We all just participated in our annual November election, and the Voting Methods Team is here to analyze the election through the lens of voting methods. Boulder County election results are available here.
Let’s tackle some of the questions about local elections that we’ve heard during this election season. We’ll cover questions on the city of Boulder (and touching on Broomfield), Longmont, and the St Vrain Valley School District (SVVSD), and conclude with general questions that pertain to multiple jurisdictions, including the Boulder Valley School District (BVSD).
Boulder (and touching on Broomfield)
First, some terminology: There are many forms of Ranked Choice Voting (RCV). Boulder started using the Instant-Runoff Voting (IRV) form this fall. A century ago, Boulder was a pioneer of the proportional multi-winner form of RCV which goes by several names, but the standard one is Single Transferable Vote or STV. As the Voting Methods Team, we like to use the most accurate name for each voting method.
Q: Was Boulder’s Instant-Runoff Voting (IRV) mayoral election a success?
A (part 1): For the IRV portion, the ranking and tabulation worked as intended, and we are not aware of any administrative or software errors, unlike New York City in 2021 or Alameda County (Oakland School Board) in 2022. Congratulations to the city and county clerks’ staff and everyone else who helped voters understand how to mark their ballots. However, we do see room for improvement in the way election results are reported.
- The most common comment that we heard about IRV after the election has persuaded the Voting Methods Team to advocate for single-candidate elimination rounds instead of batch elimination. So, in the Boulder contest, eliminate Tweedlie at the end of Round 1, then eliminate Speer at the end of Round 2. As one team member wrote, “batch elimination reduces one of the benefits of IRV, in that people can't see what the transfers were from each individual candidate as they were eliminated.” Colorado rules currently mandate batch elimination when possible. The Voting Methods Team has been in touch with the Sec of State’s office to get that rule changed.
- We would like more clarification in the percentages reported. The Daily Camera wrote, “Brockett won with 51.9% of the vote [sic].” If Brockett had won the first round of counting with that percentage of the votes – in other words, no instant runoff had been triggered – then we wouldn’t quibble.
- However, Brockett’s 16,823 votes are actually 51.1% of the 32,893 votes cast in the election. Exhausted ballots and overvotes accounted for 478 ballots or 1.5% of the votes. (Exhausted ballots were those that ranked Tweedlie or Speer #1 and did not rank Brockett or Yates so those voters did not have a say in the final round.) In New York City in 2021 Eric Adams won the IRV Democratic mayoral primary with only 43% of the vote. It’s this very scenario that led LWVCO to advocate successfully to remove the “majority criterion” from the proposed LWVUS concurrence before its adoption.
- Alternatively, one could measure percent support by dividing by the number of ballots cast in the mayoral contest. There were 34,249 ballots cast with more than 1,000 people leaving the mayoral contest entirely blank. Using this measure, the final-round percentages would be Brockett 49.1% support, Yates 45.5% and “Neither Finalist” 5.4%.
A (part 2): We are less confident that holding an election for mayor was to Boulder’s benefit. In addition to the fact that the mayor in Boulder is a “weak” mayor who runs the meetings, but has no extra voting power, consider these other observations:
- Boulder was going to lose at least one, maybe two, experienced public servants. Unlike the multi-winner city council contest, the mayoral contest has only one winner. Yates or Brockett or both of them would lose and no longer be in municipal government. In 2022 Boulder voters passed “2D: Charter Clarifications.” The opposition pointed out the problem of not letting a person run for both council and mayor. (You may recognize the author of the linked opinion, but she wrote in an individual capacity, not as a representative of LWVBC.)
- Mayoral-candidate quality was not noticeably improved.
- Three experienced city council members and one other, arguably minor, candidate vied to be mayor. Under the old system, Yates and Brockett would have had to run for and win city council seats, which seems likely, and then Speer (who was midway through her council term), Yates and Brockett could have vied to get the most support from their council colleagues to become mayor. IRV may or may not have led to more mayoral candidates. Maybe another council member would have thrown his or her hat into the ring or maybe the field would have cleared for Brockett to become mayor uncontested.
- RCV proponents often point to RCV electing more women and minorities. That should certainly be the case for proportional RCV (aka STV), but Boulder used winner-take-all IRV, not STV, and three of the four candidates in Boulder were WASP or WASPish males, including the winner. At least three of the four candidates, including the victor and the sole female, spoke of their or their family member’s sexuality or gender differences. Boulder has a strong history of electing female mayors under the previous council-choose-the-mayor system.
- Boulder budgeted about $200,000 for the mayoral contest. The city had hoped that Broomfield would share the ~$75,000 software license cost, but Broomfield did not hold an IRV election because none of its 6 contests had more than 2 candidates. (The argument that IRV leads to more candidates running definitely wasn’t true this year in Broomfield.)
- The argument that IRV saves money by eliminating a runoff election was never going to be true in Boulder or Broomfield because neither city holds runoff elections. Even when IRV eliminates runoffs, an RCV Resource Center “study is unable to show that implementing ranked choice voting has been responsible for any financial savings or liabilities in the cities that have chosen to use it.”
- The Voting Methods Team strongly believes that elections should be fully funded. We also strongly believe that spending money to implement a better voting method is money well spent. However, LWVBC was neutral on the 2020 ballot question to begin electing a mayor using IRV because LWVBC didn’t (and still doesn’t) have a position on directly electing a weak mayor. LWVCO supported Broomfield’s adoption of IRV because Broomfield only proposed changing from Plurality to IRV – an obvious improvement.
- Mudslinging in the mayoral contest has led to LWVBC members calling on our Local League to make a public statement about the negative campaigning. There was also some mudslinging in the council race, but many people seemed unaware of it.
