cast of 176,929.

66. The screenshot above was strongly objected to by counsel for the 1st to 4th respondents because it was from another cellphone belonging to Nyambati; and, for want of a certificate. But I note that at page 206 of the amended petition there is certificate of electronic record. Furthermore, DW1 confirmed that he sent that information to Nyambati. I find that the objections were groundless.

67. The conclusion by PW13 was that the results in form 37C were not accurate or verifiable because the voter turnout remained in doubt; and, 30,167 votes could not be accounted for. He also testified that 10 days after the declaration of the results, DW1 generated a second form 37C. But according to DW1, the form used to declare the results is only one and appears at pages 141-150 of their volume 4. He said the other one was a draft. I will shortly revisit the matter.

68. That marked the close of the petitioner’s case. As I stated before, the 1st and 2nd respondents did not take to the stand or call any witness. The 3rd and 4th respondents on the other hand called 7 witnesses. The first one was Njoroge Anthony Douglas (DW1), the County Returning Officer who is sued as the 3rd respondent. He relied largely on the Response and his replying affidavit sworn on 8th October 2022.

69. The court allowed him to amend paragraph 32 of the Response to delete the words “That Form 37As tendered in evidence by the 3rd and 4th respondents are forgeries” to read instead by the petitioners. A further attempt to amend paragraphs 41 and 63 of the affidavit was declined for considered reasons.

70. The witness said the presiding officers were issued two booklets of Form 37As but were instructed to only use one. He testified that as a general rule there should be no alterations to form 37A. But if there are errors in filling the data, the presiding officer can amend by crossing, countersigning, stamping or make such corrections. He explained that a “transposition error” occurs when transferring results from forms 37A to 37B or from 37B to 37C. He said results are collated in an Excel sheet with columns and rows which may skew the data to a wrong column.

71. He gave the example in Form 37C at page 151-There was no entry for Osiemo Daniel. The transposition was to the left, it is blank, not even a zero. At page 110 Volume 1, he has ‘00’.

72. He gave another example at page 130 of Volume 4: Nyambati had 0 votes. In Form 37A he was awarded 131. In the second Form 37C he is allocated his rightful 0. According to him, the trans positional error affected all candidates.

73. The witness stated that not all the errors were trans positional. There were typos for instance and omissions that skewed data to one side of the form. He admitted to a number of errors in form 37C as detailed in paragraphs 68, 69, 70, 73, 71 and 72 of the affidavit but the combined errors denied Nyambati 432 votes and added Nyaribo 17 votes. The votes of all the other candidates were unaffected.

74. He also freely admitted that there were two forms 37C. He received 643 forms 37A and four forms 37B from the four constituencies of Kitutu Masaba, West Mugirango, North Mugirango and Borabu. He then tallied them to create form 37C in the presence of some agents. The final form was completed and signed on 12th August 2022, a fact which is contested by the petitioner.

75. He said that when he prepared the first form 37C he signed it together with some agents but discovered it had some errors. When re-examined about the issue the witness said-In the draft 37C, the votes for Bosire and Ondieki were not captured as totals per polling stations for Kitutu Masaba, North Mugirango and part of Borabu. As expounded in paragraph 33, 34, 35 and 36.

76. He then prepared a second form “about an hour later” which he signed together with the same agents. In his opinion, form 37C remained a draft until the formal announcement of results. In his own words: “It does not matter even if the form is signed. Corrections can be done resulting in a new form”.

77. He admitted that he did not announce to the public at the tallying center about the draft. He claimed that he noted the mistake in the first form 37C before he declared the results and informed the agents present. He did not also look out for candidates or their agents.

78. He admitted that the first form 37C contained results as per paragraph 21 of the petition and paragraph 9 of the petitioner’s affidavit save that the votes garnered by Osiemo Daniel were 785 and not 985.

79. He testified that the form contained, “arithmetical errors in terms of summation of the total number of valid votes cast discernable from the total number of votes garnered by each candidate such that from the said entries, the total number of valid votes cast is 207096 (with Osiemo votes 785 and not 985) and not 179719”

80. At paragraph 29 he stated that the amended form 37C did not alter the number of votes garnered by each candidate as reflected in the first form 37C. He testified that the error in the total number of valid votes cast was occasioned by an error in the total number of votes in polling stations in Kitutu Masaba, North Mugirango and Borabu for two candidates, Bosire Anthony Moseti and Bwo’ndieki John Nyariki.

81. The witness made extensive reference to the tables at pages 33, 34, 35 and 36. In the end he denied the allegations that the amended form 37C was a scheme to justify the declaration of results in favour of the 1st respondent.

82. Regarding the WhatsApp message referred to by PW13, he reiterated paragraph 26 of his affidavit and said that he shared the message with Nyambati on 14th August 2022. However, he realized that he had attached the wrong form when someone alerted him in a social forum. He then sent another message to Nyambati on 22nd August 2022 attaching the “correct form 37C”.

83. The witness was referred to page 142 of volume 4 IEBC bundle containing the first Form 37C. In that document the total voters were 179,719 (55. 59%). The second-last column lists the valid votes as 176,929. The last column (rejected ballots) are 2,790. He however insisted that it was not the form he used to declare results; and, that the correct one appears at pages 142-150 of Volume 4 of the same bundle.

84. When pressed further, he answered-The declaration form should be signed by the Returning Officer, and any present candidate or agent. It should have valid votes and votes cast which means the same thing. It must show rejected ballots. Rejected ballots are not votes. We tally votes cast in favour of candidates. Valid votes and people who cast votes are the same.

85. When re-examined, he added that “We do not have rejected votes. It is rejected ballots; the latter being votes not in favour of any candidate.” Applying that logic to the impugned form 37C he stated that-The total voters are 207,096. Rejected are 2,791. Voters are people who cast votes for a candidate. Those are the ‘voters who turned out to vote’. It would include rejected ballots. In the first form (page 156) the total number of rejected ballots plus votes cast agree with ‘total number of persons who turned out to vote’. I did not use the same formula in form at page 150. In that form if you add 207,096 + 2791 = 209,887. It would be incorrect if you use formula at page 156. If you add all votes garnered by candidates, you get 207,096. Every voter who came to vote and their vote was counted. So everyone cast a vote. Valid votes are for those who identified candidates. Rejected ballots are not valid votes.In the 2nd Form 37C on 1st Form 37C contain [sic] details of those who have made a decision at the election. The rejected ballots are persons who made no clear decision. They cannot be said to be among number of voters who turned out to vote.

86. The witness conceded that the Response did not make clarifications about the alterations in forms 37A in Volume 2 page 91 of the IEBC bundle. He said that 106 votes for Nyambati were omitted from form 37C at page 149. He claimed it was “an error of omission” explained at paragraph 73 of his affidavit.

87. When cross examined further he conceded that IEBC is required to provide accurate and verifiable figures. He clarified that the tables in his affidavit were meant to “address constituency errors. They are not the total of the corrected summation. We attached the final figures in final Form 37C. What I have shown in the affidavit are samples. In my view the response by IEBC is sufficient to clarify the errors”.

88. At paragraph 19, he deposed that the total number of registered voters in the county is 323,291. He denied the allegation by the petitioner that the total number of voters who turned out to vote was 179,719 representing (55. 59%) of the total registered voters. He testified that the total number of voters who turned out to vote was 209,887 or 64. 92% voter turnout.

89. The witness was put to task about the stamp impressions on Form 37A for Motagara DEB Primary School 2 of 2 which showed an outer square and an inner circle. In the form, two agents were only identified by one name. The form was neither signed nor dated. The witness was referred to another bundle of IEBC forms which had stamps with a circular border. He claimed that all the stamps were from IEBC and that the round rubber stamp is mounted on a square wooden block. He opined that the disputed mark may arise when the inkpad dries up. He showed examples of forms at page 122 Volume 1 and page 109 Volume 2 where the stamps have square edges.

90. The witness denied that the alterations were meant to favor Nyaribo. He argued that since the margin of victory was over 30,000 votes, the errors did not substantially affect the results. He maintained that the election was won by the 1st respondent with 81,980 votes and that the certificate issued to him was in accordance with Regulation 83(1)(g).

91. He claimed that the forms attached to the petitioners’ affidavits are either inaccurate or forgeries. He expounded it at the tables in paragraphs 45, 47, 49, 50 and 53. He also responded to allegations in each of the witness affidavits annexed to the petition from paragraphs 68 to 81.

92. The witness was then referred to video clips in the petitioner’s exhibit DOA1. He said the second clip was shot at the tallying center. In his view, there was nothing wrong with the agent displaying party symbols there as the restrictions only applied to the polling centers. He had no control over choice of agents which was the prerogative of the candidates or their parties. He opined that even when a list of agents is given to a presiding officer, it is purely for reference purposes.

93. The witness was shown a publication of the County Assembly for members of County Assembly Service Board for Nyamira which included Leonard Okari Mogaru. Since this document was introduced by the petitioner after close of his case, it was marked for identification as PMFI1.

94. DW1’s final conclusion was that the gubernatorial elections were conducted in a transparent, accountable and verifiable manner and in accordance with the Constitution, the Elections Act and the regulations.

95. DW2 was Sumbwa Ambuko. Although he was the Constituency Returning Officer for Kitutu Masaba, paragraphs 7 and 20 of his deposition stated that he supervised elections for Borabu constituency which was not the case. Those offending paragraphs in the circumstances are for striking out.

96. He received 214 forms 37A from various Presiding Officers from which he prepared the form 37B. He announced the results at the Constituency Tallying Center in the presence of some candidates and agents on 11th August 2022. He then forwarded the form 37B through his deputy, Bob Asstless Mauko, to DW1 at the County Tallying Center. I should add that Mauko swore an affidavit but was not called as a witness.

97. DW2 queried the authenticity of forms 37A attached to the petitioner’s affidavit. When cross examined by Mr. Kamwaro, he said that he never received any complaints about forms 37B and there were no discrepancies between that form and form 37C.

