An Chúirt Uachtarach

The Supreme Court

Between

1. There are three issues on this election petition: firstly, is a choice valid that begins beyond the number ‘1’ or a mark which in its context indicates a clear first preference by being followed by ‘2’ and ‘3’ etc; secondly, are ballot papers rejected by a returning officer before the count begins as invalid, and thus not counted as part of the total valid poll, ever to be included in a recount; and, thirdly, has the burden of proof been met by the petitioner to enable a remedy on this appeal should error in the counting of votes have been demonstrated? That remedy is considered by the majority as a total recount.

2. Proof is the most important issue as it is central to the remedy sought, which is to upset the result of this election. It is the point of this dissent. Arguably, this petition would not have been necessary had the questioned ballot papers been discovered to, and then checked by, the petitioner. Each of these were both stamped and marked with a green pen by the returning officer so that the numbers ‘3’, ‘4’, ‘5’, etc were replaced by the numbers ‘1’, ‘2’, ‘3’ etc. Of itself, this illustrates the error which appears to have been collective to a number of returning officers in consequence of prior experience where local and European Parliament were run together: a not insubstantial number of voters thought that by starting on the European ballot and expressing a preference in numerical sequence starting with ‘one’ or ‘1’, that they had to continue through that numerical choice on the local election ballot resulting in their first preference becoming a ‘3’ or whatever, and then being continued sequentially. The difference between the last 3 candidates being less than 2 votes and 5 votes, on the proportional representation system, it is considered by the majority that the error of including ballot papers where preferences started numerically with other than a ‘1’ is, in the words of section 5 of the Local Elections (Petitions and Disqualifications) Act 1974, “likely to have affected the result of the election”.

3. While it is thus clear that the majority view on this election petition is based on that statutory imperative, this dissent is predicated on the basis that the section requires that relief is only available under the Act of 1974 where affect on the result of the election is “established”. Every plaintiff in every ordinary contract or tort case has to demonstrate the probability of the case they make: in the absence of establishing that their evidence is more likely to be correct than any countervailing evidence or possibility, the plaintiff fails. An electoral case should not be different. There is also the question of the means of proof. In an electoral case, with physical ballot papers which are preserved, the ones said by a petitioner to be wrongly counted and leading to an incorrect result can be physically taken out and scrutinised. That is the advantage the paper ballot system has over electronic voting machines: the evidence is there. In this case, by the simple expedient of identifying and taking out for judicial scrutiny the 57 ballots both stamped and corrected by the returning officer in a green biro pen and calculating the effect of any the error of their inclusion on the result, whether the result would have been different or not thereby could have been established. That was not done by the petitioner. That is what has happened on prior election petitions. There is thus no proof whether the now to be discarded 57 ballot papers favoured Dan Kiely. Elections are the expression of the people’s reposing of trust in representatives, chosen under conditions of high scrutiny and subject to legal form. There should be no questioning of an election result unless the test as set out in the Act of 1974 is met as a probability by any petitioner, whereby not just error in the running of an election is demonstrated but also a probable affect on the result of that election is also established. What the legislation requires is that the petitioner demonstrates as likely, not merely possible, that an impermissible method has produced a false outcome. That has not happened here.

4. In the light of the precise wording of the legislation, it is to be doubted if the citation of the prior cases on electoral petitions is of assistance. Of most relevance is Boyle v Allen [1979] ILRM 281 at 330, where Sheridan J pointed out that the onus of proof is on the petitioner “to show, by evidence, that (a) an incorrect adjudication took place and (b) that such an adjudication … was … likely to affected the outcome of the election.” The test is not whether the result could or would have been affected. Were that so, the petitioner must succeed simply because the numbers are so close. But, it is not the test. Mathematical proof of probable effect is possible by the extraction and counting of the 57 ballots clearly marked by the returning officer. In that context, nothing less than a demonstration of why these inclusions go either against the petitioner, or in his favour, and do not cancel each other out, is required by law. Proof on the balance of probabilities is required. The petitioner is required to establish a false result. As the relevant excluded ballots are capable of identification, mathematical calculation is not only feasible but is open. It is to be noted that the helpful submissions of the Minister for the Environment, Community and Local Government on this appeal include a reference to a 1999 petition, Larry Kavanagh v Laois County Council, noted in an article by Louis Brennan in Counts, Recount and Petition: Laois County Council Election 1999, published by the Institute of Public Administration (2000), where Judge Moran considered 21 different disputed ballot papers in issuing his decision. That work has been made available to the Court and it includes a facsimile of the disputed ballot papers. Similarly, the Minister’s submissions note the case before Hamilton P on the petition challenging European election, In the Matter of the European Parliament Elections for the Constituency of Leinster 1989 (The Petition of Michael Bell), in November 1989, where the judge examined the relevant ballot papers and established to the satisfaction of all sides that even if these ballot papers had been included, the petitioner would not have been elected either way; referenced in Whelan - Politics, Elections and the Law (Dublin, 2000). The transcript of that decision of 24 November 1989 has been furnished. Hamilton P considered all the relevant ballot papers, it appears with assistance from an expert, and did a count which established that the election result would have been the same. There was no reason not to do the same on this election petition. Here, as there were 57 ballot papers to be excluded, it is not beyond proof to show the probable effect of that on the outcome of this election. These precedents, while not binding this Court, indicate that only by producing the disputed ballots to a judge can a candidate on an election petition “establish” that the result of the election would have been different.

