Nonprofit, nonpartisan journalism. Supported by readers.


Expect final Minneapolis mayoral results Wednesday after ranked-choice voting

This time, we should know the winners and losers in the mayor’s race sometime on Wednesday and all of the other results by that Friday.

Forget your Election Night victory party in Minneapolis and think instead of scheduling your party sometime later in the week.

But voters should know the results sooner than they did in 2009, when it took nearly three weeks to determine the winners in the first city election to use ranked-choice voting.

This time, we should know the winners and losers in the mayor’s race sometime on Wednesday and all of the other results by that Friday.

“On Election Night, we will only be counting the results of people’s first ranked choices,” said City Clerk Casey Carl, who explained that, by state law, officials are required to count the races in the order in which they appear on the Nov. 5 ballot.

They will start with the mayor’s race, move on to the City Council races (in an order yet to be determined) and then count the Board of Estimate and Taxation, Park Board at Large and finish up with Park Board by District.

Before the counting begins, they will establish a threshold number of votes a candidate in each race will need to be declared a winner with only the results of the first-choice votes. That threshold will be based on the number of ballots cast, not votes, and will be higher than the usual standard of 50 percent plus one.

“It’s a falsely inflated threshold so that if someone passes that higher threshold we know definitively that it is mathematically impossible for another candidate in that race to defeat them,”said Carl. If no candidate in a race reaches the threshold, the election clerks will declare that a runoff is required.

“We have an obligation for accuracy first,” said Carl of the threshold, which he notes is driving some of the campaign people “Looney Tunes.” 

The inflated threshold will prevail only on Election Night. “My best guess is that by 10 o’clock or 11 o’clock, at the latest, we will have completed all of the counting of the first-choice rankings in every race.”  

On the afternoon after the election, they will begin counting second and third choices — the runoff process — starting with the race for mayor. Carl said he expects to finish the mayor’s race on Wednesday.

On Election Night, the totals from all 117 city precincts will be transmitted wirelessly to the Hennepin County Elections Division, which will then send them to both the Minneapolis Elections Office and to the Secretary of State’s Office.

At the same time, the paper ballots will be transported to the Elections Training Center and Warehouse, where they will be secured as a backup to the electronic system.

The secretary of state’s staff will post the raw numbers on the department’s website as they are received. This will allow anyone with a calculator to add up the first-, second- and third-choice votes for any candidate and come up with a “winner.”  But that result might not be accurate.

That’s because under-votes, over-votes and voter intent come into play.

If a voter doesn’t list a first choice but does choose a second and third choice, that second choice will be counted as a first choice, and the third choice will move up to second.

“Say you rank someone as your first choice, you don’t rank anyone your second choice, and you rank someone your third choice, then you have under-voted in that race,” said Carl. The voting machine will capture that there is no vote in column two.

“We’ll have to normalize that ballot according to our voter-intent rules,” said Carl.  Those rules in Minneapolis mirror state statutes that require a ballot to be counted if it is possible to determine voter intent.

Because ranked-choice voting is not covered by state statute, Carl said, the City Council established the rules for voter intent in 2009 and updated the rules this year.

“In the past, there were specific instances, for example, where if you cast a ballot with a certain type of marking on it, we just threw it out and didn’t count it,” said Carl.

The new rules place the onus on election clerks to find a way to count ballots whenever possible. “In those previous instances, when we didn’t count a ballot in 2009, we will count it in 2013.”

“In 2009, if you skipped the first column, nothing else was counted,” said Carl. “You had to vote at least for column one” in a race for any of your votes to be counted.

Back to the under-vote and how it will be normalized. The voter’s first choice will remain a first choice, and the third choice will move up and become the second choice. This is what makes calculating a winner based on the secretary of state’s raw data dicey.

In ranked-choice voting, first choices are counted first and then, if needed, the count moves to second choices. But if a voter had no second choice, and that voter’s third choice moves to second, that is not reflected in the raw data.

The key here is that in ranked-choice voting when a winner is declared, the clerks quit counting.  If they declare a winner after tabulating the first two choices, the third choices will not be counted. But if you are looking at the secretary of state’s raw data you will not know that someone’s third choice has moved up to second and put some candidate in the win column.

An over-vote is when someone votes for too many candidates. The voting machine will catch this error, and the voter will be given a chance to try again.

Minneapolis election judges also have the ability to drop from the subsequent counts candidates who have no opportunity for a victory, which makes the raw data even less likely to be reliable.

With 35 candidates in the race for mayor, it is likely that some of them will not receive many votes. They can be dropped, and their votes no longer tabulated in runoffs, based on both the number of votes they receive and the spread between their totals or percentages, compared with others in the same race.

“The spread between the majority who are going to continue on and those who aren’t has to be so extensive that we feel very comfortable we can defeat them and move on,” said Carl, who added that the decision to drop a candidate from the counting process will be made on an individual basis but that more than one candidate could be dropped in any of the rounds of counting.

