The republic of uganda in the supreme court of uganda at kampala

The Commission (may) ……….otherwise adapt any of those provisions as may be required to achieve the purposes of this Act …….to such extent as the Commission considers necessary to meet the exigencies of the situation.” (Emphasis is added)

Clearly, it is a condition for the exercise of that power that adapting provisions of the Act to meet the exigencies of the situation must be restricted or limited to what is required to achieve the purposes of the Act to exigencies of the situation. It follows that the power cannot be invoked to adapt a provision of the Act, if doing so would defeat the purposes of the Act. In my view reduction of the period for display of the voters’ rolls was not required to achieve the purposes of S.25 of the Commission Act. On the contrary it served to defeat those purposes. In my considered opinion therefore, the reduction of the period for display was ultra vires the powers of the Commission.

(3) If it be true, as deponed by Chairman Kasujja, that the tribunals to handle complaints, were set up during the update exercise, that was inconsistent with the letter and spirit of the provisions in S.25 of the Commission Act. Tribunals are supposed to handle disputes generated from objections raised under S. 25(3) “during the period of the display of the voters’ roll.” In the circumstances I could not see what the tribunals he mentioned could have handled during the updating exercise. I am inclined to infer that during the short display of 3 plus 2 days no fresh Tribunals were set up to handle objections as required in sub-section (5).

The importance of the system of voter registration in a democratic election cannot be over emphasised. It is through it, that the voters to participate in the election are identified and ascertained. To ensure that only persons entitled to vote remain on the voters’ register the two exercises of updating and displaying the register must be carried out diligently and meticulously. In the instant case, owing to what I must call self-imposed shortage of time, the 2nd Respondent did not execute the exercises diligently, let alone meticulously. However, there was not sufficient reliable evidence to show the overall impact of this failure.

For the Petitioner, Frank Mukunzi deponed that he carried out an analysis of data related to the Presidential election which “revealed an error margin of over 50% in the Electoral Commission’s figures of the voters’ register”. Counsel for the Respondents strongly criticised that evidence on two grounds. First, his C.V. revealed that Frank Mukunzi had no professional training in statistics, and he was therefore, not competent as an expert witness; secondly his analysis was based on erroneous or doubtful premise. I was not inclined to disregard the evidence purely on the ground that the deponent lacked formal training in statistics. The subject of his report was one which he probably could handle from the practical experience in data analysis he claimed to have. However upon going through the report I was not impressed that it was well founded. Accordingly I did not place any reliance on it and I do not find it necessary or useful to discuss it here.

Nevertheless, there was scanty and scattered evidence to the effect that some names of deceased persons and of non-citizens who had left the country, were still on the register on polling day. And conversely there was equally scanty evidence of persons entitled to vote but who did not vote because their names were not on the register. Needless to say that this was largely a consequence of not carrying out the two exercises properly. However, the proportion of names which wrongfully remained on the voters’ register was not established. Nor was the proportion of the eligible voters who were disfranchised because their names were omitted from the voters’ register. I should point out that the 2nd Respondent’s contention that failure to clean up the register would not per se be a ground for annulment is not entirely correct. If the failure translated into disfranchising a large proportion of citizens entitled to be registered as voters, it could be construed as non-compliance with the principle of voter registration which affected the result. That, however, was not the situation proved in the instant case.

Unlawful possession of voters’ card:

A related irregularity the Court received affidavit evidence on was unlawful possession of voter’s cards. A voter’s card is the identification mark that links the voter to his registered name. The voter has to produce it to the Presiding Officer in order to be given a ballot paper for voting. Under the Commission Act it is an offence for a registered voter to hold more than one valid voter’s card (see S. 26); and it is also an offence for any person to be unlawfully in possession of any voter’s card which is issued in the name of a voter or which is blank (see S. 28).

