With Day 2 mayoral ballot retabulations still to go, MPRC poll was accurate

It’s quiet here at City Hall as elections officials check scanning equipment and prepare to open the State of Maine Room doors to the media and the public around 10 a.m.

It’s a good moment to size up the results of last week’s Maine People’s Resource Center poll with the reality of the Portland vote as we have it so far.

The MPRC, which conducted the telephone poll of 477 “likely Portland voters” then ran simulated “instant runoffs” like what the city will do later this afternoon (once they get through ballot scanning and software checking), found Michael Brennan coming out of proverbial Round 1 in first place with 27.4 percent of the first choice votes.

Using ranked choice voting for the first time, Portland voters were able to rank the candidates from No. 1 all the way to No. 15 if they so chose. On Election Day, only the first choice votes were counted, with today’s retabulation activity due to kick in if nobody received more than 50 percent of the votes.

My previous blog post explains the stages of that whole process.

So how did MPRC do in terms of accuracy?

Like I said, Brennan was projected in the poll to end Election Day with a lead with 27.4 percent of the first choice votes. In reality, he ended Election Day with a lead with 26.7 percent of the first choice votes. Pretty close.

The poll placed Ethan Strimling second after first choice votes were counted with 21.6 percent. Reality? Strimling in second with 22.4 percent. Pretty close again.

In third the poll had Nicholas Mavodones with 12.6 percent and David Marshall coming in fourth with 7.2 percent. Jed Rathband placed fifth in first choice votes in the poll with 6.7 percent.

The actual first choice votes counted by city officials on Election Day found Mavodones at 15 percent, Marshall at 7.7 percent, and Rathband at 7.1 percent.

If the poll is similarly accurate in its simulation of the reallocation of second choice votes, as will be done today to definitively determine the winner, Brennan will move on to victory. The poll found that after 12 rounds of “instant runoffs,” Brennan claimed 61 percent of the votes to Strimling’s 39 percent.

Where the poll broke with reality a little bit is it placed Deering High School teacher Markos Miller as the sixth place finisher after Election Day vote counts. City councilor Jill Duson actually came in sixth, with Miller finishing the day seventh.

The poll also overestimated the electability of John Eder and Ralph Carmona, and underestimated support behind Christopher Vail and Richard Dodge.

The poll lumped Vail and Dodge in one umbrella “bottom six” option for survey respondents, while breaking Eder and Carmona off as individuals in its top 9 that poll takers could rank individually. As it turned out, Eder, who admittedly threw his support behind fellow candidate Strimling about a week before Election Day, and Carmona were in the bottom six, while Dodge and Vail were among the top 9.

In fairness to the MPRC, those rankings won’t be a factor in who wins the overall race, and the organization was using its best information at the time (media reports, campaign financing and political experience) to guess which candidates would be in the top tier and which would be in the bottom.

Seth Koenig

About Seth Koenig

Seth has nearly a decade of professional journalism experience and writes about the greater Portland region.