98. He said that he advised presiding officers to only use one booklet of form 37A and deposit the other in a ballot box. He also said the two booklets had different identification numbers. He conceded that the mere fact that the petitioner’s forms were at times faint or illegible did not mean they were forgeries.

99. He opined that a person who casts a vote, whichever way he marks the ballot, is a voter. However, a vote is different from a ballot and in his view a rejected ballot is not a vote. When questioned by Mr. Muyundo, he confirmed that the amended petition has not complained about the use of the 2 booklets. In the end he was of the view that the elections in the Kitutu Masaba Constituency were conducted in a free, fair and transparent manner.

100. Julius Okeyo (DW3) was the Returning Officer for Borabu Constituency. He received 128 forms 37A and tallied the results into form 37B. He signed it on 12th August 2022, announced the results and forwarded it to DW1. He said that the results in form 37B corresponded with the tally in form 37C. He disputed the averments by two of the petitioner’s witnesses, Yabesh Ntabo Mokua and Doris Magoma Nyamanga. The two did not take to the stand.

101. Upon cross examination by Mr. Mokua the witness conceded that he could not comment about West Mugirago because he did not perform any duties there. Strangely this witness said that he supplied the presiding officers with three booklets of forms 37A and each had 16 pages but he said the officers were told to only use one and the others sealed in the ballot box. He testified that he never received any report of a misplaced ballot box of Omonono; and, that all the ballot boxes and forms 37A were accounted for.

102. DW4 was Susan Kemunto Ngare, the Presiding Officer at Kiabonyuru Primary School Polling Station 1 of 2 in Borabu. She said that voters were identified through the KIEMS kit upon producing their IDs which were captured in the logs. At the end of the exercise, she counted and verified the votes in the presence of the agents. The results were enumerated at paragraph 7 of her affidavit. She prepared form 37A annexed to her affidavit and appearing at page 122 of IEBC bundle. She announced the results in the presence of five agents who signed the form. She also allowed them to take images of the forms.

103. She denied the allegations by the petitioner that there was a variance between the votes cast and those entered in 37C. She conceded however that there was an alteration in the total number of registered voters and the total number of valid votes cast but which did not affect the total votes of any of the candidates.

104. Sylvia Jepchumba (DW5) was the Constituency Returning Officer for West Mugirango Constituency. She received all the 166 forms 37A and prepared form 37B. The results are tabulated in paragraph 13 of her deposition and were announced in the presence of some agents and candidates. She then handed all forms 37A and 37B to DW1. In her opinion the results in 37A and 37B were correctly tallied in 37C produced by DW1.

105. She disputed the averments in the affidavit of Doris Magoma, a witness for the petitioner but who did not testify. At paragraphs 21, 22, 23, 24, 25 and 26 of her affidavit she has analyzed the results presented by the said Magoma vis a vis the certified forms of the IEBC to demonstrate that the forms relied on by Magoma were not authentic.

106. In the final analysis her view was that the elections in West Mugirango were conducted in accordance with the Constitution and that the results were true representation of the people of West Mugirango.

107. In cross examination the witness said that the results of Nyamira DEB streams 1 to 8 “had no issues”. She said that the forms exhibited by the petitioner and those of the IEBC are different for steam 1. The witness said she received no complaints from Motagara Polling Station or for Makairo DEB Polling Station.

108. The witness was taken to task on the forms 37A for Makairo Polling Station 1 and 2. She admitted the form for stream 2 is printed as stream 1/2 but the digit 2/2 inserted by hand. However, she confirmed that the results of the two forms related to corresponding streams and were signed by different agents and presiding officers for each stream. Two of the agents were from UDA Party. The two forms also had different serial numbers GB085686 and GB085870.

109. She said there were two booklets of forms 37A but the returning officers were to use only the first booklet. Each booklet had six copies. The original was supposed to be submitted to the Constituency Returning Officer, the first carbon copy pasted on the ballot box and another pasted to the door of the center. In the event that the remaining copies were not sufficient for all agents, they could take an image of the form.

110. DW6 Ruto Wesley was the Constituency Returning Officer for North Mugirango Constituency. He received all the 135 original forms 37A and the ballot boxes. He then prepared form 37B and announced the results in the presence of agents and some candidates on 11th August 2022. He then handed all forms 37A and 37B to DW1. In his opinion the results in 37A and 37B were correctly tallied in 37C produced by DW1.

111. He disputed the averments in the affidavit of Andrew Ombongo (PW6) that forms 37A and 37C for North Mugurango differ. In cross examination the witness said that he never got a complaint regarding Form 37B but he could not tell whether the results in form 37C are grouped by constituencies. He confirmed that the letter at page 457 Volume 5 by Nyambati was delivered to him by Kinaro (PW13). He acknowledged receipt and made remarks at the bottom of the letter. In re-examination he stated that he got no further comments from Nyambati. His view was that the allegations in the letter did not amount to malpractices.

112. He said that it would be an election offence if one “knowingly makes a wrong entry”. He said the code indicated in the above letter could not be for a ballot box. The serial number for ballot boxes are engraved on the side by a machine and not labelled by names of polling stations. But there would be a record showing that a ballot box Serial Number “X” had been allocated to polling station “Y”.

113. DW7 was Martha Ongeni, the presiding officer at Mwongeni Youth Polytechnic Polling Station 2 of 4. She testified that she counted and verified the votes in the presence of the agents who also signed the form. The votes are tabulated in paragraph 7 of her affidavit. She prepared form 37A annexed at page 2 of volume 4 of IEBC bundle. She also said that the form corresponds with the copy annexed by the petitioners appearing at page 126 of volume of petitioner’s bundle. She also allowed the agents to take photographs of the form.

114. She denied the allegations by the petitioner that there were variances between the number of votes cast or as entered in form 37C. She said that form 37A did not have any alterations in the results of candidates and that the only digit that was corrected was the last digit in the figure of 448 for the total number of voters.

115. That marked the close of the 3rd and 4th respondents case.

Report on Scrutiny 116. Following a successful motion by the petitioner, the court ordered a partial scrutiny of election materials in a number of disputed polling stations. IEBC was to provide an inventory of the ballot boxes used for the gubernatorial election at Omonono Primary School polling station 2 of 2 and Nyanchoka Tea Buying Centre 1 of 1 including their serial numbers and the seals serial numbers.

117. A partial scrutiny was also ordered but limited to a recount of the votes cast for the governor at Motagara DEB Primary School 2 of 2, Kenyamware DEB Primary School 1 of 1, Nyankoba Polling Station 1 and Nyabigege DOK Primary School Polling Station 1 of 2. In addition, there was to be a re-tally of the results in the final Forms 37B for Kitutu Masaba, Borabu, West Mugirango and North Mugirango; and, the final Form 37C for Nyamira County from the original Forms 37A submitted to the County Returning Officer.

118. On 19th December 2022, the Deputy Registrar formally submitted his scrutiny report to the court and to all the parties. The original report is dated 15th December 2022. Under the Petition Rules it now forms part of the proceedings in this petition and parties were at liberty to address it in their final submissions. I will analyze some of its key findings in the course of this judgment.

Final Submissions 119. The petitioner filed final submissions dated 1st January 2023 together with a volume of authorities and a comprehensive case digest. The 1st and 2nd respondents’ submissions are dated 16th January 2023 with a list of precedents. The 3rd and 4th respondents filed their submissions on even date accompanied by a bundle of authorities.

120. On 18th January 2023, all learned counsel for the disputants addressed the court on those submissions. I am greatly indebted to all of them for their detailed submissions and courtesy to the Court. I will make references to the submissions in the course of the judgment. If I do not make direct reference to all the cited cases and materials, it is not for their lack of relevance.

Issues for Determination 121. The following five broad issues were agreed for determination:i.Whether the gubernatorial election for Nyamira held on 9th August 2022 was conducted in compliance with the Constitution, the Elections Act and Regulations.ii.Whether there were irregularities, illegalities or malpractices in the election.iii.Whether the 1st respondent was validly elected as the Governor for Nyamira County.iv.Whether the petitioner is entitled to the reliefs sought in the amended petition; and,v.Who will bear the costs of the petition?

122. A close scrutiny of those issues reveals that they are largely subsumed under the third one: This court is simply being asked to determine whether the 1st respondent (and by extension the 2nd respondent) were validly returned as the Governor and Deputy Governor of Nyamira County.

Legal Framework and the Burden and Standard of Proof 123. I will commence with a synopsis of the applicable legal regime. Article 1 of the Constitution proclaims the sovereignty of the people. The political rights enshrined under Article 38 are partly exercised through regular elections. Articles 81 and 82 on the other hand call for free, fair, transparent and credible polls.

124. Article 81(e) of the Constitution provides as follows:The electoral system shall comply with the following principles-(e)free and fair elections, which are-(i)by secret ballot;(ii)free from violence, intimidation, improper influence or corruption;(iii)conducted by an independent body;(iv)transparent;(v)administered in an impartial, neutral, efficient, accurate and accountable manner.

125. The Elections Act and the Regulations thereunder are the legislation contemplated by Article 82. The amended section 83 (1) of the Act; and, which applies to the present dispute, provides-(1)A Court shall not declare an election void for non-compliance with any written law relating to that election if it appears that-a.the election was conducted in accordance with the principles laid down in the Constitution and in that written law; andb.the non-compliance did not substantially affect the result of the election.

126. The above provision is framed in negative language. It means that there is a rebuttable presumption in favour of the respondents that the election was conducted properly and in accordance with the law. It is also implied that elections are not always perfect. See Steven Kariuki v George Mike Wanjohi and othersNairobi, High Court Petition No 2 of 2013 [2013] eKLR. Consequently, not all malpractices will lead to nullification of the result. See the dissenting opinion of Njoki SCJ in Raila Odinga & another v IEBC & 2 others, Supreme Court of Kenya, Presidential Petition 1 of 2017 [2017] eKLR, Morgan and others v Simpson and another [1974] 3 ALL ER 722.