There were more than a dozen witnesses who deponed that they saw different people in unlawful possession of voter’s cards distributing them to persons not entitled to the cards. I will mention a few. Ojok David Livingstone of the Petitioner’s monitoring team for Mbale Municipality deponed that on polling day, following a tip-off about a lady distributing voter’s cards, he went with a police officer to look for Nakintu and found her at her home. She admitted that she had distributed 11 voter’s cards out of 50 she had received for distributing to the 1st Respondent’s supporters. She surrendered the balance of 39 voter’s cards together with a bottle of JIK, a tablet of soap and a drying rug which she said were to be used to remove marking ink from the finger of anyone who had voted. She was taken with the voter’s cards and the other items to the police station. Wafidi Amiri also monitor for the Petitioner in Mutoto, Bungokho, Mbale deponed that on polling day after observing suspicious conduct of the R.D.C., the Sub-county Chief, the local Movement Chairman, and one Sonya David, he followed the Sub-county Chief who was being driven by Sonya and when he caught up with the latter he found him carrying about 50,000 voter’s cards. He raised alarm and with assistance of the public arrested him to Mbale Police Station. The Sub-county Chief, Wamae Kenneth deponed in reply that he was ambushed by Wafidi Amiri. When he heard Wafidi raising alarm that the tax collector was stealing votes, he run away for fear of being lynched. He explained that he was returning the balance of undistributed voter’s cards to the sub county headquarters and that they were not 50,000 but less than 3000. I did not believe the Chief’s explanation. His conduct of running away showed that he knew he was in the wrong. Others witnessed less quantity of cards. Maliki Bukoli of Mbale Municipality witnessed the police arrest on Mukonge with 5 blank voter’s cards. Sulait Kule, the Petitioner’s monitor in Kasese received from one Kanunu 16 voter’s cards with a report that they had been given to him to supply to others for voting illegally. He took them to Kasese Police Station. Kakuru Sam of Rukungiri deponed that on 7th March when he collected his one voter’s card, he saw one Nshekanabo being given a stack of about 30 voter’s cards. Karenzyo Eliphaz also of Rukungiri, deponed that on 6th March at Rwenyerere polling station in Kihihi, he witnessed many new voters being openly denied their voter’s cards on the ground that they were the Petitioner’s “rebels;” and later their voter’s cards and those of voters who had died or migrated were given to LC I Chairmen to distribute to the owners which they never did. This offence appeared to have been spread, but the extent to which the voter’s card in unlawful possession were used to facilitate non-voters to vote or to enable registered voters to vote more than once, was not established.

Irregular Voting:

In sub-paragraphs 3(1) (i), (j), (m), (o) and (x) of the petition, the Petitioner pleaded several forms of irregular voting, namely stuffing ballot boxes, multiple voting, pre-ticking of ballots, and under-age voting. Evidence brought to prove these irregularities were of two types. First there was direct evidence of witnesses to the irregularities. Secondly there was documentary evidence which the

Petitioner contended showed, and invited the Court to infer from, that there was massive rigging through stuffing and multiple voting. I will discuss the latter first.

Copies of a large number of “Declaration of Results Forms” were produced in the evidence as annextures to a number of affidavits. It is important to understand the particulars on the forms on which the said contention was based. On the form, the Presiding Officer was required to enter not only the number of votes cast for each candidate, but also to record-

• the total number of votes cast for all the candidates

• the total number of rejected (invalid) ballot papers

• the total number of ballot papers counted

• the total number of spoilt ballot papers

• the total number of ballot papers issued to the station, and

• the total number of unused ballot papers

As I understood the lay out, the total ballots “counted” include “valid” ones cast and “invalid” ones rejected, but exclude the “spoilt” ones. Then the three (valid, invalid, and spoilt) with the “unused” balance added together give the total number of the ballot papers officially “issued to the polling station” However, the filling of the forms produced in evidence, did not all conform to that understanding. In many of them, the figure entered for the total ballot papers issued, equaled the aggregate of the valid, invalid and spoilt ones only, excluding the unused ones. That gave the impression that all ballot papers issued to the station had been used up, and that there were unexplained ballot papers that remained unused. Much reliance was placed on this, by the Petitioner and the several witnesses who annexed copies of those forms to their affidavits, to advance a theory to the effect that the unused ballot papers, being over and above the total “issued to the polling station”, were unexplained extras, and must be equivalent to ballot papers irregularly introduced or stuffed into the ballot box. To illustrate the point the Petitioner annexed copy of one such form to his affidavit accompanying his petition, and later, he annexed to his Affidavit in Reply to the 2nd Respondent, 93 copies of similar forms from different polling stations in 19 districts. Copies of other similar forms were annexed to affidavits of other witnesses. In his supplementary affidavit, Frank Mukunzi, deponed that he had “analysed the Declaration of Results from 254 polling stations distributed throughout the Country.” He annexed to the affidavit a summary of findings in which he concluded:

