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

Special polling stations:

A major complaint about the special polling stations for soldiers was that many of them were not published or disclosed until they were seen on the polling day or their particulars were found in the results. Counsel for the Petitioner referred to those “as sham polling stations.” Although strictly, this was not a specifically pleaded complaint, I found it to be sufficiently linked to the pleading in sub- paragraphs 3(1) (a) and (b) of the petition, that it was inescapable to consider the evidence, which showed that some of the special polling stations were not set up in compliance with law. The evidence was also in two categories, namely direct and indirect. The direct evidence was from witnesses who deponed that on polling day they found, inside or outside army barracks, voting in progress at polling stations which were not expected to be there. The indirect evidence is in the affidavit of Mukasa David Bulonge who obtained, from the Commission, tally sheets for the election results in respect of Kitgum, Gulu and Kamwenge Districts on which were included results from such undisclosed polling stations.

Altogether three documents, containing particulars of special polling stations for soldiers, were produced in evidence. The first was the Uganda Gazette of l9th February, 2001 read together with the amendments in the issue of 9th March. The Second was the ungazetted list of all polling stations distributed on 11th March. The third is an undated detailed list of all the special polling stations for soldiers. It does not appear to have been published or distributed. It came in evidence as an annexture to Chairman Kasujja’s supplementary affidavit. For ease of reference I will call that “the Chairman’s list”. In the said supplementary affidavit Chairman Kasujja explained that all army polling stations had been listed in the Gazette as “Outside Quarter Guard” but ultimately the number of soldiers had determined the ballot boxes (and therefore number of polling stations) to be used. That explanation would have been plausible if all the special polling stations were included in the list of 1lth March, 2001, since it is reasonable to assume that by that date the number of soldiers registered to vote was known to the Commission. But not all the polling stations set up for, and used by, the soldiers were included in that list, as was shown in the evidence which I proceed to summarise.

One complaint was about Upper Mbuya. It was by Ebulu Vicent, the Petitioner’s polling monitor for Mbuya Barracks who deponed that on polling day, he unexpectedly found seven polling stations inside the barracks and had to send for more polling agents to deal with what he called “this crisis situation.” Capt. Ondoga, the Division Political Commissar, while insisting that they were all outside the barracks, admitted that there were seven polling stations for Upper Mbuya. In the gazette list of polling stations for soldiers there had been listed under Mbuya Parish, one polling station “Outside Quarter Guard.” However in the list of 11th March, this polling station was not listed, though a total of 18 polling stations were listed for Mbuya I and Mbuya II parishes. It was only in the undisclosed Chairman’s list that seven polling stations for Upper Mbuya and four for Lower Mbuya appeared. It was not explained how such a large number of polling stations was omitted from the “final” list of polling stations. It appears to me therefore that they were not set up in compliance with the provisions of the relevant law and the principles laid down therein.

The second complaint was raised in the affidavit of James Oluka, the Petitioner’s polling agent at Akisim Barracks in Soroti Municipality. He deponed that there were supposed to be only two polling stations at those barracks identified as Akisim Barracks “A-D” and “E-Z”, outside the barracks. He was surprised to find two more inside the barracks, which were later brought outside also. Two officials, namely Omuge George William, the Returning Officer for Soroti, and Cpl. Oyo James, the Political Commissar, tried to explain what happened. They deponed in their respective affidavits that there were three polling stations, i.e. Akisim Barracks A-D, Akisim Barracks E-Z, and a third one which the former called “Akisim Barracks Outside Quarter Guard” and the latter called “Akisim (Outside Quarter Guard) barracks polling station.” What appeared in the Gazette and the Chairman’s list under Akisim Ward of Soroti Municipality was “Polling Station Cell II (Outside Quarter Guard).” In the list of 11th March only two polling stations were listed and identified as “Akisim Barracks A-D” and “Akisim Barracks E-Z.” The alleged third polling station was not listed in any of the three documents. As for the fourth polling station the two officials explained that an additional ballot box had been brought to Soroti barracks for use by soldiers recently transferred from Olilim barracks, in Katakwi District, It formed the fourth polling station at Akisim barracks. However I failed to trace any polling station for “Olilim Barracks” in any of the three documents. The only polling station bearing the name “Olilim” and which was included in the ordinary list gazetted on 22 December 2000, and in the list of 11th March 2001, was Olilim Primary School, in Usuk County, Ngarian Sub-county. Both the gazetted special polling stations and the Chairman’ list include only two polling stations for soldiers in Katakwi i.e. Oburatum and Okuliak. My conclusion on all that evidence was that the two extra stations at Akisim Barracks were not located lawfully.
In his supplementary affidavit Mukasa David Bulonge made two points. The first was that he had found in the tally sheets in respect of Kitgum and Gulu Districts, six and eight polling stations respectively, which were neither gazetted nor included in the list of 11th March. He did not, however, disclose what he found, if any, from the tally sheets in respect of the third district, Kamwenge. Of the six he named in Kitgum tally sheets, I found that at least three, i.e. Ngomoromo “A-E”, “F-N” and “O-Z” in Lamwo County, Lokung Sub-county, Pawor Parish, were on the undisclosed Chairman’s list, leaving three appearing in the tally sheets only. Then of the eight in the Gulu tally sheets, it appeared that he miss-spelt the name of one, because I could not trace it in the tally sheets.