127. In Kakuta Maimai Hamisi v Peris Pesi Tobiko and others, Nairobi High Court Petition 5 of 2013 [2013] eKLR, I had the following to say on the subject-The crux of the matter is whether all the grounds raised by the petitioner sufficiently challenge the integrity of the poll. The court has to weigh the impact of the alleged irregularities on the outcome of the election. See James Omingo Magara v Manson Nyamweya and others Kisumu, Court of Appeal, Civil Appeal 8 of 2010 [2010] eKLR. In the context of a general election for six electoral offices, it would be impractical to demand or expect mathematical precision

128. The centrality of evidence in a dispute of this nature was emphasized by the Supreme Court in Odinga & 16 others v Ruto & 10 othersPresidential Election Petition E005, E001, E002, E003, E004, E007 & E008 of 2022 (Consolidated) [2022] KESC 56 (KLR) -(28) The law of evidence complements the existing civil and criminal substantive and procedural laws in this country. The outcome of a case depends on the strength, accuracy and reliability of evidence. In an adversarial court system like ours, the courts and Judges are ‘blind’, in the sense that they do not carry out any investigative roles or gather evidence on behalf of the parties before them. They depend on and determine disputes from what parties present. Consequently, cases are won or lost on the evidence placed before the Court.

129. Closely linked to this is the legal and evidential burden of proof. The formal responses to the petition by all the respondents join issues on all the allegations in the amended petition. The respondents have thus put the petitioner to strict proof. The legal burden of proof remains throughout on the shoulders of the petitioner: the petitioner must lead cogent evidence to impeach the poll. The cardinal precept of the law of evidence is that he who alleges must prove; and to the required standard of proof. See section 107 and 109 of the Evidence Act.

130. If the petitioner marshals sufficient or quality evidence, he will have discharged that burden at which point then the evidentiary burden shifts to the respondents and calls for a rebuttal. Ultimately, the petitioner must show that the preponderance of evidence inures in his favour. If he fails to rise to that standard, the petition must be dismissed. See Henry Tiole ole Ndema v IEBC & 2 others, Kitale High Court Election Petiton 2 of 2017 [2018] eKLR.

131. The line between the legal and evidentiary burden was restated by the Supreme court in Odinga & 16 others v Ruto & 10 others [supra]-(32) Suffice to stress that the Court has been consistent that a petitioner who seeks the nullification of elections for alleged non-conformity with the Constitution or the law or on the basis of irregularities and illegalities, has the duty to proffer cogent and credible evidence to prove those grounds to the satisfaction of the Court. Once the Court is convinced that the petitioner has discharged that burden, then the evidentiary burden shifts to the respondent (who in most election-related cases is IEBC), to present evidence by way of rebuttal of the assertion.

132. In the earlier decision in 2013, the Supreme Court in Raila Odinga and othersvs Independent Electoral and Boundaries Commission et al, Petitions Nos. 3 & 4 of 2013 (consolidated); [2013] eKLR, the court had held-Where a party alleges non-conformity with the electoral law, the petitioner must not only prove that there has been non-compliance with the law, but that such failure of compliance did affect the validity of the elections. It is on that basis that the respondents bear the burden of proving the contrary. This emerges from a long standing common law approach in respect of alleged irregularity in the acts of public bodies. Omnia praesumuntur rite et solemniter essa acta: all acts are presumed to be done rightly and regularly. So the petitioner must set out by raising firm and credible evidence of the public authority’s departures from the prescriptions of the law.

133. The standard of proof in election petitions is higher than a balance of probabilities in ordinary civil cases but not beyond reasonable doubt as required in criminal cases. See Raila Odinga and others v Independent Electoral and Boundaries Commission and 3 others, [2013, supra]; Henry Tiole ole Ndema v IEBC & 2 others [supra]; Rishad Amana v IEBC & 2 others, Malindi, High Court election petition 6 of 2013 [2013] eKLR.

134. However, with regard to allegations by the petitioner of a criminal nature such as unlawful use of public resources or corrupt practices, the standard of proof is beyond reasonable doubt.

135. Lastly, it bears repeating that the parties to the petition are bound by their pleadings. See generally Jackton Ranguma v IEBC & 2 others, Kisumu High Court E.P. 3 of 2017 [2018] eKLR, Henry Tiole ole Ndema v IEBC & 2 others [supra].

Analysis and Final Determination 136. I will now return to the matter at hand. Like I stated earlier, the court is simply being asked to determine whether the 1st respondent (and by extension the 2nd respondent) were validly elected.

137. A good starting point is the failure by the 1st and 2nd respondents to testify or call any witness. According to their learned counsel, they formed the impression that no case had been made out against them by the petitioner or his 13 witnesses to warrant any rebuttal.

138. That was a perilous gamble. Learned counsel for the petitioner equated it with the empty chair doctrine expounded by Robert H Steir Jr; Revisiting the Missing Witness Inference-Quieting the Loud Voice from the Empty Chair. See also Joel Makori Onsando & another v IEBC & 4 others [2018] eKLR.

139. I find that the posturing by the 1st and 2nd respondents was a grave and unnecessary risk. The consequences are fourfold: Firstly, although the two affidavits filed with their Response remained on the record they were never adopted by the court or admitted by consent of the parties as required by the Rules.

140. Secondly, I am entitled, and I do infer, that the testimony would have been unfavourable to the two respondents. I do so on the authority of the Supreme Court in Mawathe Julius Musili v IEBC & another, Petition 16 of 2018 [2018] eKLR-(86) On this principle, when a party – in this case the IEBC – fails to produce the witness (Returning Officer), the Court may infer that the testimony would have been unfavourable to the IEBC. [underlining added]

141. Thirdly and more importantly, since the evidence was not tested in cross-examination, it means that the two depositions had little if any probative value.

142. Fourthly, the net effect was that the 1st and 2nd respondents offered no meaningful rebuttal to the allegations made by the petitioner and his 13 witnesses.

143. I am well guided in those conclusions by both the Court of Appeal and the Supreme Court. In the Court of Appeal decision in Sumra Irshadali Mohammed v IEBC & another Election Petition Appeal 22 of 2018 (unreported), the learned judges held-It is clear to us and as rightly held by the trial court, that the R.O did not appear to give evidence, despite the fact he had filed a lengthy replying affidavit. It means therefore there was no response to the Petition, consequently it cannot be said that the appellant did not discharge the burden of proof. Having adduced oral and 36 documentary evidence before it, we do not know what else the appellant was required to do. The trial court did put a high and unreasonable premium on the absence and failure of the R.O to adduce evidence. The burden shifted the moment the issues raised by the appellant were not answered. [underlining added].

144. When the above matter reached the Supreme Court in Mawathe Julius Musili v IEBC & another [supra] the learned judges addressed the question of failure by a respondent (in that case the IEBC) to call evidence in an election petition. They held-(86)On this principle, when a party – in this case the IEBC – fails to produce the witness (Returning Officer), the Court may infer that the testimony would have been unfavourable to the IEBC. The absence of the Returning Officer inevitably weakens the IEBC’s response and evidence.(87)However, a Court may look the other way if the reason given for the absence of the Returning Officer is satisfactory and credible. Such, however, was not the case in the instant matter. Both Superior Courts found the replying affidavit inadmissible. In our view, this was an entirely proper finding on the issue of the probative value of the Returning Officer’s affidavit. Now such a finding begs one question: what was the effect of this finding, as a factor in the disposal of the case? [Underlining added]

145. The learned judges of the Supreme Court answered their question as follows-(91) The evidential burden therefore shifted to the 1st respondent, to show that this was not so. The evidence relied on by the 1st respondent was its replying affidavit sworn by its employee: the Returning Officer. This evidence was held to be of no probative value by the Superior Courts. In the circumstances, we are in agreement with the Appellate Court: it cannot possibly be concluded that the 1st respondent discharged its evidential burden. [Underlining added]

146. It is true that the 1st and 2nd respondents extensively and at times aggressively cross-examined the petitioner and his witnesses. But a party’s case cannot be built entirely on cross examination. In Christopher Odhiambo Karan v David Ouma Ochieng & another, Supreme Court Petition 36 of 2018 [2018] eKLR, the learned judges were emphatic that whereas cross examination is an important tool of testing evidence, the answers in cross-examination cannot form a basis of a party’s case or build a defence. They held-(85)The purpose of cross examination is therefore principally to unearth the truth. Cross-examination is regarded as one of the most important evidentiary safeguards of the adversarial system and an entrenched right of the defendant. In Juan Torres & another v Michael Njai, Civil Appeal No. 154of 20115; [2017] eKLR, the Court of Appeal thus emphasized that cross examination is intended to help the court to reach a just decision.(86)In the Court of Appeal decision of John Wainaina Kagwe Vs Hussein Dairy Limited Mombasa Court of Appeal 215/2010 (Githinji, Makhandia and Murgor JJA), the Court of Appeal was categorical that “answers in cross-examination cannot form a basis of a party’s case or built [sic] a defence. They must tender evidence in support of the allegation.”[Underlining added]

147. For all those reasons, I readily find that the 1st and 2nd respondents failed to offer any or sufficient rebuttal to the evidence by the petitioner. Had the matter ended there, I would have found that the petitioner had substantially discharged his evidential burden. But there were two other respondents: The County Returning Officer and the IEBC (the 3rd and 4th respondents respectively) and who called witnesses.

148. Furthermore, I must still interrogate whether the petitioner proved to the required standard the complaint of a criminal nature or furthering a corrupt practice made against the 1st and 2nd respondents and an agent known as Leonard Mogaru; and, whether all the five issues framed earlier have been answered in favour of the petitioner.

149. I will commence with the disputed results for Omonono Polling Station.PW13 made a written complaint to the Constituency Returning Officer for North Mugirango that a ballot box for the station (which is situated in Borabu Constituency) was found at Kebabe Girls, in North Mugirango. Like I stated earlier, the witness largely gathered that information from PW11. But he also said he saw the ballot box. An image of the box taken by a witness was not produced at the trial.