“My analysis reveals that in the 254 polling stations that I have analysed approximately 34.9% of the votes tallied for a presidential candidate are “ghost votes.” Due to the randomness of the process by which I selected this data for analysis, I have no reason to believe that the percentage of ‘ghost votes’ nationwide would differ from the percentage in my sample population — this means that approximately 2,579,802 of the 7,389,691 votes tallied nationwide for one or another presidential candidate are ‘ghost votes’ and should not be counted.”

The analysis was based on the same theory that the forms showed that there were more ballot papers that what had been issued to the polling stations. Frank Mukunzi did not produce the forms that were subject of the analysis. I assume that they were not different from those produced in Court, which I have carefully scrutinised. The forms produced in Court did not all show the same pattern. Some were filled in the manner I understood they were supposed to be filled, so that the figures tallied well, showing no “extra” ballot papers. Others showed totals that were incomprehensible. There were even a few where the section for totals was left unfilled.
My considered conclusion, after thorough scrutiny of the forms, was that the more probable explanation was that the Presiding Officers, made erroneous entries either through miscalculations or through misunderstanding of what was required to be filled especially in the space for ballot papers “issued to polling station.” During submissions, the learned Solicitor-General, suggested that there was a misprint and that the expression should have been “issued at polling station,” instead of “issued to polling station”. With due respect, I do not agree. However, a Presiding Officer who understood it in the same way suggested by the learned Solicitor-General, could, and probably did, record the total of issued ballot papers excluding the unused ballot papers.

To illustrate that lack of understanding, or misinterpretation of the form, may be the reason behind the untallying figures, I will use four forms, two specifically referred to by the Petitioner, and two others which I picked at random from two different areas but which were produced in evidence to prove the theory. In the affidavit accompanying the petition, the Petitioner asserted that at Bukaade Primary School Polling Centre “the number of votes cast exceeded the number of ballots issued for the polling station”, the former being 856 votes, and the latter being 650 ballot papers. The form was in respect of results at Bukaade P/S polling station Code 04; Sub-county Bukanga: Code 01; Constituency Luuka County: Code 040; District Iganga: Code 07. The votes recorded as obtained by the candidates were: Awori and Bwengye 0 each, Besigye 60, Karuhanga 2, Kibirige and Museveni 397 each, which totals 856 votes as deponed by the Petitioner. However, the figures filled in the section for totals were:

Total number of —

• valid votes cast for candidates 459

• rejected (invalid) ballot papers 05

• ballot papers counted 464

• spoilt ballot papers 03

• ballot papers issued to polling station 650

• unused ballot papers 183

Evidently the Petitioner’s assertion in the affidavit conformed with his theory. As the form stands the 183 unused ballot papers appear to be excess. However a closer look shows that the totals are incomprehensible. To start with if the numbers of votes shown for each candidate are added up, they do not come to 459, which was the figure filled in as the total of “valid votes cast for candidates.” It becomes comprehensible as the total, only if the 397 votes recorded for either Kibirige or Museveni is deleted as a mistaken entry. If that is done then all figures fall in place. The valid cast and invalid rejected add up to the 464 counted ballot papers. Then the valid, invalid spoilt and unused (i.e. 459 + 5 + 3 + 1 83) add up to 650 as the ballot papers issued to the station. I was convinced that this was a case of mistaken entry on the form and not evidence of ballot box stuffing.

