The 1,289 Votes That Decided the Election

Mark Dion, City Councilor and former State Legislator, was announced this morning as the winner of Portland’s 2023 mayoral election and is now mayor-elect. In the final round of votes, he edged out 2^nd place candidate Andrew Zarro by just 643 votes, or just a few percentage points, to seize the prize.

More thorough analysis of the election and its victor will be published in the Portland Townsman over the next several days.

In the preliminary vote counts that have been released, however, a substantial statistical outlier has been identified which played a crucial role in the election of Mark Dion that – had it been more typical – may have resulted in a different outcome.

Ranked-Choice Voting

As background, Portland uses a form of ranked-choice voting called Instant Runoff Voting (IRV) to elect its mayors. What this means is that on the mayoral ballot voters have the opportunity to rank as many of the candidates as they like, choosing their top choice, their 2^nd choice, their 3^rd choice, and so on. Any candidates they do not rank are considered to be equally desirable in terms of the election system.

When votes are tabulated, the City Clerk’s office determines how many of the “top choice” votes each candidate received. If there is a majority of over 50% of votes, the process stops here and a winner is declared. If there is no majority, then the candidate with the least votes is eliminated from the running and all of the ballots which had selected that candidate as their 1^st choice are re-allocated to their 2^nd choice. At this time, once again, the Clerk’s office determines how many ballots each still-running candidate received and checks to see if there is a majority. If not, then the candidate with the next-least votes is eliminated and their ballots re-allocated to their 2^nd and 3^rd choices (skipping over eliminated candidates,) and so on. This process is repeated for as many times as is necessary to establish a majority.

More information about IRV can be found in this Townsman guide to 2022’s election reforms.

An important facet of IRV, however, is that voters do not have to rank all candidates. For example, in this election there were 11 people who wrote in the name of George Rheault as their first choice and then left the rest of the ballot blank. Rheault, though receiving a notable number of votes for a write-in candidate, was eliminated before any of the five on-ballot candidates. So, what happens to their votes? In the event of a ballot containing no further valid choices, in that all the non-eliminated candidates have been left unranked when the ballot is to be re-allocated, the ballot is simply removed from the election. It will no longer be considered in calculating what constitutes a majority of votes.

For this reason, while an individual voter may have any number of personal motivations for not ranking every candidate, the voting system “sees” this as a voter having no preference among the remaining candidates and thus irrelevant.

The 1,289 Votes

Ahead of the election on Tuesday, many voters and pundits had conceptualized the mayoral race as a landscape on which the candidates stood, some closer to one another than others. Mark Dion, with his emphasis on public order and cost-effective government, was perceived to stand to one side, while Pious Ali and even more so Dylan Pugh, both running on more communitarian agendas, were perceived to stand on an opposite side. Andrew Zarro and Justin Costa were typically perceived as somewhere in the middle.

For this reason, though Mark Dion led the pack in the first ballot, he was not considered a shoe-in to the win the whole thing. He was still well short of a majority, with only ~36% of the vote. His nearest competitor, Andrew Zarro, had ~27% of the vote, and 3^rd place Pious Ali had ~23%. The other candidates had far too few votes to win.

If, it was thought, Ali’s voters perceived Zarro as “closer” to them (ideologically) than Dion, then it would make sense for Ali’s voters to rank Zarro above Dion. With the large chunk of votes from Ali, plus sundry votes picked up from the minor candidates, Mr. Zarro seemed to have a path to victory in successive rounds.

This theory was basically true, with Ali’s voters ranking Zarro above Dion at over twice the rate as they ranked Dion over Zarro… but there was an unexpected hitch.

When Rheault was eliminated from the race, just 10% of his ballots were removed from the election for having no valid ranked choice. When Pugh was eliminated from the race, just 9% of his ballots were so removed. When Costa was eliminated from the race, again, just 10% were removed. In total, only 268 ballots were removed from the election this way, at a consistent rate of around 10% per elimination.

But after Ali was eliminated, a solid 22.2% of his ballots were removed from the election. 1,289 votes.

This is a surprisingly high number of ballots to be “removed from circulation,” and these 1,289 votes which could have gone to either candidate instead, for all intents and purposes, blinked out of existence.

If these votes had been split between Zarro and Dion at the same rate as the rest of Ali’s re-allocated ballots, the difference between the two finalists would’ve been whittled down to only about 100 votes… still a Dion victory but a much, much narrower one. But there’s reason to believe these voters may have ultimately favored Zarro over Dion, had they ranked their choices. This is because the high abstention rate can probably be attributed to two groups. First, Portland’s progressive left; as an example, DSA chapter co-chair Wes Pelletier indicated that he was intentionally leaving Zarro off of his ballot, and this writer can confirm several other progressive voters who did the same. And second, voters with barriers to understanding the ranked-choice system, especially those in immigrant communities or other groups with low English proficiency or education. Such groups are noted as struggling with non-First-Past-the-Post systems, at times.

Correction 11/8/2023: A previous version of this article incorrectly described Wes Pelletier’s role in the DSA. Thank you to Wes for the correction!

The Difference?

Why did Mr. Zarro fail to appeal to these voters, who might have been expected to favor him over his competitor, Mr. Dion? It could have been any number of reasons. Local politics is notoriously non-ideological, and this anonymous voter for example claims that Zarro’s business behavior is what motivated them to leave the Councilor off of his ballot. Mr. Zarro’s campaign was also among the most policy-oriented of any candidate’s, so perhaps one or more of the policies which he promised to pursue alienated some voters. In a city in which the votes of religious groups is mattering more than it has in decades due to trends in immigration, perhaps the Councilor’s past controversies about faith contributed to the dissension. It’s impossible to know for sure.

No candidate is owed votes by anyone, it is interesting however that when the choice came down to Andrew Zarro and Mark Dion, a substantial number of voters decided that they were equally displeased with either result. The two men’s agendas and platforms, though not as diametrically opposed as they are to other candidates’, are quite different, and one wonders whether those who left Mr. Zarro off their ballot will come to regret this decision.

In any case, these hypothetical scenarios are just that. Mark Dion ran a strong campaign and focused on issues important to Portlanders, and the intrigues of this slice of the electorate does not take away from his enormous success.

With his new mandate, what can we expect our new mayor to do? Keep an eye on the Portland Townsman for further analysis and reporting on this period of transition.

Ashley D. Keenan – Ashley is an editor of The Portland Townsman, writer, local small business-owner, and originally a Downeast Mainer. Her work primarily covers the mechanics of local government, the ongoing housing crisis, responsible market economics, and New England culture and history. She lives in Portland with her fiancé and can be personally reached at ashley@donnellykeenan.com.