City Council races will be counted after the winner is declared in the mayor’s race. The order in which those races will be counted will be determined Oct. 29 during a 10 a.m. public accuracy test of voting machines at the elections training center at 732 Harding St. NE.  The order for counting Park District races also will be determined at that meeting.

Unlike a recount where ballots are examined in public, ballot counting in this election will take place in a private secured area at City Hall. There will be an area in the rotunda where vote tallies will be posted, but the information first will first be made public on the Minneapolis elections website.

You can also learn about all our free newsletter options.

Comments (25)

  1. Submitted by John Ferman on 09/24/2013 - 01:10 pm.

    RCV Threshold

    Carl’s description of the first round threshold is not in harmony with the ordinance language. In the first round only ballots with first choices are used to evaluate the threshold. The threshold is the defined Threshold which does not include partially defective ballots (voter intent can not be determined). There is a defined Maximum Possible Threshold but it is only required to be calculated. I would suggest Karen obtain Chapter 167 of City Ordinances and cross check what Carl said with what the ordinance says.

    • Submitted by Ken Bearman on 09/24/2013 - 03:31 pm.

      Casey Carl is right

      The threshold at the beginning, for first choice votes, includes all ballots as defined in 167.20: “Maximum possible threshold means the number of votes sufficient for a candidate to be elected under a first ranked choice tabulation under sections 167.60(b) and 167.70(b). In any given election, the maximum possible threshold equals THE TOTAL BALLOTS CAST that include votes, undervotes, skipped rankings, or overvotes for the office …”. [My emphasis]

      It’s not only defined, it’s used to determine winners “under a first ranked choice tabulation”.

      Casey Carl read, understood, and explained the ordinance correctly.

  2. Submitted by Robert Moffitt on 09/24/2013 - 01:43 pm.

    You know what result I expect?

    Whoever wins, the loser will be ranked choice voting, a bad solution to problem we never had.

  3. Submitted by Hiram Foster on 09/24/2013 - 04:33 pm.


    It has been quite the disaster. It’s strange but many people seem to have forgotten that we had ranked choice voting in the last mayoral election.

  4. Submitted by Walt Cygan on 09/24/2013 - 04:39 pm.


    I appreciate the enthusiam for RCV, but as my friend Sue knows, I am not a fan. My expectation is a remarkably low turnout for a highly contested mayor’s race, with a significant number of ballots having the mayor’s race left blank. Assessing a 35-candidate field is like reading a 35-page memo. By the time you get to then end of it, are you still paying attention? I think not.

  5. Submitted by Roann Cramer on 09/24/2013 - 10:50 pm.

    RCV, simple to understand and we get democracy with it.

    The Mayor’s office in Minneapolis should be elected by a majority of the voters. Waiting one short day to get the results in case of a runoff is worth it. Voters can chose there three candidates in their preferred order. Having a majority of the voters elect the next mayor ensures that voters retain their power. I believe it will improve turnout and voter participation.

    • Submitted by Dan Hintz on 09/25/2013 - 01:34 pm.


      It would be nice to have a mayor elected by a majority. Unfortunately, RCV does not necessarily provide majority winners. Well, not unless you don’t count the votes of people who voted for the wrong candidates.

      By the way, how is your work as Treasurer for Fairvote Minnesota going? We now have three Fairvote directors commenting on this article.

      • Submitted by Ken Bearman on 09/25/2013 - 02:30 pm.

        Gracious me!

        What disqualifies Minneapolis residents who’ve studied voting systems — in my case for nearly 17 years — from supporting RCV without getting ad hominem comments?

        And what are Mr. Hintz’s bona fides?

        • Submitted by Dan Hintz on 09/25/2013 - 03:14 pm.

          Bona fides

          This isn’t about bona fides – this is about honesty and ethics – the failure of Fairvote higher-ups to disclose their positions when they comment.

          And where is the ad hominem? Pointing out that the claim about majorities is false and the lack of disclosure of the Fairvote commenters is an as hominem.

          • Submitted by Ken Bearman on 09/25/2013 - 10:20 pm.

            It’s about facts and issues

            Ad hominem is about the person, not about what the issues s/he writes or talks about. The issue in this article is RCV. Being a volunteer, unpaid member of a board could be relevant to a person’s knowledge of facts or issues. It isn’t at all like being a member of a for-profit group that stands to gain monetarily or otherwise from advocating for an issue. Is the implication that board members — both, not three, BTW — can gain something by Minneapolis using ranked choice voting (RCV)? It can’t happen, and it’s not what we’re discussing.

            As I wrote, the issue is RCV in Minneapolis. Some speculate that turnout will be low. The City said it’s preparing for a high turnout, but we’ll know on November 6 when we have facts. Some speculate that if we had a primary, voters would be able to winnow 35 candidates to two. There’s no way to know, except that such voters would have had 35 candidates to evaluate and less information (in August) to do so. The fact is RCV is how we’ll vote, so maybe the doubters could hold their criticism until the facts are in — until after Minneapolis uses it in a non-incumbent election for Mayor and several Council seats.