150. The retort from Julius Okeyo (DW3) was that he never received any report of the misplaced ballot box; and, that all the ballot boxes and forms 37A for the entire constituency were accounted for.

151. The court ordered a partial scrutiny at Omonono. IEBC was required to provide an inventory of the ballot boxes used for the governor’s election at Omonono Primary School polling station 2 of 2 including their serial numbers and the seals serial numbers. The IEBC presented the Polling Station Diary (PSD). The Report by the Deputy Registrar concludes that the diary and its entries was not specific as to whether it related to stream 1 or 2.

152. Learned counsel for the petitioner submitted that by failing to present the ballot boxes “when ordered by the court…. was an effort of concealing critical evidence”. That is not entirely true because the order on scrutiny did not require the 3rd respondent and IEBC to present the ballot boxes for inspection.

153. The PSD contained the inventory. However, the diary did not specify for which stream. True, the details of the serial number of the box, the seals serial numbers before and after the poll are there, but in the absence of the box it became difficult to make a comparison.

154. Despite that shortcoming, it would not be right to condemn the 3rd and 4th respondents in view of the express terms of the order on scrutiny. However, the net effect is that it remains doubtful whether the inventory provided was for stream 1 or 2. What this means is that the allegations by the petitioner that a ballot box for Omonono polling station 2 of 2 which is situated in Borabu Constituency was found at Kebabe Girls, in North Mugirango were not meaningfully challenged. The allegation was not disproved by the casual rejoinder from Julius Okeyo (DW3). Paraphrased, the petitioner proved this allegation slightly above a balance of probabilities.

155. The petitioner deposed in paragraph 135 that a ballot box of Nyanchoka TBC “was found open at the tallying station without any reasonable explanation”. Like I stated, there is a photograph and a certificate of electronic record of Yabesh Ntabo in Volume 2. However, the witness did not take to the stand and the photograph was never admitted in evidence.

156. I again ordered for partial scrutiny in the two streams at Nyanchoka TBC on terms similar to those at Omonono in the preceding paragraphs. The IEBC equally failed to present the ballot boxes for comparison but tendered the PSD. The remarks by the Deputy Registrar were that “the entries for the serial numbers used after polling have all been altered (crossed) by hand”.

157. But I find that the allegation was not proved for four main reasons: Firstly, the only evidence was from the petitioner. Secondly, he learnt of it from a potential witness, Yabesh Ntabo, who did not testify and whose deposition was never adopted or admitted into evidence. Thirdly, the petitioner freely conceded that the photograph was taken at night; and, he could not tell the time or place. Fourthly, the photograph was never admitted in evidence. I also remain alive that boxes are not to be opened after the poll but are supposed to remain sealed unless there be an order of the court. The upshot was that it was hearsay and not proved.

158. The next issue is whether the petitioner proved that the 1st and 2nd respondents engaged in a corrupt practice by hiring Leonard Okari Mogaru as their agent; or, whether Mogaru was an employee of the County Public Service Board as pleaded at paragraphs 161,162 and 163 of the amended petition.

159. Citing among other cases, Karanja Kabage v Joseph Kiuna Kariambegu Ng’ang’a & 2 others, Nairobi, Court of Appeal Civil Appeal 301 of 2013 [2014] eKLR; Frederick Otieno Outa v Jared Odoyo Okello & 4 others, Supreme Court Petition 6 of 2014 [2014] eKLR; and, the decisions of the Supreme Court of India in Indira Nehru Gandhi vs Shri Raj Narain & Anr [1975] AIR 2299, Dr. Y.S. Parmar v Shhira Singh Paul & another [1959] AIR 244, I was implored to find that that the appointment of Mogaru as an agent was proof of the corrupt practice; and, that it would be sufficient to nullify the election.

160. My further findings are as follows: Firstly, the Compact Disk marked DOA1 containing four short MP4 video clips was played in court. The respondents did not adequately challenge its admissibility or the certificate of electronic recording then. They have chosen to attack it belatedly in their final submissions. The video clearly shows Mogaru at the county tallying centre wearing a red and white shirt emblazoned at the back: Mogaru Supports Governor Nyaribo.

161. Secondly, he signed the two contentious forms 37C at the tallying centre as an agent of UPA Party. I was unconvinced by the submission by the 1st and 2nd respondents that “he was not the same and only person” seen at the tallying centre in the red and white shirt above.

162. Thirdly, DW1 confirmed that the two forms are signed by an agent from UPA Party bearing that name. The details of his identification card number appear on the form. Fourthly, and more importantly, the 1st and 2nd respondents or any of their witnesses did not testify in this matter. The latter did not thus rebut that evidence at all. See Mawathe Julius Musili v IEBC & another, Petition 16 of 2018 [2018] eKLR.

163. From that analysis, I readily find that the said Mogaru was an agent of the UPA party and showed his preferences openly at the county tallying centre.

164. I will now turn to the allegation that Mogaru or the 1st respondent contravened sections 14 or 15 of Elections Offences Act. I will commence with section 15 which provides as follows-(1)A public officer who—(a)engages in the activities of any political party or candidate or act as an agent of a political party or a candidate in an election;(b)publicly indicates support for or opposition against any party, side or candidate participating in an election;(c)engages in political campaigns or other political activity; or(d)uses public resources to initiate new development projects in any constituency or county for the purpose of supporting a candidate or political party in that constituency or county, commits an offence and is liable on conviction, to a fine not exceeding one million shillings or to imprisonment for a term not exceeding three years, or to both.(2)A person who knowingly aids in contravention of subsection (1) commits an offence and is liable, on conviction to a fine not exceeding one million shillings or to imprisonment for a term not exceeding three years, or to both such fine and imprisonment.(3)A candidate who knowingly aids in contravention of subsection (1) shall not be eligible to contest in the election.

165. Did the petitioner adduce sufficient evidence to prove criminal conduct or a corrupt practice to the required standard? The answer is in the negative for five main reasons. Firstly, and as I stated earlier, parties are bound by their pleadings. The petitioner pleaded that Mogaru was an employee of the County Public Service Board. A belated attempt to further amend the amended petition was declined for reasons on the record.

166. According to the publication of the County Assembly marked as PMFI1. Mogaru was a member of County Assembly Service Board for Nyamira, a different entity altogether. Since this document was introduced by the petitioner after close of his case, and during cross-examination, it could only be marked for identification. The question arising is whether answers given in cross-examination by the respondents’ witness (DW1) about that document established the petitioner’s claim.

167. Learned counsel for the petitioner, cited Doabia & Doabia; Law of Elections and Election Petitions for the proposition that I should not stretch the doctrine of strict proof too far “to the extent of making it nigh impossible to prove an allegation of corrupt practice”.

168. He also made extensive reference toEco Bank (NIG) Ltd vs Uchechukwu Aghazu (2019) LCN/12760 (CA). In the relevant paragraph, the court stated-Through the questions put to her by counsel in cross-examination she now gave detailed explanation as to why the cheque was written in the name of the bank staff instead of in the name of the bank. In a similar vein through questions put by learned counsel for the appellant in cross-examination PW1 now explained why the names of the signatories were not indicated in the Certificate of Deposit. Learned counsel cannot now turn round and complain that the answers to the questions put by him were not pleaded. I agree with learned counsel for the Respondent that all the facts tabulated by the Appellant as wrongly admitted are relevant and admissible as some of them explain or introduce the facts in issue in the pleadings while some others support an inference suggested by the facts in issue.

169. But I am well guided on that point and bound by the Court of Appeal in Kenneth Nyaga Mwige v Austin Kiguta & 2 others Court of Appeal, Nairobi, Civil Appeal 140 of 2008 [2015] eKLR where the learned judges held-[20] Once a document has been marked for identification, it must be proved. A witness must produce the document and tender it in evidence as an exhibit and lay foundation for its authenticity and relevance to the facts of the case. Once this foundation is laid, the witness must move the court to have the document produced as an exhibit and be part of the court record. If the document is not marked as an exhibit, it is not part of the record. If admitted into evidence and not formally produced and proved, the document would only be hearsay, untested and an unauthenticated account.

170. Furthermore, the petitioner cannot meander outside his pleadings in which he stated that Mogaru was an employee of the County Public Service Board. I find that it was thus not established that Mogaru was an employee of the County Public Service Board as pleaded.

171. Secondly, the petitioner did not establish that Mogaru was appointed as an agent of the 1st and 2nd respondents. From the declaratory forms, it is apparent he was an agent for the UPA Party. Learned counsel for the petitioner, Mr. Omwanza, was of the view that it makes no difference. I do not entirely agree. The difference between a political party agent and a candidate’s agent and the implications was well explained in Alfred Nganga Mutua & 2 others v Wavinya Ndeti & another, Supreme Court Petition Nos. 11 & 14 of 2018 (Consolidated) [2018] eKLR.

172. Thirdly, Regulation 62 (6) expressly denies admission into a polling station to someone displaying party symbols, badges, paraphernalia or dress code signifying support to a party or candidate. I have not come across an express rule barring an agent from doing so after the close of polls at the county tallying centre. The logic is obvious: at that stage, voting and counting is long gone; and, there are no longer voters capable of being influenced.

173. Fourthly, DW1 and the IEBC had no control over the nomination of agents by candidates or parties. Their role was to ensure that the agents had an appointment letter, identification documents and had executed the oath of secrecy. The petitioner did not lead any evidence to show that the 3rd and 4th respondents were complicit in any scheme by Mogaru or the 1st and 2nd respondents to further a corrupt practice.

174. Fifthly, and more importantly, the petitioner pleaded that the appointment of Mogaru amounted to criminal conduct. The petitioner’s case is that Mogaru’s activities, and by extension that of the 1st and 2nd respondents, amounted to a “corrupt practice” in violation of section 15 of the Act. Learned counsel for the petitioner submitted that appointing Mogaru was a corrupt practice that “vitiated the 1st and 2nd respondents’ ability to vie for elections under section 15 (3) of the Elections Offences Act”.