5. The figures recorded as primary fact by the trial judge, His Honour Judge Moran of the South Western Circuit, are that there were 230 doubtful ballots. The returning officer rejected 173 of these and admitted 57 as valid. Turning shortly to whether a recount involves a reconsideration of all 230 ballot papers, it is the 57 admitted that are here of importance. The transcript of the hearing elucidates that the extraction and calculation of the effect of these renders it possible to demonstrate the probable outcome of their exclusion from the recount. Those challenging an election result must do that. On 18 June 2014, the second day of the hearing before Judge Moran, the deputy returning officer Michael McMahon, responsible for the Listowel electoral area, explained how the issues concerning ballot papers with a numerical sequence beginning later than a ‘1’ were dealt with by him. Any ballot paper which was rejected was stamped ‘rejected’ and was put aside. This did not form part of the count and, consequently, those ballot papers were never included in the bundles upon which the proportional representation system is based. As to an admitted ballot, where there had been a dispute or a potential dispute, this was stamped ‘valid’ and then subject to the procedure as explained by the witness thus:

In some cases where I made it clear, without interfering with a mark that was on the ballot paper, I made it clear what my ruling was in either in admitting it or in terms of future counts because I didn’t want a situation where a vote might be disputed or that the staff wouldn’t be clear what my decision was … Beside the three, four and five in a green biro I wrote one, two, three … to give a clear indication – these papers are then going – being given back to the counters who had already raised them as doubtful papers and I’m giving a clear indication to those counters of what I’m allowing.

6. The trial judge, dealing with this issue, although holding against the petitioner on the issue by deciding that a recount cannot include a re-examination of doubtful ballots, and in holding that the returning officer did not have a duty to ensure that the petitioner was present at any adjudication of doubtful ballots, and holding that it was valid that later numerical sequence beyond one commencing with ‘1’ allowed such votes as valid, also held as follows:

First, Mr Patrick Enright, solicitor and election agent for Mr Kennelly, one of the candidates, testified that he was at the adjudication of doubtful ballots, that very little was said by any of the persons present, but he, Mr Enright made the most representations, and that very little, if anything, which he said influenced the returning officer to change his mind. Mr Enright’s recollection is that there was just one doubtful ballot which aroused any real interest. Secondly, the petitioner, on whom is the burden of proof, has not produced any evidence to contradict the said evidence of Mr Enright. Thirdly, there was no controversy about the adjudication of doubtful ballots and there is no reason to conclude that the returning officer would have, or should have, altered any decision by reason of the presence of the petitioner or by reason of any representation made by the petitioner or anyone on his behalf.

8. On the main issue of ‘3’, ‘4’, ‘5’ etc becoming ‘1’, ‘2’, ‘3’ and so on, it is difficult not to feel sympathy for the returning officer. Experience in prior elections had shown two distinctive approaches taken by the Irish electorate. One was a form of expression which was by demonstrating a first preference with a tick or an ‘X’ and continuing with the next preferences as ‘2’, ‘3’, ‘4’. That process demonstrates a clear choice and one which, on any reasonable analysis, is unambiguous. There is nothing to indicate that the admission of such ballots falls foul of the relevant legislation. The other, however, demonstrates a dangerous ambiguity. Certainly, it is the case that the evidence before Judge Moran was that where a European election was coupled with a local election, a not insignificant group of people continued on their preferences from one ballot paper and into the other as if the choices were amalgamated. Where people had chosen ‘1’, ‘2’, ‘3’, ‘4’ etc on the European ballot they expressed their first and following preferences on the local ballot paper by a ‘5’, ‘6’, ‘7’ and so on. This resulted in the flaw demonstrated in the 57 local election papers in question here.

