COURT FILE NO.: CV-22-00690801 DATE: 20230329 ONTARIO SUPERIOR COURT OF JUSTICE
BETWEEN:
ROBERT PELLA Applicant – and – CORPORATION OF THE CITY OF TORONTO, TORONTO CATHOLIC DISTRICT SCHOOL BOARD, AND JOSEPH MARTINO Respondents
Counsel: Philip Hugh Horgan and Raphael Fernandes, for the Applicant Glenn K. L. Chu and Alex Redinger, for the Respondent Corporation of the City of Toronto Jonathan Lisus and Philip Underwood, for the Respondents Toronto Catholic District School Board and Joseph Martino
Heard: March 22, 2023
Dineen J.
[1] The Applicant ran for school trustee in the October, 2022 Toronto municipal election and was defeated by a single vote. He seeks an order under s. 58 of the Municipal Elections Act for a recount.
Factual Background
[2] The Applicant and the Respondent Mr. Martino were candidates for school trustee for the Respondent Toronto Catholic District School Board (“TCDSB”) in Ward 1 during the municipal election held on October 24, 2022.
[3] This election used composite ballots on which voters chose candidates for multiple offices. In Ward 1, voters were given ballots with candidates for three elections: mayor, city councillor, and trustee for one of four school boards.
[4] Ballots were counted using electronic tabulators, sometimes known as vote tabulating machines or VTMs. Voters would fill out their ballot and place it in a secrecy folder, and staff at the polling station would feed the ballot into the tabulator. The tabulator would scan the ballot and register any votes it detected. Voters were instructed to cast their votes by filling in an oval next to the name of the candidate they choose.
[5] Under certain circumstances, the tabulators were programmed to make a beeping noise after scanning a ballot rather than registering votes. These circumstances included: if the ballot contained an “overvote” with the tabulator detecting votes for more than one candidate for a single office; if no votes at all were detected; or, if the ballot was unreadable.
[6] In the case of an overvote, the City’s policy is that election staff should alert the voter to the fact that an overvote has been detected and offer a fresh ballot. If the voter declines the ballot then staff are to press a button directing the tabulator to register the vote for other races with no vote registered for the election with the overvote. There was one such vote in the Ward 1 TCDSB election.
[7] The policy is different for “undervotes.” Where the tabulator detects votes for some but not all elections on the ballot, it simply registers the votes that it does detect. Undervotes are very common in municipal elections and particularly for school trustee races where voters without school-age children may have little or no interest in the outcome of the election. In the election at issue, there were 258 Ward 1 ballots with the TCDSB candidates listed where no trustee vote was recorded.
[8] For certain types of votes, the voter is not present during the tabulation process. These include mail-in votes for voters who pre-registered to vote by mail, as well as votes collected at long-term care homes from voters with mobility restrictions whose votes are tabulated after-the-fact.
[9] For these types of votes, where an overvote, blank, or unreadable ballot is detected, an adjudication process takes place where senior elections staff attempt to determine whether the intent of the elector can be ascertained according to adjudication guidelines provided by the city. If so, a replacement ballot would be prepared that reflected the voter intent determined by the adjudicators. Candidates or scrutineers are permitted to be present for this process. There were a total of 92 votes adjudicated throughout the city’s 25 wards. The Applicant understandably did not have a scrutineer present for this process and it is unknown whether any of the adjudicated votes related to the Ward 1 TCDSB trustee election, though all ballots have been preserved.
[10] The final count for the Ward 1 TCDSB trustee election was released in the early hours of the morning of October 25, 2022. Mr. Martino received 2,148 votes to the Applicant’s 2,147. The Applicant’s disappointment at this result was increased by the fact that preliminary results before the counting of long-term care home votes had shown him ahead.
[11] The Applicant asked the TCDSB to pass a resolution ordering a recount of the election pursuant to s. 57 of the Municipal Elections Act. When this did not occur, he brought this application.
The statutory framework
[12] This election is governed by the Municipal Elections Act, 1996. The Act authorizes the use of vote-counting equipment including tabulators at s. 42(1)(a).
[13] While the governing legislation for federal and provincial elections mandates a recount where the winning margin is very close, section 56 of the MEA mandates a recount only in the case of a tied vote. It permits municipalities or local boards to implement policies that may require recounts in other circumstances but the City of Toronto has not done so.