175. As I stated earlier, in allegations of criminal conduct such as those prohibited by sections 14 and 15 of the above Act, the standard of proof is beyond any reasonable doubt. I do not in any manner condone the behavior or antics displayed by Mogaru. I find it to be despicable.

176. The petitioner had also pleaded at paragraph 162 of the amended petition that the 1st respondent contravened section 14 of the Election Offences Act. That section deals with use of public resources. Besides references to the employment of Mogaru, no other material evidence was led on that allegation.

177. From the evidence I have highlighted above and the legal analysis, I readily find that the petitioner failed to prove beyond reasonable doubt that the 1st and 2nd respondents used public resources or engaged in a corrupt practice in contravention of sections 14 and 15 of Election Offences Act. This finding equally applies to the allegation by PW10 that a polling clerk at Manga TBC Polling Station was “telling people to vote for Nyaribo”. It was a bear statement unadorned by concrete evidence.

178. One of the key planks in this petition is that the 3rd and 4th respondents altered the results in forms 37A. It is thus important to interrogate the nature and extent of the errors and whether they substantially affected the outcome of the poll.

179. But before I delve any deeper into those entries, I will comment on two preliminary matters. The first is about the existence of two booklets of pre-printed forms 37A. This matter arose largely in the cross-examination of the witnesses presented by the IEBC. The witnesses clarified that the two booklets had different serial numbers; and, that presiding officers had firm instructions to only use the first booklet and seal the other one inside the ballot box.

180. However, the issue of the two booklets was not pleaded in the amended petition at all. There was also a paucity of evidence on whether it impacted on the election. Noting that the petitioner is bound by his pleadings, I need not make a finding on it.

181. The second preliminary matter revolved around competing claims about the authenticity of the forms 37A annexed by the petitioner against those submitted by the IEBC. Learned counsel for the petitioner submitted that “the forms 37As as provided by IEBC are different from the ones in possession of the petitioner and which materially affected the overall outcome of the results”. The respondents on the other hand condemned the forms exhibited by the petitioner as outright forgeries.

182. Where does the truth lie? At the pre-trial conference, I ordered the IEBC to furnish the court with certified copies of all the forms used to declare the results. They were lodged in court by the IEBC on 7th November 2022 and used at the trial. As I will detail later, some were markedly different from those by the petitioner.

183. In the absence of clear evidence that the certified copies were fraudulent, the court would be minded to rely on them as the official results of the election. Equally, the IEBC was required to prove that the petitioner’s forms were a fraud. The standard of proof for fraud is higher than a balance of probability; it approaches, but is below, proof beyond reasonable doubt. See Ratilal Patel v Lalji Makanji[1957] E.A 314 at 317, Umila Mahindra Shah v Barclays Bank International Ltd[1979] KLR 67.

184. Some of the certified forms presented by the IEBC were poor carbon copies. A good example is at Kenyamware DEB Primary School 1 of 1: the certified copy of Form 37A displayed by the IEBC at page 159 of the response is too faint and illegible. The copy attached by the petitioner to the amended petition is equally indecipherable. One cannot clearly see the candidates’ results or the signatures of the agents or presiding officers. The same is true of the form annexed by his witness, Wesley Kerina (PW9).

185. Furthermore, at Motagara DEB Polling Station 2 of 2 there were two conflicting forms 37A: In one Nyambati had 88 votes but only 1 vote in the other. From the scrutiny report and recount, the petitioner was correct that Nyaribo obtained 88 votes and not 1 vote as indicated in IEBC’s form 37A. But at the same station, the 1st respondent garnered 122 votes and not 88 as indicated in the form 37A. From those few examples, the disputants were not entirely forthright about the true entries. It is a classic case of the kettle calling the pot black.

186. It is thus important to interrogate the source of the forms 37A displayed by the petitioner. The petitioner said he received some forms in hard copies and others electronically. One of his witnesses Elvis Kibeni (PW10) said he got some in hard copies while others were received vide WhatsApp messages.

187. The respondent’s witness, Martha Ongeni (DW7), was the presiding officer at Mwongeni Youth Polytechnic Polling Station 2 of 4. She allowed agents to take photographs of the form 37A. That version was supported by Susan Ngare (DW4), the Presiding Officer at Kiabonyuru Primary School polling station 1 of 2 in Borabu.

188. I thus find that the quality of the images could be affected by the type of camera or cellphone used by the agents, lighting, angle and so forth. DW7 admitted that the official form 37A corresponds with the copy annexed by the petitioner. In short, it would be incorrect to condemn all the forms by petitioner as unauthentic.

189. The IEBC also complained that the petitioner was selective on the quality of forms 37A that he annexed. For instance, Sumbwa Ambuko (DW2) claimed that where the petitioner alleged disparities, his forms were blurred, illegible or with poor lighting. But in the images of forms in which he alleges alterations by IEBC, they were somehow taken in better lighting. I find that the allegation is not entirely true because with respect to forms 37A attached to affidavits of PW4 and PW5 (who complained of disparities in the forms 37A and 37C) the forms are clear and legible.

190. I will now turn to alterations, cancellations or overwriting. According to DW1, if there are errors in entering data into form 37A, the presiding officer can amend it by crossing, countersigning, stamping or make such other corrections.

191. I will start with one polling station where there is an allegation that a wrong form was used. The petitioner claimed that at Makairo Polling Station 2 of 2, the Presiding Officer used the form 37A for stream 1 of 2. In his view, the results from one stream are thus unaccounted for. IEBC’s witness, Sylvia Jepchumba (DW5) admitted the form for stream 2 was printed as that for stream 1 of 2. The digit 2 of 2 was thus inserted by hand.

192. I have closely examined the form. I find that the results in the two forms relate to independent streams. Each of the stations has its results signed off by agents and presiding officers in the respective stream. Two of those agents were in fact from the UDA Party. The two forms also have different serial numbers GB085686 and GB085870. The claim by the petitioner that the results of one stream were unaccounted for is thus untrue.

193. The petitioner submitted that IEBC failed to supply the original forms 37A for Ikobe Tea Buying Centre 1 of 1 and Kenyamware DEB Polling Station 1 of 1. The report by the Deputy Registrar states that IEBC did not supply the form for Ikobe TBC 1 of 1. The truth however is that a certified copy for Ikobe TBC 1 of 1 was filed by the IEBC in court on 7th November 2022. It has only one alteration in the last digit (7) in the score of Nyaribo of 237 votes. Nyambati got 94 votes there. The form is signed by the Presiding Officer, his deputy and 3 agents from KANU and ODM. The results tally with those in form 37B for West Mugirango.

194. I thus find nothing untoward in it. Furthermore, the complaint at Ikobe TBC 1 of 1 appears at paragraph 134 of the amended petition as one of the 64 polling stations in West Mugirango “that did not have agents”. At paragraph 139, the petitioner pleaded that form 37A was altered in the results for Nyaribo. Like I have stated in the preceding paragraph, there were three agents there. I would however grant it to the petitioner that none was from Nyambati’s party, UDA. I have also found that there was only one minor alteration of the last digit in Nyaribo’s votes.

195. However, at Kenyamware DEB Primary School 1 of 1, the certified copy of Form 37A displayed by the IEBC at page 159 of the response is too faint and illegible. The copy attached by the petitioner to the amended petition is equally unreadable. One cannot clearly see the candidates’ results or the signatures of the agents or presiding officers.

196. Noting the centrality of Form37A in elections, I ordered a partial scrutiny at Kenyamware DEB Primary School 1 of 1 limited to a recount of the votes. The report by the Deputy Registrar established that the 1st respondent had 39 votes while Nyambati garnered 280 votes. The Deputy Registrar could not also make a comparison with the form 37A from IEBC because, like I stated above, it was illegible.

197. But the recount left no doubt about the numbers. The true complaint by the petitioner can then only be that Nyambati was denied 60 votes in form 37C which gives him 220 votes. I will discuss shortly about the errors in transfer of data from forms 37A into 37C and the effect on the outcome of the results.

198. The petitioner submitted that at Motagara DEB Polling Station 2 of 2 the form 37A was “clearly incongruent with other forms annexed” to their bundle. The petitioner also claimed that agents were given differing copies of forms 37A. In one of the form 37A there, Nyambati had 88 votes but only 1 vote in the other. I ordered partial scrutiny there which supported the petitioner’s case that Nyaribo obtained 88 votes and not 1 vote as indicated in form 37A. That allegation is accordingly proved to the required standard.

199. I must however add that the scrutiny also revealed that the 1st respondent was denied 34 votes at that polling station. His score was 122 and not 88 as indicated in form 37A. Some candidates such as Momanyi had been unfairly added 82 votes; Moseti 4; and, Okioma 18. In short, I cannot say that the discrepancies there only affected Nyambati.

200. The petitioner singled out 38 forms with “multiple alterations” in the entries for the votes of candidates, the total number of valid votes, the registered voters or rejected ballots. The disputed returns in North Mugirango Constituency are highlighted at paragraphs 22 to 62 of his affidavit; those in West Mugirango Constituency at paragraphs 63 to 102; in Kitutu Masaba Constituency from paragraphs 103 to 116; and, from Borabu Constituency at paragraphs 113 to 123. I referred earlier to a table containing his analysis.

201. Examples can also be found at Ting’a DOK Primary School 3 of 3 (whose alteration is stamped), Nyairang’a Primary School 2 of 2 and Motagara DEB Primary School 2 of 2. I have dealt with the latter station in detail in paragraphs 198 and 199 of this judgment. I will analyze the issues at Nyairang’a Primary School 2 of 2 below.