Five of them, i.e. Kasubi “A-A”, “B-L”, “0-0”, “114-N’ and “P-Z” appeared to be within the description of what appeared on the Chairman’s list and on the list of 11th March. In the Gazette there appeared under Gulu Municipality, Sub-county Bar-Dege, Parish Gulu Barracks, “Polling Station Kasubi (Outside Quarter Guard)” The list of 11th March showed twelve polling stations under Gulu Barracks Parish. Code Nos.01 and 02 were named Airfield I and II respectively. Code Nos.03 to 12 were named Gulu Barracks and distinguished with letters “A (A-L)” up to “C (0-Z)” The Chairman’s list on the other hand listed only five polling stations as they appeared in the tally sheets. The remaining two were recorded in the tally sheets as “Bibia Outside Quarter Guard B-N and O-Z” with a third one recorded separately without the letters. All three however were under Bibia Parish, Atiak Sub-county, Kilak County. In the Gazette there appeared only “polling station: Bibia Outside Quarter Guard,” and in the list of 11th March simply “Bibia Barracks” The Chairman’s list however has all three named “Bibia Outside Quarter Guard” and identified as “A-A “, “B-N” and “O-Z” My conclusion from this evidence was also that there were set up and used (at least three in Kitgum and two in Gulu) special polling stations for soldiers which were not disclosed prior to polling.

I am constrained to observe that it remains an unexplained puzzle to me, how those polling stations, 2 in Soroti, 3 in Kitgum, and 2 in Gulu obtained polling materials when they were not reflected even on the Chairman’s list.

The second point Mukasa Bulonge made with a thinly veiled innuendo was that he observed in those polling stations he named, that “the 1st Respondent got results that sharply contrast with the pattern of results got from polling stations that were gazetted and/or in the list submitted on 11th March 2001.”

Learned Counsel for the Petitioner submitted that out of the so-called sham polling stations arose ballot stuffing, chasing of agents, multiple, under-age and ghost voting, and falsifying of results. Needless to say, with due respect to counsel, that his conclusion was speculation based on suspicion. A court of law does not base a decision on speculation or suspicion. However, the suspicion and speculation were understandable in circumstances where previously undisclosed polling stations were used for soldiers. That inevitably undermined the principle of transparency.

There was also an anomaly regarding separate registration of soldiers as voters. Chairman Kasujja, explaining a discrepancy in numbers of registered voters deponed in his supplementary affidavit in reply:

“28. That after the referendum of June 2000 the register on cleaning had about 9,308,173 voters, after the update referred to above the number rose to 11,093,948 voters. After display and clean up, the number reduced to 10,672,389. This however did not include soldiers and adults living with them and when they were included the number rose by 103,447 to 10,775,836.”

29. …………………………….

30. ………………………….

31. That the number of registered voters rose from 10,672,389 to 10,775,836 due to inclusion of voter soldiers who had previously been registered separately to vote within the barracks and as a result the number of polling stations rose from 17,136 to 17, 308”

It is apparent that the 2nd Respondent compiled a separate “register” or “roll” for soldiers. This, in my view is another area where the 2nd Respondent misconstrued the law and principles underlying it. Section 38 of the Act empowers the Commission to make special provision “for the taking of votes” of the persons specified therein, including soldiers in restricted areas. The section reads:

“38. The Commission may make special provision for the taking of votes of patients in hospitals or persons admitted in sanatoria or homes for the aged and similar institutions and also for persons in restricted areas such as soldiers and other security personnel; but the Commission shall publish in the Gazette a list of the restricted areas under this section.”