In his affidavit in reply to the 2nd Respondent, the Petitioner also deponed that the number of ballot papers issued to the polling station at Ishaka Adventist College were 477 and were equal to the ballot papers counted, yet there were 253 ballot papers unused. The form relating to that station was, with several others, annexed to the affidavit of John Tumusiime, the Petitioner’s Chairman for Bushenyi, who also averred that it showed an “unexplained anomaly of a large number of unused ballot papers, where the ballot papers issued to the station did not exceed those actually used.” The particulars of the station are Polling Station Ishaka Adventist College: Code 02; Ward IV: Code 20; Sub-county Ishaka Town Council: Code 01; Constituency Igara County West: Code 019; District Bushenyi: Code 04. The recorded votes obtained by each candidate were: 1 for Awori, 1 44 for Besigye, 0 each for Bwengye, Karuhanga and Kibirige, and 330 for Museveni, which add up to 475 votes. The figures filled in the totals section were as follows:
Total number of:

• valid votes cast for candidates 475

• rejected (invalid) ballot papers 2

• ballot papers counted 477

• spoilt ballot papers 0

• ballot papers issued to polling station 477

• unused ballot papers 263

In his submissions to the Court, Mr. Mbabazi, the Petitioner’s counsel who handled this issue, specifically referred to this form in support of his argument for the theory. He contended that the 263 ballot papers reflected the number of ballot papers which were obtained other than from the official issue, and were illegally stuffed in the ballot box. The possibility of the Presiding Officer having by mistake excluded the number of unused ballot papers from the total of ballot papers issued to the station, however, was not ruled out.

The first of the forms I picked at random, for purposes of the illustration, was in respect of: Polling Station Ngugo: Code 03; Parish Ngugo: Code 04; Sub-county Bugamba: Code 01; Constituency Rwampara: Code 03; District Mbarara: Code 27 — where the totals recorded are simply incomprehensible. The votes recorded for the four candidates were Awori and Kibirige 1 each; Bwengye and Karuhanga o each; Besigye 53 and Museveni 664 votes, which all add up to 719 votes. Yet, the totals were recorded on the form as follows:

Total number of:

• valid votes cast for all candidates 519

• rejected (invalid) ballot papers 13

• ballot papers counted 534

• spoilt ballot papers 2

• ballot papers issued to station 660

• unused ballot papers 120

I failed to find any permutation of the figures that would make sense. The form was annexed to Mary Francis Semambo’s affidavit, in support of the petition, and was produced as proof of stuffing ballot boxes. To my mind, it did not prove the theory.

The second form which I picked at random was an annex to the affidavit of Anteli Twahirwa also submitted among others as proof of stuffing. It is from Polling Station Murukoro II: Code 05; Parish Butare Code 14; Sub-county Muko Code 03; Constituency Rubanda County West: Code 048; District Kabale Code 09. The candidates’ votes were recorded as: 50 for Besigye; 465 for Museveni; and 0 each for the other four candidates, adding up to 515. The totals of ballots were recorded thus:

Total number of:

• valid votes cast for all candidates 515

• rejected (invalid) ballot papers 2

• ballot papers counted 560

• spoilt ballot papers 1

• ballot papers issued to polling station 518

• unused ballot papers 42

On this form although at first glance, the unused 42 ballot papers appeared to be extra, on further observation it becomes clear to me that the “unused” were included in the “counted” instead of being included among the “issued” If that error were rectified the counted ballot papers would be (515 + 2) 517 (not 560) and the issued would be 517+1 +42 (560) (not 518) and there would no extra.

It seems to me that the theory was adopted and put forward without any attempt at verification. The Commission had in use, forms on which the ballot papers issued to each polling station were accounted for, recording inter alia the serial numbers. If in the course of his analysis, Frank Mukunzi had taken off time to verify the figures recorded on the results forms with those recorded on the accountability forms, his analysis report would most probably have been different and carried more credibility. Similarly, my impression is that the Petitioner and the other witnesses who subscribed to the theory must have examined the results forms rather superficially. Be that as it may, I was not satisfied that the Declaration of Results Forms which were produced in Court amounted to evidence from which the Court could infer stuffing of ballot boxes or multiple voting at all.