202. One of the weaknesses in the petitioner’s case is that a number of his prospective witnesses who made extensive complaints about the results in forms 37A or the authenticity of IEBC’s returns did not take to the stand. They included Yabesh Mokua Ntabo, a chief agent for Borabu; Abednego Saisi, Constituency Manager for North Mugirango; and, Doris Magona Nyamanga a chief agent for West Mugirango. Their affidavits were consequently not adopted by the court and having not undergone cross-examination, they are of no probative value. I reached the same conclusion earlier on the failure by the 1st and 2nd respondents to testify or to offer any witnesses.

203. The petitioner questioned the high number of 53 rejected ballots at Nyankoba Polling Station 1; and, 34 rejected ballots at Nyabigege DOK Primary School Polling Station 1 of 2. I ordered, suo motu, for a recount and examination of the entries in Forms 37A at the two stations. From the Scrutiny Report, the ballot box for Nyankoba was intact. The recount showed that 48 ballots that had been marked rejected were actually valid. Candidate Bosire gained an additional 27 votes from that pile; Nyambati 11 and Nyaribo 4.

204. The ballot box for Nyabigege DOK Primary School Polling Station 1 of 2 was equally intact. Of the votes condemned as spoilt, 33 turned out to have been valid. Nyambati gained an extra 12 and Nyaribo 1 from that bunch. The obvious inference to be drawn here is that the petitioner as well as other candidates were affected by the erroneous decisions of the presiding officers. The two candidates most prejudiced by those actions were Bosire and Nyambati.

205. It is important to keep in mind that the results at the polling station are final. See IEBC v Maina Kiai & 5 others, Court of Appeal, Nairobi, Civil Appeal 105 of 2017 [2017] eKLR. The learned judges delivered themselves as follows-It is clear beyond peradventure that the polling station is the true locus for the free exercise of the voters’ will. The counting of the votes as elaborately set out in the Act and the Regulations, with its open, transparent and participatory character using the ballot as the primary material, means, as it must, that the count there is clothed with a finality not to be exposed to any risk of variation or subversion. It sounds ill that a contrary argument that is so anathema and antithetical to integrity and accuracy should fall from the appellant’s mouth. [underlining added].

206. The significance of the above precedent is that the primary and most important declaratory form is 37A: The forms 37B and form 37C series of forms are spreadsheets merely transferring and consolidating the scores in forms 37A.

207. I must add that the polling station affords an aggrieved candidate or agent additional safeguards. For instance, such a candidate or agent who is present has the right to a recount. Regulation 80 of the Elections (General) Regulations 2012 provides as follows: -(1)A candidate or agent, if present when the counting is completed, may require the presiding officer to have the votes rechecked and recounted or the presiding officer may on his or her own initiative, have the votes recounted: provided that the recount of votes shall not take place more than twice.(2)No steps shall be taken on the completion of a count or recount of votes until the candidates and agents present at the completion of the counting have been given a reasonable opportunity to exercise the right given by this regulation.

208. First and foremost, the petitioner or Nyaimbati’s agents failed to take advantage of this right except for one request at Nyairanga referred to by Antony Miranyi (PW11). He testified that the request for a recount was declined. It was not clear when or who made the request and why it was denied.

209. As I have indicated, Regulation 80 only applies to a candidate or agent who is present at the count. PW11 conceded in cross examination that he was not there during the counting of the votes. Upon further cross examination he admitted that the UDA Party had an agent at Nyairanga but “he was harassed”. The witness could not recall the name of the agent but he was “a presidential UDA agent”. I thus find that this claim was unproven.

210. I also find that some of the alterations or changes in forms 37A affected candidates’ results. I took those seriously and I will make specific comments on a number of the forms. However, others were mathematical errors in totaling the numbers of votes cast for candidates; or, the total number of voters who turned out to vote; or, the valid votes cast. Generally speaking, the summation or arithmetical errors did not affect the votes for each candidate.

211. A few examples will suffice for my analysis. DW7, the presiding officer at Mwongeni Youth Polytechnic Polling Station 2 of 4 confirmed that the form 37A displayed by IEBC was similar to the one produced by the petitioner. I find that it did not have any alterations in the results of candidates. However, the last digit was corrected in the figure of 448 for the total number of voters. She denied the allegations by the petitioner that there was a variance between the votes cast and those entered in 37C.

212. Susan Kemunto Ngare (DW4) equally conceded that there was an alteration in the total number of registered voters and the total number of valid votes cast but which did not affect the total votes of any of the candidates.

213. At Nyairang’a Primary School 2 of 2, the form contains a stamped alteration. It was signed by 2 agents. In Nyagokiani Primary School Polling stations 1 and 2 the forms are signed by 4 and 3 agents respectively while at Embaro DOK Primary School Polling Stations 1 and 2, the same were signed by 3 and 4 agents respectively. In all of those cases, the results in the primary form 37A were unaffected. The dispute seems to be on the transfer of the data from those primary forms into the eventual form 37C.

214. The same analysis applies to Chaina Primary School 1 of 1, all the six streams at Ikonge Primary School Polling Stations, Riamanoti Primary School 1 of 1, Nyangena TBC 1 of 1, Nyankono TBC 1 of 1, Nyabisimba DEB 1 of 1, Nyangorora DEB Primary School 1 of 1 and Omokirondo Primary School 1 of 1

215. I am alive that forms 37A are pre-printed forms that are populated by hand by the presiding officers. It is done at the end of the count. The entire exercise in some cases went for days and obviously took a toll on the IEBC officials and agents. For instance, DW4 testified that her team “completed the count after 24 hours and may account for some human errors”. Errors are bound to arise and may be corrected by crossing, countersigning or stamping. However, the exercise should be transparent. That is why agents sign the form and are given copies.

216. Although there were allegations that officers employed by the 3rd and 4th respondents deliberately or fraudulently altered the results in forms 37A to favour Bosire or Nyaribo, there was no concrete evidence on the point. The partial scrutiny and recount showed that erroneous decisions by presiding officers, say in Nyabigege DOK Primary School 1 of 2; and Nyankoba Polling Station 1 affected other candidates. But like I said earlier, the two candidates most prejudiced by those actions were Bosire and Nyambati.

217. The petitioner contended that Nyambati’s agents were blocked from polling stations or ejected altogether. He averred that 138 polling stations did not have UDA or Nyambati’s agents “due to discriminative restrictions imposed upon them by the 3rd and 4th respondents”. He listed 39 polling stations in Borabu; 64 in West Mugirango; and, 42 in North Mugirango. In some instances, the agents were allowed into the stations late.

218. Elvis Simba Kibeni (PW10) testified that at Nyaisa SDA Primary School Polling Station 1, Manga TBC Polling Station 2 and Mokwerero DEB Polling Station 1, the presiding officers were “so biased that at some point, some of our agents were ejected”.

219. At paragraph 9 and 10 of his deposition, he avers that Nyambati’s agents were chased away from polling stations; and, that he had to intervene for them to be admitted back. However, he conceded that he did not attach a list of those witnesses. He also conceded that he did not specify the names of agents who were denied access.

220. Walter Maake (PW12) testified that he successfully intervened to have some agents admitted although some were allowed access as late as 3:00 p.m. However, his efforts did not bear fruit at Tonga Primary School 2 and Bugo Primary School 1 of 1.

221. He also conceded that some forms 37A that were issued were signed by UDA agents for instance the one for Nyaisa SDA Polling Station annexed to the affidavit of PW8 and the one for Manga TCC Polling Station 2 of 2 at page 343 of Volume 5. The witness admitted that he never raised issues regarding forms 37As for Kitutu Masaba.

222. In many of those cases, the results in the primary form 37A were unaffected. I am alive that the true dispute is whether the personal agents of candidate Nyambati were present or signed the forms and whether it affected the result. But I have also noted that some of those agents are from the UDA Party, on whose ticket he was running.

223. PW13 claimed he was denied entry into the county tallying center but upon cross examination he admitted that he was denied entry by security officers outside the premises and despite showing them his badge and credentials. He conceded that he never got in touch with the County Returning Officer (DW1). As I will discuss later, DWI confirmed that PW13 was not present when both 37C forms were signed or when the final announcement was made.

224. It is thus imperative to examine the right of a candidate to have agents at polling stations. Section 2 of the Act defines agent in the following terms:Agent means a person duly appointed by-a.A political party or an independent candidate for the purposes of an election under this Act; orb.A referendum committee for the purposes of a referendum under this Act, and includes a counting agent and a tallying agent.

225. The Regulations on the other hand define agent “as a person appointed under section 30 of the Act”. Section 30 in turn provides as follows-(1)A political party may appoint one agent for its candidates at each polling station.(2)Where a political party does not nominate an agent under subsection (1), a candidate nominated by a political party may appoint an agent of the candidate's choice.(3)An independent candidate may appoint his own agent.(3A) A registered referendum committee may appoint one agent at each polling station. [underlining added]

226. It follows that a party candidate would only be entitled to appoint an agent if his party does not nominate one. My considered opinion is that any offer to party candidates to appoint their own agents or additional agents would offend section 30 of the Act. Such an offer would at best be gratuitous.

227. Fundamentally, Nyambati was a UDA Party candidate. He was not an independent candidate. The petitioner and some of his witnesses conceded that the party had agents at polling stations. Doubt is completely removed because a wide number of the results declaration forms were signed by UDA agents.

228. It is true that some of those agents were dedicated to the presidential election. But I have kept in mind that there were 12 candidates running for governor. Quite apart from demands of space at the polling stations, I take judicial notice that all the six elections were being conducted at the same time in each polling station; and, that such agents remained party agents. I thus find there was no transgression of the law or Regulations merely because Nyambati did not have his personal agents in some of the polling stations.

229. There was then an allegation by Chrisantus Nyakundi (PW8) that the Presiding Officer at Nyaisa SDA Primary School Polling station 1 “was assisting some voters alone without involving the agents” and that when he complained “she confiscated [his] badge and threw [him] out”. He claimed that he was not allowed back until 23:00 hours by which time the count had ended.