            Speaking of facts, the local organization is FairVote Minnesota (FVM) — capital ‘V’ + Minnesota — not FairVote (capital ‘V’), the national group unaffiliated with FVM. The board members, as with most nonprofits, are there because they’re studied election issues and have thought in depth to reach conclusions about the deficiencies of winner-take-all elections. As I suggested, they gain nothing for themselves — nor does FVM — by supporting an issue they understand well and believe in.

            People certainly are entitled to their opinions about issues. But credentials ARE important when people claim things are or aren’t a certain way or that X, Y, and Z are true or false. Without a basis other than “I think so”, how do they expect to convince other people to agree with them?

            • Submitted by Dan Hintz on 09/26/2013 - 08:48 am.

              Ad hominem

              Sorry Ken, but critcizing Fairvote members for their lack of ethics is not disclosing their postions is not an ad hominem. Neither is calling them out for false and misleading astatements about RCV or pointing out that the court cited and fined them for dishonest and illegal campaign practices. Those are criticisms of the actions of Fairvote members, not of the characteristics of those people.

  6. Submitted by Hiram Foster on 09/25/2013 - 06:12 am.

    The Mayor’s office in Minneapolis should be elected by a majority of the voters.

    It is almost certain to be the case that the eventual winner will not be the first choice of a majority of voters. What is happening here is a redefining of the word “majority”, not any change to an underlying political reality. Meanwhile we have an election where no definition of candidates or issues ever occurred. Ranked Choice Voting has been and is a disaster for the people of Minneapolis.

    • Submitted by Steve Titterud on 09/25/2013 - 09:07 am.

      By the measure of your first sentence, the current systerm…

      …fails just as well !! Yet you don’t object to the current system putting plurality winners in office.

      Do you see that criticizing RCV for plurality winners while NOT criticizing the current election system for the EXACT same fault is contradictory, inconsistent, doesn’t hold water – that it is not a valid argument at all ?

      But your real objection seems to be that some proponents go too far in their claim of always producing a “majority” winner. I agree. They forget to identify the “majority” they are talking about, which is not necessarily a majority of the total valid ballots cast, the notion most people usually think of as the “majority”.

      The “majority” in the RCV regime is rather a calculated, round-by-round reassessment of vote totals based on the business rules of RCV. Both proponents and opponents of RCV misuse this term, by which I mean they often do not clearly specify which kind of “majority” they are talking about, attempting to impose their own notion of “majority” on the opposing side.

      This casts heat, rather than light, on the subject matter – much like the claim in your comment above.

  7. Submitted by Connie Sullivan on 09/25/2013 - 11:22 am.

    Even the RCV advocates know that not having a primary–which was a huge “selling point” for the new system–has been disastrous for the Minneapolis mayoral race: 35 candidates, with options going onto the back of a two-page ballot (nice, if your name is Andrew, under the alphabetical listings, but not so much, if it’s Winton or Woodruff). The advocates are now advocating that Minneapolis require a higher bar for entry to candidacy than the mere filing of a form with $20. Some sign of life beyond one’s own ego fulfillment.

    RCV may not be too complicated for most folks to understand in the voting part. But, boy! does RCV ever fail the test of transparency! The counting is impenetrably complicated and opaque. We can’t even agree on what “Threshold” means? And the city elections clerk has the “ability”–or the “discretion”?–to determine who has an impossible to comprehend too-low-to-ever-win first choice ballot count, so he drops them and they’re NEVER COUNTED?

    An “arbitrarily high” threshold is established for round one, so that even the famous 50-plus-one vote definition of majority by RCV sellers doesn’t obtain? That’s obtuse, and opaque, and it will lead to people’s doubting the validity of the entire mayoral election result. No conspiracy theory; people will doubt a result if they can’t understand it. And, believe me, the public will not understand it.

    Voting system wonks have to understand that reality is more difficult to deal with than mathematics. It’s politics that they’re avoiding, like the candidates, who so far have presented us with mostly mush (wanting to be Your Number Two Choice).

  8. Submitted by Kelly O'Brien on 09/25/2013 - 11:50 am.

    Worth the Wait

    We could start our electoral process in August with a primary that very few participate in, and then three months later we find out the winner. Or we could save ourselves more than $200,000/election by using ranked choice voting once, in November (when more voters participate), and get our results mostly within 24 – 48 hours.

    RCV is better democracy. More voters, majority winners.

    • Submitted by Dan Hintz on 09/25/2013 - 12:47 pm.

      Three strikes and you’re out

      Fairvote apologists like to claim that elimination of primaries saves money, but there is no evidence to bear that out. Minneapolis spent more on its last RCV election, which to be fair, was its first. But if you are going to make that claim, there must be some evidence to bear that out.

      The same is true with voter turnout. Its all theory and no evidence.

      The big (and third) strike is that RCV provides majority winners. That one fails, not from lack of evidence, but because it is just outright false. Some RCV apologists will argue that they achieve majorities by only counting certain people’s votes, but in the real world if the winner is chosen by less than half of those who showed up to vote, no majority exists.

Leave a Reply