9. While returning officers are understandably keen not to disenfranchise people, as Clarke J states in his judgment, the problem is in the ambiguity of apparent choice. Perhaps some people decide who is the candidate whom they like the least and then work backwards. Possibly others decide that there are certain candidates they definitely will not vote for; consequently if they start with their last preference it will not necessarily be the number ‘15’ that marks that choice even though there are in fact 15 names on the ballot paper. How does one know that a person has not promised a third or fourth choice to a particular candidate and decides to mark a ‘3’ or ‘4’ beside their name and then go back to the others, but forgets to fill in their ‘1’ and ‘2’? These, certainly, are speculations and admittedly unlikely ones. But statistically unlikely things happen within a sample of thousands. If the discretion of returning officers was unlimited, and if the legislative rule allowed for any form of choice by a voter no matter how expressed, provided it was sequential, the decision made by the returning officer in this case would have been possible.

10. The problem with such an unfettered discretion is that it makes matters uncertain. Even where that not so, it is beside the point. The legislation governs this situation. The Local Elections Regulations 1995 make it clear that no ballot paper is valid unless the preference begins with “the figure 1 standing alone, or the word “one” or any other mark which, in the opinion of the returning officer, clearly indicates a first preference …” Thus, an “X” or a "✓ " or “an duine seo” or “this person” or some other indication followed by a ‘2’, ‘3’ and so on is possible as the expression of a valid choice under the Regulations. What that naturally says in that particular context is “this is the woman I want to vote for and here are my second and third preferences.” Should such a single mark be followed not by a ‘2’ but by a ‘3’ or any lower number, then there is a problem. Similarly, commencing a numerical sequence at any random number beyond ‘1’ cannot be within the legislation. Since Article 80(1) gives a definition of a first preference as “any mark which, in the opinion of the returning officer, clearly indicates a first preference”; a clear choice of a candidate in unambiguous terms as the one most preferred for election is not invalidated by a mark on its own or a mark followed sequentially by ‘2’ and ‘3’ etc. By requiring, however, either a ‘1’ or a ‘one’ or a mark which in its context states that this is the start of a sequence, perhaps ‘a’, ‘b’, ‘c’, the Oireachtas has excluded numerical sequences randomly beginning beyond an unambiguous first choice. Regrettably, therefore, since the mistake amounts to the elector disenfranchising themselves in error, the 57 ballot papers having that defect were not within the range of discretion open to the returning officer for inclusion in the vote.

11. Insofar as confusion has arisen in the past where two elections to different representative bodies are run together, and may be predicted as likely to return in the future, clearly different colours as to ballot papers, an education campaign through the media, or a simple instruction and giving out each ballot paper by the returning officer would surely correct the error. A suggested wording might be: “Now mark this one here ‘1’, ‘2’, ‘3’ and so on in order of your choice and this other ballot paper you do exactly the same thing marking it ‘1’, ‘2’, ‘3’ in order of your choice: they are separate.”

12. Much of the evidence before Judge Moran was taken up with the issue of the exclusion of the 173 rejected ballot papers and as to why, when a recount might be called, these were not also subject to reassessment for possible inclusion in the bundles. The transcript makes it clear that these rejected ballots were ruled out from any form of counting in accordance with the legislation. The time to raise any query in relation to that process was in the early stages of scrutiny. The duty is on the candidate to be present, as opposed to placing an impossible burden on the returning officer to require their attendance. Article 78 of the Regulations makes it clear that it is the duty of the returning officer to scrutinise the ballot papers “with the purpose of discovering any papers liable to be rejected as invalid”. The invalidity may arise from, for instance: a voter not voting at all; or registering a protest through the ballot paper without also filling in a preference; or deciding to identify themselves by, for instance, signing their choice or otherwise writing some identifying matter; or, by mistake or mischief, spoiling their vote. Once rejected, such ballots are not thereafter made part of the count. The count is defined in Article 80(1) of the Regulations as:

… all of the operations involved in counting the first preferences for candidates; all the operations involved in the transfer of the surplus of candidate deemed to be elected; all operations involved in the transfer of the votes of an excluded candidate; or all operations involved in the transfer of the votes of two or more candidates excluded together.

14. In the result, while an error has been demonstrated, the burden of proof has not been met by the petitioner. In consequence, the petitioner is not entitled on this analysis to a remedy on this appeal.