230. The claim by this witness can be misleading. Under Regulation 72 (1) where a voter suffers from disability or is illiterate for example, the presiding officer shall permit the voter to be assisted by a person of the voter’s own free choice. The involvement of agents would only come in under Regulation 72 (2) if the voter who requires assistance is not accompanied by a person qualified to assist him. In that case the presiding officer shall assist the voter in the presence of agents.

231. Another issue that arose related to authenticity of the IEBC stamps affixed to some forms 37A. For instance, at Motagara DEB Primary School 2 of 2, the petitioner as well as Wesley Kerina (PW9) referred to multiple alterations in the form. The IEBC stamp impression has a square border and appears different from the round one in other certified forms.

232. There is a similar impression on the form for Nyangoge Primary School 1 of 3. The County Returning Officer (DW1) theorized that the square imprint resulted from “a drying ink pad and use of pressure by the maker”. It may well be so. He emphasized however that they are all IEBC stamps.

233. I hold the opinion that the lack of an IEBC stamp on Form 37A or the presence of it in an unfamiliar shape is not fatal. For starters, there is no formal space set out in the form to affix the stamp. Secondly, the IEBC stamp constitutes what the Regulations refer to as non-strategic materials. Thirdly, there is no express requirement in the Act or Regulations that the Presiding Officer must stamp Form 37A. This is unlike the case of a ballot paper which must be stamped. See Sammy Kemboi Kipkeu v IEBC & 2 others, Eldoret High Court E.P. 2 of 2017 (unreported). See also IEBC & another v Stephen Mutinda Mule & 3 others, Nairobi, Court of Appeal, Civil Appeal 219 of 2013 [2013] eKLR.

234. There is then the matter of missing signatures by agents. The signatures by agents are not mandatory. Regulation 79 (6) and (7) provide that the absence of an agent or candidate at the time of announcing results does not invalidate it. Form 37A provides for signatures by agents or candidates if present. Aggrieved agents or candidates have the right to protest or refuse to sign the form. Doubt is removed completely because at the foot of the form, there is a space intituled: Refusal to Sign (if any). See also Oscar Omoke Ocholla & 2 others v IEBC & 4 others, Court of Appeal, Nairobi, E.P. Appeal 23 of 2018 [2018] eKLR; Sammy Kemboi Kipkeu v IEBC & 2 others, [supra].

235. One intriguing aspect of the amended petition is that there was no significant attack on forms 37B from the constituencies. For instance, when Elvis Kibeni (PW10), the chief agent for Nyambati in Kitutu Masaba, was referred to paragraph 13 of the affidavit of Sumbwa Ambuko (DW2) and the form 37B showing results of all the gubernatorial candidates in that constituency, he (PW10) said he had no complaints with it.

236. This is also clear from the rebuttal by the four Constituency Returning Officers, DW2, DW3, DW5 and DW6. For instance, Ruto Wesley (DW6) was the Constituency Returning Officer for North Mugirango Constituency. He never received a complaint regarding the form 37B. It partly means that the petitioner was largely satisfied with the entries of data from the primary forms 37A into the 37B series but was predominantly aggrieved by the entries into the impugned forms 37C.

237. That said, the re-tally exercise by the Deputy Registrar revealed that the total variance between the entries in form 37B for Kitutu Masaba and form 37C for all the total votes garnered by the 12 candidates was 200 votes only. That is derived from the figure of 69,127 indicated by IEBC. However, the re-tally shows that the correct figure should be 69,073.

238. From the re-tally exercise, the rejected votes in Kitutu Masaba are 912 in form 37B but recorded as 914 in form 37C; a difference of two votes only. In Borabu, it revealed a variance of 1013 votes across the forms 37A, 37B and 37C. The total rejected votes in form 37A were 568 whereas in forms 37B and 37C they are shown as 564 and 553 respectively. In North Mugirango, the total rejected votes in form 37A were 440 whereas in forms 37B and 37C they are 439, a difference of 1 vote.

239. In West Mugirango, the votes cast in form 37B and 37C tallied save for the item of rejected votes. The latter were 884 in form 37B and 885 in form 37C; a difference of 1 vote. I have kept in mind that IEBC failed to provide the original form 37A for Ikobe TBC Polling Station. But as I observed earlier a clear and certified carbon copy was lodged in court on 7th November 2022.

240. The petitioner also led evidence to show that in many instances, the results from the primary forms 37A were not accurately reflected in form 37C. Learned counsel for the petitioner submitted that given the magnitude of the transposition errors and the number of votes involved, the conduct was not in good faith or an honest mistake but a deliberate and well-calculated scheme to benefit the 1st and 2nd respondents.

241. I have paid close attention to the evidence from Elvis Kibeni (PW10) on this issue. He attached forms 37As but not the impugned form 37C. He said that he received the forms 37A between 9th and 10th August 2022 but only saw form 37C on 12th August 2022.

242. A few other examples will suffice.At Omonono Primary School 2 of 2, the IEBC conceded at paragraph 76 of the response that a “transpositional error” occurred when transferring data from Form 37A into 37C. An erroneous 0 in the column for one candidate, Momanyi George, pushed data to other columns affecting the results of four other candidates.

243. At Nyamonge Primary School 1 of 2, the petitioner and Andrew Nyangau Ombogo (PW6) testified that Nyambati had 36 votes as reflected in Form 37A but was given 0 in Form 37C. The error was freely admitted by IEBC in paragraph 75 of its response explaining it as a “transpositional error”.

244. A similar error is found in Riabaruoti Market 1 of 1, where the 1st respondent garnered 105 votes reflected in Form 37A but was added 10 votes in Form 37C. This was clear from the evidence of Omariba Joseah Gwaro (PW4) and was conceded by the IEBC.

245. At Kerongo DOK Primary School 2 of 2, the 1st respondent obtained 115 votes as per Form 37A but which was inflated to 150 votes in Form 37C. At Tinga DOK polling station, Nyambati garnered 106 votes as per form 37A but in form 37C he was given 1 vote. There is an express admission of those errors at paragraphs 70 and 79 of the IEBC Response. The IEBC has equally owned up to similar errors at Nyasumi DOK Primary School 1 of 1.

246. But in all those cases, the results in the primary form 37A were unaffected. The IEBC conceded at the trial that owing to wrong data entry in Form 37C, Nyambati lost a total of 432 votes while the 1st respondent gained 17 votes.

247. But the arithmetic is now a little different due to the results of the partial scrutiny and recount. There were additional votes from the recount or revision of the true number of rejected ballots. At Nyabigege DOK Primary School Polling Station 1 of 2 Nyambati gained an extra 12 and Nyaribo 1. At Nyankoba Polling Station 1 candidate Bosire gained an additional 27 votes from that batch; Nyambati 11 and Nyaribo 4.

248. As detailed earlier, the partial scrutiny at Motagara DEB Polling Station 2 of 2 also established that Nyambati garnered 88 votes and not 1 as indicated in form 37A. The 1st respondent on the other hand obtained 122 votes and not 88 as indicated in the same form.

249. Leaned counsel for the petitioner submitted that there was “reverse engineering of form 37C to reach a pre-determined outcome”. The key evidence on this issue was from PW13. He was not present when both 37C forms were signed or when the final announcement was made. He said that on 14th August 2022, DW1 sent a WhatsApp message to Nyambati attaching the initial form 37C. The screenshot appears at page 458 of Volume 5.

250. On 22 August 2022, DW1 sent Nyambati a copy of the second form 37C. According to the first form, the voter turnout was 179,719 and the total valid votes cast were 176,929. In the subsequent form 37C, the votes cast were 207,096 which PW13 claimed burst the dam of voter turnout of 179,719 and the valid votes cast of 176,929.

251. In his written submissions, the petitioner contended that DW1 failed to give a reasonable explanation for the 10 days’ delay before he shared the final form 37C with Nyambati; or, why he had shared the earlier form with the same candidate 2 days after the declaration of results. The petitioner opined that the second form was generated to mollify “the uproar in the social media”.

252. Like I stated earlier, unlike the pre-printed forms 37A, form 37C is an Excel spreadsheet merging the data from the primary forms 37A and 37Bs. DW1 does not deny that he sent the two messages or that there were two differing forms 37C. He said that when he prepared the first form 37C he signed it together with four agents but discovered it had some errors. When re-examined about the issue the witness said-In the draft 37C, the votes for Bosire and Ondieki were not captured as totals per polling stations for Kitutu Masaba, North Mugirango and part of Borabu as expounded in paragraph 33, 34, 35 and 36.

253. He then prepared a second form “about an hour later” which he signed together with the same agents. In his opinion, form 37C remained a draft until the formal announcement of results. In his own words: “It does not matter even if the form is signed. Corrections can be done resulting in a new form”.

254. But what DW1 claimed to be a draft was in fact a form that had been dated and fully executed by him and the agents present. At that moment, it may as well have been used to declare the winner of the gubernatorial election. I thus refuse to accept his casual answer that it was a draft. When he now prepared the second form to correct the errors above, he did not announce it to the public. True, he executed the second form together with the same agents but the element of transparency was beginning to fade away.

255. I however find the following. There is no express Regulation that bars the County Returning Officer from correcting such errors in form 37C. The Regulations only require the process of tallying to be done in the presence of candidates, agents or observers if present. Secondly, the existence of an error in such a form is not sufficient to annul the results. In Oscar Omoke Ocholla & 2 others v IEBC & 4 others, [supra] the Court of Appeal was faced with a form 35B that had an error in the number of registered voters. The learned judges delivered themselves thus-With respect to the form 35B we are satisfied that the error in the total number of registered voters did not affect the final results of the elections, and therefore such error was also not sufficient to nullify the elections.

256. Furthermore, the results for the candidates in both disputed forms remained the same. The introduced changes related to the number of total number of voters who turned out to vote, the valid votes cast and the rejected ballots.