- The impetus of a campaign finance complaint lodged against the Working Families Party National PAC was a negative mailer specifically advising how to rank two candidates and going negative against a third. IRV proponents like to claim that ranking candidates reduces negative campaigning. Usually, the comparison is with Plurality single-winner contests (which Boulder has never had except for recall or special elections), but, as we pointed out for the 2021 NYC contest, better voting methods don’t always preclude negative campaigning.
- Moving from council colleagues choosing the mayor at a televised meeting to the electorate voting in a single-winner contest was probably more responsible for the mudslinging than which voting method was used.
A (part 3): Looking forward, our LWVBC program positions guide our advocacy on some potential ballot measures.
- Some Boulderites and Broomfielders are advocating that the mayor should be elected to a 4-year term. LWVBC’s support for “a system of staggered terms in which at each election voters have the opportunity to elect at least half of those who will sit on the municipal governing body” means that LWVBC would be in opposition to only changing the mayoral term.
- The Voting Methods Team is asking the Boulder City Council to study proportional voting methods as a better way to elect city council. From 1917 (when Boulder became a home rule municipality) until 1947, Boulder used a proportional voting method in its elections. LWVBC supports “using a voting method which promotes proportional representation in city council elections.” The Town of Superior just created a Home Rule Charter Commission to draft a charter. Perhaps LWVBC should ask Superior to consider a proportional voting method to elect a future town council.
Q: If you want to give your vote completely to one candidate in an IRV contest, is the best thing to do to just vote for them as your first choice and not vote for anyone else? (A version of this question was the most popular question we got before the election.)
A: The answer is a resounding NO - it does NOT help your first-choice candidate to refrain from voting for anyone else, and your vote is most likely to have the most impact if you rank as many candidates as you actually have preferences among!
Here’s how it works: in Round 1, ONLY first-choice votes are counted. (So in Round 1 it doesn’t make any difference whether or not you ranked additional candidates.) If no one gets a majority of first-choice votes, then the candidate who got the FEWEST first-choice votes is eliminated.
In Round 2 (and later rounds, if applicable), if your first-choice candidate is still in the race, then your vote continues to count for that candidate and ONLY that candidate, and it still doesn’t matter at all whether you ranked additional candidates.
Where the benefit of ranking additional candidates comes in is if your first-choice candidate is eliminated before the final round. In that case, if you didn’t rank any other candidates, then you don’t get a say in any of the rounds after your favorite is eliminated. If you DO rank additional candidates, then your vote continues to count for whoever is your highest-ranked candidate who is still in the race.
So the upshot is, rank as many candidates as you have genuine preferences among, and you won’t hurt any of your higher-ranked candidates by including lower rankings. The more candidates you rank, the more likely you are to continue to have a vote in all rounds until a winner is determined.
Longmont:
Q: Did vote splitting impact any of Longmont’s municipal elections?
A: Two of the Longmont council races (Wards 1 and 3) do look like good arguments for better voting methods due to two similar candidates splitting the vote and enabling the third candidate to win. Additionally, the 3 wards are so unrelated to definable constituencies that a plan of proportional representation might have some traction in the future.
St Vrain Valley School District
Q: Why didn't the St Vrain Valley School District have a school board election?
A: Because, as is common in SVVSD elections, there was just one candidate per seat. But then, interestingly, after the candidate-filing deadline, one of those candidates dropped out of the race due to purchasing a house out of district. The school board accepted applications to fill the seat, interviewed three applicants and appointed one of them the day after the November election.
- Like in Broomfield this year and in past years, finding even two candidates to run in single-winner contests can be difficult. Converting from single-winner contests to a multi-winner at-large contest allows for voters to weigh in or more of the elected officials and provides a larger population from which to recruit candidates. More voice and more choice lead to more exciting, competitive races. Competitive races lead to higher turnout.
Multiple Jurisdictions:
Q: Why are there so many recalls, e.g., Oct 3 Louisville and Nov 7 Broomfield?
A: The signature threshold for recalling a ward member is typically considerably lower than the threshold for recalling someone elected at-large. (The group behind the recall effort can usually find someone to replace the recall target, or the group would be wasting time forcing the recall.) The last known recall in the all-elected-at-large Boulder City Council and BVSD was in 1974 which succeeded in recalling Council Member Tim Fuller, but not Mayor Penfield Tate II. In the mid-1990s an abortive attempt to recall BVSD Director Stephanie Hult failed due to not collecting enough signatures.
Q: Should we use a voting method that promotes proportional representation on city councils, school boards and other elected governmental bodies?
A: The Voting Methods Team is a fan of proportional representation! We see both the multi-winner contest aspect and the proportional aspect as good features.
As mentioned in the St Vrain Valley School District Q&A above, multi-winner contests lead to more contested elections. Also, LWVBC member Doris Flax (who is soon moving to Ohio) has been elected to school boards in two different states; she objects to BVSD and SVVSD’s single-winner contests with a district-residency requirement for candidates. When two good candidates run against each other, only one can win. (For instance, the Boulder Valley Education Association endorsed both District A candidates this year, but voters had to choose one candidate on their ballot.) She also feels strongly that “board members serve and make decisions for the ENTIRE school district... so they should [continue to] be elected by the ENTIRE school district.”
As we note in our Proportional Multi-Winner Voting Methods pamphlet, proportional voting methods can eliminate gerrymandering and give representation to minority groups in proportion to their percent of the electorate. We are proud of U.S. Representative Joe Neguse for being a co-sponsor of the Fair Representation Act. We were also happy to hear that Easthampton, MA, voted in November to use Single Transferable Vote (STV, aka Proportional RCV) for their multi-winner elections.
----------------------------------------