While it is not necessary for the purposes of this petition to conclusively interpret the application of this provision, largely because the point was not canvassed, I am still constrained to observe that by virtue of the ejusidem generis rule, “restricted areas” should mean areas where the occupants, like patients in hospitals, and the aged in senatoria, are for some reason, like military operations, not able to come out of the areas readily, for voting with the rest of the community. It does not appear to me that barracks fall within that description. Be that as it may, while the provision clearly envisages the creation of separate polling stations, it does not envisage the setting up of a separate voters’ register as appears to have been done. Inevitably each polling station would have its voters’ roll containing only names of voters in the restricted area served by that polling station. However, by virtue of S. 18 of the Commission Act, each such voters’ roll forms part of the voters’ roll for the Constituency in which the restricted area is situated, and through that becomes an integral part of the national voters’ register. In my view having a parallel voters’ register for soldiers, which was not integrated in the national voters register is incompatible with the law, and the principle underlying voter registration, and transparency.

Furthermore there was some other evidence which tended to portray the electoral process for the army as parallel to, rather than integrated into, the national electoral process For example in his affidavit Cpl. Oyo, the Political Commissar at Akisim Barracks, deponed that he went through the Army Headquarters at Bombo, to call for the names of the soldiers from Olilim together with a ballot box to be transferred to Soroti. That gave the impression that there was a role in the electoral process for the Political Commissar and the Army Headquarters which is not apparent in the law. That would probably have passed off as an innocuous administrative procedure if it did not lead to unexplained polling stations unlawfully set up. Even more glaring is the evidence of Major Nuwagaba John of Bombo Army Barracks. He deponed that at the time of the Presidential elections he was a Captain in UPDF and was “in charge of the electoral process in the army that is organising conducting and administering the presidential elections.” He confirmed the evidence in the affidavit of one Ongee Marino that on 1 2tI March he flew to Kitgum by helicopter to deliver three ballot boxes containing electoral materials for three Ngom Oromo polling stations for soldiers. He arrived at 7.30 p.m. and held a meeting with the Returning Officer, the District Registrar and Brigade Political Commissar “concerning the voting of soldiers in Ngom Oromo” and thereafter flew back leaving the boxes with the District Registrar. Ongee Marino, the Petitioner’s monitor for Kitgum deponed that he objected to the three additional ballot boxes being taken to Ngom Oromo and they were taken to the Police Station at about 10 p.m. However the Returning Officer refused his request to open them so as to ascertain the contents.

Further evidence on parallel treatment of the special stations is in regard to the policing. Although Chairman Kasujja insisted in his supplementary affidavit in reply that on polling day, the police and election constables appointed by the Commission, were in charge at polling stations, while the Army was in charge of national security generally, in a circular letter to Returning Officers dated 22 February he instructed them to liaise with local Commanders/Political Commissars for the latter to identify army personnel who can act as Election Constables and additional Polling Assistants. This was corroborated by John Kisembo the Inspector — General of Police who after explaining the command structure of the Joint Security Task Force in his affidavit, deponed:

“11 policing of the polling stations and tallying centres during the electoral process was only under the Uganda Police save for the army barracks for which the Electoral Commission had made other arrangement” (emphasis is added)

With the foregoing in mind, I am constrained to reiterate my earlier observation that the concept of “taking of the votes of persons in restricted areas” appears to have been misconstrued. In my opinion, the Commission ought to re-consider its application of that provision if the possible abuse of the process and the general negative impact on transparency are to be avoided.

For the reasons I have endeavoured to articulate, I answered issue No.2 in a qualified negative. I was satisfied that while to a large measure the Presidential election was conducted in compliance with the principles laid down in the Act, there was non-compliance with some of the principles as I have indicated above and as summarised in the judgment of the Court.

ISSUE NO. 3

The third framed issue was:

“Whether, if the first and second issues are answered in the affirmative, such non-compliance with the provisions and principles of the said Act, affected the result of the election in a substantial manner.”

My answer to this issue was that the non-compliance with provisions and principles of the Act, found to have occurred, did not affect the result of the Presidential election in a substantial manner. Before I discuss my reasons for that conclusion, I will summarise counsel’s arguments submission the issue. The argument mainly centred on the test to be applied in determining the issue.

For the Petitioner, two counsels addressed the court. First, Mr. Mbabazi submitted that in assessing if non-compliance affected the result of the election the Court had to take into account the nature of the non-compliance. He contended that there were two categories of non-compliance; namely, non-compliance which goes to the root of the Constitution, and non-compliance with provisions of the Act only. He submitted that noncompliance which goes to the root of the Constitution IS a substantial noncompliance and must be deemed to have affected the result of the election in a substantial manner. He included in that category, non-compliance with the principle of voter registration which he submitted was so substantially violated that in the end there was no final national voters’ register. In the alternative, he submitted that the sum total of all the proved irregularities was so substantial that it affected the result of the election in a substantial manner. His submissions were augumented by Mr. Walubiri who, in reply to submissions by the Respondents’ counsel submitted that determination of the issue involved value judgment, whereby the test is qualitative rather than quantitative. He argued that noncompliance with principles cannot be quantified in terms of numbers in the manner counsel for the Respondents demanded. He stressed that the essential principle to be adhered to in a democratic society is that elections shall be free and fair. If, therefore, in an election, non-compliance goes to the root of that principle, the court must annul the result. He invited the Court to follow the approach taken by the Court of Appeal of Tanzania, in its decision in ATTORNEY GENERAL AND OTHERS vs. KABOUROU (1995) 2 LRC 757, rather than the precedents relied on by counsel for the Respondents. He specifically urged the Court not to follow the approach taken by the High Court of Tanzania in MBOWE vs. ELIUFOO (1967) EA 240, on the ground that it is not consonant with the democratic setting and values Uganda chose and entrenched in the new order, set out in the 1995 Constitution.