[14] This application is brought under s. 58 of the Act, which provides:
58 (1) A person who is entitled to vote in an election and has reasonable grounds for believing the election results to be in doubt may apply to the Superior Court of Justice for an order that the clerk hold a recount.
(3) If satisfied that there are sufficient grounds for it, the court shall make an order requiring the clerk to hold a recount of the votes cast for all or specified candidates, on a by-law, or for all or specified answers to a question, and shall give the clerk a copy of the order as soon as possible.
[15] Section 60 sets out the manner in which a recount should be done:
60 (1) A recount under section 56, 57 or 58 shall be conducted in the same manner as the original count, whether manually or by vote-counting equipment.
(3) Despite subsection (1), if the judge who orders a recount under section 58 is of the opinion that the manner in which the original count was conducted caused or contributed to the doubtful result, he or she may, in the order, provide that the recount shall be held in a different manner and specify the manner.
The positions of the parties
The position of the Applicant
[16] Following the election, the Applicant sought information about the tabulators from the City and the manufacturer and raised a number of issues about the vote count. By the time the application was heard the issues had narrowed substantially. The Applicant rested his case for a recount on three categories of votes: the 258 undervotes, the single overvote, and the unknown number of adjudicated mail-in votes.
[17] The Applicant submits that it is possible that the undervotes include cases where the vote tabulators failed to register a faint or partially filled in vote. He cites the evidence of a scrutineer for a City Council candidate in Ward 1 who reported that his mother had initially filled out her ballot by making marks that went beyond the oval lines next to her preferred candidate resulting in the tabulator rejecting the vote and requiring a new ballot. The Applicant observes that if an undervote was recorded due to a similar error, the tabulators would accept the vote without alerting the voter to the problem.
[18] The Applicant relies heavily on the decision of Howden J. in Di Biase v. Vaughan (City), which criticized the reliance of the municipality in that case on the manufacturer’s standards for what will constitute an undervote, and which concluded that the programming of the computers likely resulted in a failure to count valid votes.
[19] With respect to the overvote, the Applicant notes that such votes can include situations where voter intention is clear, for instance where the voter visibly crosses out an initial vote for one candidate and then registers a vote for the other. He relies on prior decisions of this Court where recounts of overvotes have been ordered where their number equals or exceeds the winning margin including Chapman v. Kitchener, 2010 ONSC 7222 and Goldie v. Brock, 2011 ONSC 147.
[20] Finally, the Applicant argues that adjudicated ballots necessarily have some ambiguity to them and that the adjudication process took place without the presence of his scrutineers and no evidence has been led from anyone present.
The position of the Respondents TCDSB and Mr. Martino
[21] The Respondents argue that the Applicant has failed to meet the standard for a recount under s. 58 of the Act. They rely on the decision of Di Luca J. in Kett v. The Corporation of Scugog, 2019 ONSC 942, where he described the standard as follows at para. 6:
The Applicant must subjectively believe that there are grounds to doubt the election result and that belief must be objectively reasonable. The objective component of the test requires compelling and credible information which raises a reasonable probability that the election results are in doubt; see Lyras v. Heaps, 2008 ONCJ 524 at para. 25, R. v. Chehil, 2013 SCC 49, and R. v. MacKenzie, 2013 SCC 50. Speculation, suspicion and conjecture fall short of this standard. Conversely, the Applicant is not required to demonstrate a prima facie case that the election result is in doubt.
[22] The Respondents argue that the Applicant’s case is built entirely on speculation. The election was conducted properly in accordance with the City’s policies and the governing legislation. To order a recount based solely on the fact that the outcome was very close would fail to give effect to the clear legislative choice reflected in the Act to have recounts only for tie votes or cases where there is evidence of specific flaws in the voting or vote-counting process.
The position of the City of Toronto
[23] The City of Toronto provided helpful evidence and submissions about the conduct of the election but took no position on whether a recount should be ordered.
Analysis and conclusions
Are there reasonable grounds to doubt the result of the election?
Undervotes
[24] I agree with the Respondents that, on this record, the existence of 258 undervotes provides no objective reason to doubt the result. Such undervotes are common in municipal elections and particularly common for school trustee races. There is no suggestion that the proportion of undervotes in this election was unusual or unexpected. There is no evidence that any voters correctly filled out their ballot for the mayoral and city council elections but for some unexplained reason made only a faint mark when attempting to vote for school trustee.