257. Keeping in mind the decision in IEBC v Maina Kiai & 5 others [supra] about the finality of the results from the polling centre, I find that DW1 or IEBC did not deliberately or fraudulently alter the results of candidates as captured by the presiding officers in forms 37A. However, owing to wrong data entry in Form 37C, Nyambati lost a total of 432 votes while the 1st respondent gained 17 votes.

258. I have analyzed at length the errors, cancellations, over-writings or alterations to the primary forms 37A. I stated earlier that there was no significant attack on the validity of forms 37B from the four constituencies in Nyamira. Nevertheless, I have analyzed the variances between forms 37A and the entries into forms 37B. In particular, I have highlighted the key findings from the scrutiny report. Whereas the transfer of data from forms 37A into the 37B series generated lesser heat, the petitioner raised serious grievances with what the IEBC termed a “draft form 37C” and the final form 37C used to declare the winner.

259. Regarding the impugned forms 37C, I must restate that unlike the pre-printed forms 37A, form 37C is meant to amalgamate the data from the primary forms 37A and 37Bs. It is a critical form used to declare the final results of the gubernatorial election. It is also the foundation for the final certificate of the elected governor in Form 37D.

260. In the initial form 37C, the total number of voters who turned out to vote were recorded as 179,719 (55. 59%). But in the second form, they are indicated as 207,096 (64. 06%). Again, in the first form, the second-last column lists the valid votes as 176,929 while the final form 37C states a figure of 207,096. The rejected ballots in the first form are 2,790 and 2,791 in the subsequent form. Learned counsel for the petitioner thus submitted that the voter turn-out was a moving target.

261. In examining the certified final form 37C it is critical to keep in mind the centrality of the primary forms 37A; and, the binding precedent in IEBC v Maina Kiai & 5 others, [supra]. Secondly, not all registered voters turn out to vote. And those who do cast some ballots that may be rejected. I am persuaded by the simple logic that voters mean the people who cast their votes for a candidate. Those are the “voters who turned out to vote”. It must follow that rejected ballots are not votes. It also means that valid votes and votes cast mean the same thing.

262. Since the first form 37C had been signed by the County Returning Officer as well as four agents, I found it irregular and self-serving to call it a “draft”. Furthermore, DW1 was not completely open about it. He admitted that he did not announce to the public at the tallying center about the draft. He claimed that he noted the mistake in the first form 37C before he declared the results and informed the agents present. He did not also look out for any candidates who may have been present. Like I said, it compromised the element of transparency.

263. That said, I am satisfied that the first form 37C contained some errors. For instance, the total number of valid votes cast for all candidates appearing in the second-last column of the form did not include the votes of two candidates, Bosire Anthony Moseti and Bwo’ndieki John Nyariki in three constituencies of Kitutu Masaba, North Mugirango and Borabu.

264. As explained by DW1, they were “arithmetical errors in terms of summation of the total number of valid votes cast discernable from the total number of votes garnered by each candidate such that from the said entries, the total number of valid votes cast is 207,096 (with Osiemo’s votes being 785 and not 985) and not 179,719”.

265. I dealt at great length with the transpositional errors in the first form 37C and which were largely admitted by the IEBC at paragraphs 68, 69,70,71,72 and 73 of the affidavit of DW1. Firstly, the combined errors of transferring the data from forms 37A into 37C denied Nyambati 432 votes and added Nyaribo 17 votes. The votes of all the other candidates were unaffected. But to this, one must now add the votes gained or lost from the scrutiny exercise that I highlighted at paragraphs 247 and 248 of this judgment.

266. Secondly, the final form 37C that was used to declare the results was amended primarily to correct the figure of the total valid votes cast or the contentious number of voters who turned out to vote. I have dealt earlier with the changes and definitions of those terms. Fundamentally, the changes in the second form 37C did not affect the votes for any candidate.

267. In the end, the claim by the petitioner that the final form amounted to “reverse engineering” or fraud to reach a pre-determined outcome was not proved to the required standard for a number of reasons. Firstly, the petitioner admitted that he was not an agent in the election. He did not thus participate in the preparation or execution of the two impugned forms. He did not receive a copy of it either. Secondly, his key witness, Kinaro Ndubi (DW13), was not also at the county tallying centre. The latter relied largely on information obtained from a candidate, Nyambati. Nyambati on the other hand did not take to the stand. I have reached the conclusion that it is classic hearsay. Thirdly, it was common ground that the results were announced on 12th August 2022. Fourthly, the petitioner also failed to prove that the second form was fraudulently made ten days after the declaration of results.

268. There is no doubt that the inquiry by the election court has revealed a number of discrepancies and irregularities in the poll. For instance, the petitioner proved that a ballot box for Omonono polling station 2 of 2 which is situated in Borabu Constituency was found at Kebabe Girls, in North Mugirango. Another example was at Motagara DEB Polling Station 2 of 2 where the partial scrutiny established that Nyambati obtained 88 votes and not 1 vote as indicated in form 37A. The 1st respondent on the other hand garnered 122 votes and not 88 as indicated in form 37A. The exercise also significantly reduced the number of rejected ballots at Nyankoba Polling Station 1 and Nyabigege DOK Primary School Polling Station 1 of 2.

269. When all those flaws are juxtaposed against the overall findings in the scrutiny report and the remaining wide margin between the two leading candidates, I have reached the conclusion that they did not substantially affect the results. I am not thus satisfied that there was a choreographed scheme to systematically deduct the votes from Nyambati and credit them to candidates Bosire and Nyaribo; or, that there was a fraudulent stratagem contrived by the IEBC to subvert the will of the people of Nyamira.

270. I also found that the petitioner failed to prove beyond reasonable doubt that the 1st and 2nd respondents engaged in conduct of a criminal nature or a corrupt practice contrary to sections 14 and 15 of the Election Offences Act.

271. Like I stated, no election is perfect, and less so a county-wide poll with hundreds of polling stations. To be precise, this was a gubernatorial election covering four constituencies with 643 polling stations. It was also taking place in the backdrop of a general election for six electoral offices. Like I observed in Kakuta Maimai Hamisi vs Peris Pesi Tobiko [supra], it would be impractical to demand or expect mathematical precision.

272. It was never intended in the design of our laws to re-locate the campaigns to the election court. I concur on this point with my learned brother, Majanja J, in Jackton Ranguma v IEBC & 2 others, Kisumu High Court E.P. 3 of 2017 [2018] eKLR that-An election petition is not a do-over of the just concluded election. It is not an opportunity to conduct another election through the court as every election conducted in accordance with the law is presumed valid unless it is set aside by the court. [underlining added]

273. I will go back to the beginning: Section 83 of the Elections Act contains a rebuttable presumption in favour of the respondents that the election was conducted properly and in accordance with the law. It also implied by that provision that elections are not always perfect.

Disposition and Final Orders 274. My answer to issue number (ii) framed earlier is thus partially in the affirmative. I readily find that some of the records of the election in some forms 37A, 37Bs and 37C were inaccurate or tainted by irregularities or malpractices. However, the totality of those anomalies do not substantially affect the numbers or the outcome of the poll.

275. From the partial scrutiny and recount of votes in some of the disputed polling stations, the 1st respondent’s lead remained unassailable. Furthermore, the election court was dealing with specific complaints pleaded in the amended petition and the quality of the evidence. I would be hard pressed to say that the overall result did not reflect the will of the people.

276. In view of the centrality of Forms 37A; and, the binding precedent in IEBC v Maina Kiai & 5 others, [supra], I am not persuaded that the election was not conducted in accordance with the principles laid down in the Constitution and written law. It follows that my answer to issue number (i) is in the affirmative: The gubernatorial election for Nyamira held on 9th August 2022 was substantially conducted in accordance with the Constitution, the Elections Act and Regulations.

277. On a preponderance of the evidence and the authorities, the 1st and 2nd respondents were validly elected as the Governor and Deputy Governor for Nyamira County. That equally answers issue number (iii) in the affirmative.

278. It must then follow as a corollary that the answer to issue number (iv) is in the negative. The petitioner did not marshal sufficient evidence to the required standard of proof that inures in his favour. Accordingly, he is not entitled to overturn the results as prayed. The upshot being that the amended petition is hereby dismissed.

279. That leaves issue number (v) on costs. Costs ordinarily follow the event and are at the discretion of the Court. Section 84 of the Act provides that an election court shall award the costs of and incidental to a petition and such costs shall follow the cause. Rule 30 (1) (b) of the Petition Rules empowers the court to set the maximum of costs payable.

280. I grant the respondents costs of the petition. Those costs shall be paid by the petitioner.

281. Learned counsel for the respondents prayed that I certify their costs for two counsel. I see no justification for it and that prayer is declined. Although many witnesses testified; and, a tome of materials put forth, the amended petition was fairly straightforward and did not raise complex questions of law or evidence. I also note that all the evidence was taken within a week.

282. The instruction fees for the 1st and 2nd respondents (who never took to the stand or called any witness) are capped at Kshs 1,000,000. The instruction fees for the 3rd and 4th respondents are capped at Kshs. 2,000,000. The Deputy Registrar of this Court shall tax the Bill of Costs under Rule 31. I direct that the whole or any part of those costs shall be paid pro-rata to the respondents from the money deposited in court by the petitioner.

283. Lastly, a certificate of determination of this petition required under section 86 of the Elections Act shall issue forthwith.

It is so ordered.

DATED, SIGNED AND DELIVERED THIS 16TH DAY OF FEBRUARY 2023. KANYI KIMONDOJUDGEMessrs Mokaya, Maeche, Mokua and also holding brief for Mr. Omwanza for the Petitioner instructed by Nchogu Omwanza & Nyasimi Advocates.Messrs Anyoka, Ligunya, Kamwaro and Ms. Maina for the 1st & 2nd respondents instructed by Anyoka & Associates Advocates.Mr. Amimo and also holding brief for Mr. Muyundo for the 3rd & 4th respondents instructed by D. W. Muyundo & Associates Advocates.Ms. Anita, Ms. Aminah & Mr. Terer, Court Assistants.