On the other hand, Dr. Khaminwa for the 1st Respondent premised his submission on the elementary proposition that, a Petitioner who comes to court seeking annulment of an election, on the ground of noncompliance with provisions of, and principles laid down in the Act, must prove that such noncompliance affected the result of the election in a substantial manner. It is not sufficient to prove that there was non-compliance, or even to prove that the non-compliance was substantial. For the Petitioner to succeed, it must be proved that the noncompliance had substantial effect on the result. He submitted that the standard of proof in election petitions generally is very high, and that it must be even higher in a petition seeking to annul a Presidential election. He argued that an election result involves figures and numbers, and contended that, therefore, in order to prove effect of non-compliance on the election result, it was inevitable to show that it affected numbers of voters or votes. Learned counsel submitted that, in the instant case, it was not proved that a substantial proportion of the electorate was prevented from voting freely. On the contrary, according to him, the high voter turn-up of 70.3% of the registered voters, and the high proportion of 69.3% and 27.8% of valid votes cast for the 1st Respondent and the Petitioner, respectively, were significant indicators that the election was free and fair and that the electorate was free to vote according to their will. Counsel pointed out the colossal margin between votes cast for the two candidates, and argued that it was not proved that the non-compliance effected that colossal margin. He submitted that far from proving that effect, the Petitioner had pleaded uncertainty and adduced inconclusive evidence. He pointed to sub-paragraph 3(i) (y) (vii) of the petition, where it was pleaded that: “It cannot positively be ascertained that the 1st Respondent obtained more than 50% of valid votes of those entitled to vote”; and to the evidence, of Frank Mukunzi who claimed to have analysed data related to the results of the election, and concluded that “from the data that was availed, it was not possible to determine to what extent the (above) errors affected each candidate.”

For the 2nd Respondent the learned Solicitor-General associated himself with counsel for the 1st Respondent and his submissions on the burden and standard of proof in regard to this issue. He submitted that isolated incidents of noncompliance with the principles cannot have had any substantial effect on the result of the election. He reiterated that the Petitioner had failed to prove that such non-compliance as was proved, had affected the result in a substantial manner.

Section 58(6) provides:

“(6) The election of a candidate as President shall only be annulled on any of the following grounds if proved to the satisfaction of the court-

(a) non-compliance with provisions of this Act if the court is satisfied that the election was not conducted in accordance with the principles laid down in those provisions and that the non-compliance affected the result of the election in a substantial manner

(b) that the candidate was at the time of his or her election not qualified or was disqualified for election as President;

(c) that an illegal practice or any other offence under this Act was committed in connection with the election by the candidate personally or with his or her knowledge and consent or approval.” (Emphasis is added).

Issue No.3 in this petition relates to the application of paragraph (a) of that subsection. It is centred on the meaning of the phrase “affected the result of the election in a substantial manner.” The result of an election may be perceived in two senses. On the one hand, it may be perceived in the sense that one candidate has won, and the other contesting candidates have lost the election. In that sense, if it is said that a stated factor affected the result, it implies that the declared winner would not have won but for that stated factor; and vice versa. On the other hand, the result of an election may be perceived in the sense of what votes each candidate obtained. In that sense to say that a given factor affected the result implies that the votes obtained by each candidate would have been different if that factor had not occurred or existed. In the later perception unlike in the former, degrees of effect, such as insignificant or substantial, have practical meaning. To my understanding, therefore, the expression “non-compliance affected the result of the election in a substantial manner” as used in S.58(6) (a), can only mean that the votes candidate obtained would have been different in a substantial manner, if it were not for the noncompliance substantially. That means that, to succeed, the Petitioner does not have to prove that the declared candidate would have lost. It is sufficient to prove that his winning majority would have been reduced. Such reduction however would have to be such as would have put the victory in doubt. This is the view the learned Chief Justice of Tanzania, Georges C.J., stated differently in MBOWE vs. ELIUFOO (supra) when he said at p.242 D-E.

“