[25] The decision in Di Biase, is in my view distinguishable. The record in that case included evidence from a representative of a candidate who was permitted to view photographic images of the ballots and who testified that it was apparent from those that some of the ballots recorded as undervotes contained valid votes. There is no comparable evidence here. I appreciate that the Applicant in this case has not had a similar opportunity to examine the ballots but the onus is nevertheless on him.
[26] I agree with the observations of Lauwers J., as he then was, in Goldie, that ordering a manual recount of tabulated undervotes “simply because they exist” without some specific evidence suggesting a problem in how they were recorded “would be inconsistent with the basic theory of the legislation authorizing the use of the VTMs.” See also Burton v. Town of Oakville, 69 O.R. (3d) 771 (S.C.J.)
[27] Given how common undervotes are in municipal elections, the policy of the City to allow tabulators to count them without an alert to the voter is sensible and helps promote efficiency and accessibility at the ballot box. The relevant provisions of the MEA are wholly inconsistent with a legislative intention that a manual recount take place in every election where the undervotes significantly exceed the margin of victory. This is a valid legislative choice that I must respect.
Overvotes and adjudicated ballots
[28] I reach a different conclusion with respect to the overvote and the unknown number of adjudicated ballots. I agree with the Applicant that in this election with a single-vote margin, he has objectively reasonable grounds to doubt the outcome given the uncounted overvote and the ballots requiring adjudication.
[29] In my view, the very existence of these votes provides an evidentiary basis for reasonable grounds to doubt the result. Unlike undervotes which are common, rational, and expected, overvotes and other ballots requiring adjudication are anomalous and very rare. In my view, a reasonable person would view it as likely that they reflect some problem with the tabulator’s ability to detect the intent of the voter.
[30] The City’s policy assumes that the voter who cast an overvote will have been given an opportunity to have filled out a new ballot and will have declined that opportunity, making the overvote a type of spoiled ballot. In Chapman, Flynn J. refused to rely on an equivalent policy without specific evidence that it was carried out in practice. Beyond the City’s guidelines, I have no evidence before me of what took place at the polling station where the overvote in this election was cast.
[31] It is possible that the policy was carried out as written and the voter in question marked his or her ballot for both candidates as some sort of protest or spoiled ballot. But it seems perfectly plausible that the vote contained some discernable intent to vote for one of the candidates and that the voter failed to cast a replacement ballot because the local election staff failed to alert the voter to the problem, or that the voter left without waiting for the ballot to be placed in the tabulator.
[32] As for the adjudicated ballots, they are by their nature votes requiring some application of professional judgment to assess the voter’s intention. It is reasonable in my view for someone in the position of the Applicant to believe that a review of the adjudicated ballots in the presence of the candidates could likely affect the result in a one-vote election. I would not fault the Applicant for not having scrutineers present for the adjudication given how unlikely it was that the race would be as close as it was such that the result might turn on the few, if any, adjudicated ballots.
[33] I appreciate the forceful submission of the Respondents that the MEA is not designed to mandate a “do-over” of the vote-counting and adjudication process without evidence of irregularities or errors in the way in which a municipality has conducted the process. I make no criticism of the City’s policies on how under- and overvotes are recorded and in almost all circumstances those policies will lead to an outcome that reasonable observers will accept as reliable. However, in my view, in a very rare case like this one with the closest possible margin of victory and the existence of unexplained anomalous votes equal to that margin, objective grounds to doubt the result may exist even where the conduct of the election may have been exemplary. The MEA in my view permits a recount in these circumstances to ensure the appearance of fairness and to avoid disenfranchising voters.
The conduct of the recount
[34] I accordingly order a very narrow recount. I order pursuant to s. 60(3) of the MEA that the uncounted overvote be counted by hand because the tabulator’s inability to assess that specific voter’s intention has contributed to the grounds to doubt the result in these unusual circumstances. Representatives of the two candidates will also be permitted to review the results of the adjudication process for any ballots identified by the City as having been subject to that process for the Ward 1 TCDSB election as set out in the formal order of the court.
[35] Success on this application has been divided and I would make no order of costs.
Dineen J. Released: March 29, 2023