There is, however, credible direct evidence from individuals who physically witnessed the multiple voting and ballot box stuffing. Let me highlight that evidence. Agatha Tunanukye a “poll watcher” of the “Christian Joint Monitoring group” was assigned to watch over a polling station at Kaasiro, in Kihiihi, Kanungu. She deponed, that she noticed that there were many cases of individuals allowed to come over and over again to cast votes, and others who were given several voter’s cards. According to her, the Petitioner’s agents were so intimidated by state security agents, that they were rendered useless and eventually left the station in protest. Stanley Bugando was the Petitioner’s campaign agent in Kihiihi. On polling day he went to the polling station at Bushere and helped at setting up the station. He deponed that while in the process of arranging the station which was at a school, he saw one Moses Mwesigye, with three others in a room plucking ballot papers from three books which Moses had just received from the Presiding Officer. He saw as they ticked them opposite the 1st Respondent’s name and cast them in a ballot box. This was prior to the official opening of the station at 7 a.m. Patrick Senyonga John, deponed that on polling day at 6.30 a.m. he witnessed the same type of stuffing of ballot papers in a ballot box, at a polling station at Lwebitakuli parish, in Sembabule District.

Matsiko wa Mucoori, a journalist with Monitor newspaper visited Kanyarugiri in Nyamarebe sub-county, Ibanda Sub-district, where there were two polling stations, 500 metres apart. One was for civilians and the other for soldiers. He was permitted by the Presiding Officer to observe the proceedings at the latter station for some time. He deponed that the Presiding Officer, Charles Musinguzi, was a soldier and teacher at the barracks. There were no polling agents except one for the 1st Respondent who stood near, and continued to observe voters as they ticked ballot papers in the basin. He noticed “the Battalion Intelligence Officer voting more than five times, changing his clothes each time he came to vote” He also noticed many other soldiers doing the same. At one point he pointed out two soldiers to the Presiding Officer who had voted before. They were stopped from voting again. Later he landed in trouble when the Presiding Officer became weary of his curiosity. Kedega Michael averred that he saw a group of about 50 soldiers led by Lt. Peter who had voter’s cards but whose names were not in the voters’ register voting at Alero polling station, and later saw the same group voting at Paara. Hangiro John who was polling agent for the Petitioner at Kabunga Primary School polling station in Rwekiniro, Ntungamo District, deponed that known individual supporters of the Respondent were given many ballot papers each to vote with, and the LC 3 Chairman was given an unknown number of unticked ballot papers which he took away.

Tukahebwa Keneth, the Petitioner’s polling agent at a polling station in Kyenzaza Trading Centre, in Bunyaruguru, Bushenyi District, deponed that he witnessed two attempts at stuffing the ballot box. First, one Ntare Banyenzaki Abdu, a driver of Watuwa Sikora alias “Mama Chama” was intercepted trying to push several papers into the ballot box. He was arrested by an armed home guard on duty at the station. However, within five minutes the said Watuwa Sikora arrived and took away her driver and the home guard. The latter returned later having been disarmed. Later he witnessed one Kyomuhangi Allen, sister-in-law of the same Watuwa Sikora, trying to do the same thing. She was also intercepted. Thirteen ballot papers already ticked in favour of the 1st Respondent were removed from her and eventually the matter was handed over to Bushenyi Police Station under ref: SD 39/12/3/2001 and CRB 107/2001. Basajabalaba Jafari, the Petitioner’s overseer for Bunyaruguru County, on receiving a report about the incident, went to the police. He deponed that he obtained photocopies of the 13 ballot papers removed from Kyomuhangi Allen. The photocopies were annexed to his affidavit. Watuwa Schola deponed that she had gone to that polling station when she learnt that her driver had had a scuffle with a vigilante (home guard) over the identity of the driver. She was with the LC III Chairman. They found that the armed vigilante was drank. The Chairman disarmed him. She denied the description of Kyomuhangi Allen as her sister-in-law. I was not impressed by her evidence. In any case she did not rebut the evidence of the intercepted ballot papers which was corroborated by production of copies of the ballot papers. Incidentally, going by the serial numbers on them, the 1 3 ballot papers must have come from the same book. I considered this evidence to be significant, not in proving the fact of the attempt to cheat, but because with it, the evidence of others who witnessed voters casting bundles of pre-ticked ballot papers, ceased to appear to be farfetched or fictitious.