Ranked-Choice Voting - Center for Election Confidence

**Study – RCV Weakens Influence of Minority Voters**

Overview

In standard election systems, voters select their preferred candidate, and the candidate with the most votes wins. But in jurisdictions that use Ranked-Choice Voting (RCV), also referred to as instant runoff voting, a voter is tasked with ranking each candidate in the order the voter prefers them to win. If a candidate does not receive more than 50% of votes in the first round of tabulation, the candidate with the least number of votes is eliminated. Voters whose first choice was the eliminated candidate will then have their second-most preferred votes reallocated among the remaining candidates. This process continues until one candidate receives a majority of the remaining votes. While proponents claim RCV is a positive voting reform that would improve elections and voter participation, RCV causes more issues than it solves. Below exists a sampling of the many problems with RCV elections.

Ballot Exhaustion Means Ballots are Excluded from the Final Vote

Perhaps the biggest issue with RCV occurs when a voter fails or refuses to vote for more than one candidate. If that voter’s choices are eliminated after the first or second round of tabulation, his or her ballot is effectively cast aside and not used in the final vote. In other words, the ballot is exhausted. It is exceedingly common that ballots in RCV elections are exhausted and not included in the final vote due to the voter not voting for every candidate. This offends the bedrock principle of “one person, one vote” in American democracy since some ballots count while others inevitably do not.

The Maine Heritage Policy Center and Alaska Policy Forum studied 96 RCV elections across the country and found that on average 10.92 % of ballots were exhausted by the final round of tabulation and ultimately not used in the final vote.
One study found that 9.6-27.1% of ballots never make it to the final round of voting in RCV elections.
As the Foundation for Government Accountability explains, “[a]fter 20 rounds of tabulations, 53% of ballots had been exhausted, and only 8,200 ballots––or 47%–– contributed towards the final results” of San Francisco’s 2010 Board of Supervisors General Election.

Increased Voter Disenfranchisement

RCV has great potential to disenfranchise voters who do not understand how the non-intuitive system works. Where voters are accustomed to voting for only one candidate, RCV elections require voters to vote for every candidate based on preference. This leads to vulnerable voters being disenfranchised.

According to Maine Heritage Policy Center and Alaska Policy Forum, disenfranchisement in RCV elections is mostly felt by minority voters and non-English speaking voters because these voters more often fail to rank more than one candidate. Senior voters are likewise particularly vulnerable to disenfranchisement under RCV systems for the same reasons.
Election data from New York City’s RCV June 2021 primary election shows that whiter, wealthier neighborhoods were more likely to employ the new ranking system than lower-income areas of the city, many of which are home to Black, Latino and Asian communities.
In Minneapolis’ affluent voting districts, ballot error was as low as 2%. But in less affluent districts, which have a higher rate of minority voters, ballot error was as high as 20%.

Compounded Voter Confusion

The added steps voters must follow in RCV elections fuels voter confusion. As a result, RCV elections make it more likely that disadvantaged voters will be left out of the political and electoral process.

Governors in the past have vetoed RCV legislation approved by the legislature because it makes elections unnecessarily confusing. For example, California Governor Jerry Brown Jr. (D) vetoed RCV legislation for how it makes voting “overly complicated and confusing” and “deprives voters of genuinely informed choice.” He further stated, “In a time when we want to encourage more voter participation, we need to keep voting simple.” In October 2019, when vetoing an RCV bill in California, Governor Gavin Newsom (D) said: “I am concerned that it has often led to voter confusion and that the promise that ranked-choice voting leads to greater democracy is not necessarily fulfilled.”
Lisa Goodwin, Bangor, Maine’s city clerk, said many voters were confused by the state’s new RCV system––of the 4,555 ballots cast in Bangor on Election Day in 2018, about 200 (over 4%) were spoiled because of voter error from confusion over RCV, which is more than a typical election.
A 2018 Pew post-election study in Maine, the first place to use RCV in a federal election, found that 9,000 voters or 6% of ballots were not counted either due to voters being confused, their ballots being exhausted, or the voter simply choosing to leave some options blank.
RCV is also very confusing to election officials. The February 3, 2020, Democratic Party presidential caucuses in Iowa, which used a system of RCV, went so poorly that the national Democratic Party leader called for a recount after it took days to determine a winner.

Subsequent Repeals of RCV

It is not uncommon for voters to quickly have buyer’s remorse after approving RCV, which leads many jurisdictions to repeal the legislation. This is often a result of negative experiences with RCV. For example:

Burlington, Vermont adopted RCV in 2005 only to repeal it in 2010 after their unlikeable mayor was re-elected with just 29% of the first-place votes.¹
Ann Arbor, Michigan adopted RCV for mayoral elections in 1974 but repealed it in 1976 with the support of 62% of voters.
North Carolina adopted RCV for judicial vacancies in 2006, but the legislature repealed the law in 2013 after the system received mixed reviews.
After adopting RCV in 2009, voters in Aspen, Colorado repealed RCV in 2010 by 65% over frustrations with the logistics of implementing RCV.
Voters in Pierce County, Washington adopted RCV in 2006 only to repeal it by 71% in 2009 after voters disliked the system.
Voters in Alaska approved a ballot measure in 2020 to implement RCV for federal elections; however, voters will likely consider a ballot measure to repeal RCV in 2024 after RCV was the subject of much controversy in Alaska’s 2022 midterm elections.

Lack of Evidence that RCV Increases Voter Turnout

Advocates allege RCV increases voter participation, but there is little statistical support to this assertion. In fact, there is equally compelling evidence that RCV’s complexity decreases voter turnout, for voters might stay away from the polls because they do not understand the ballot or the process or do not wish to spend additional time researching candidates.

After claiming RCV increases voter turnout by 10 points, FairVote admitted RCV elections have “little impact on electoral participation.”
A study of San Francisco’s mayoral elections from 1995 to 2011 shows that voter turnout decreased in RCV elections, especially among black and white voters, younger voters, and voters who lacked a high school education.
In both Oakland, California, and Minneapolis, Minnesota, voter turnout decreased in high-minority precincts, and in Oakland, voters in high-Latino and Asian precincts tended not to rank their full ballot.
According to the Boston Globe, a post-election survey of eligible voters in Maine who did not participate in the 2018 election found that 26% did not participate due confusion over RCV.

Costs of Implementing RCV Elections

RCV proponents underestimate the cost required to adequately educate voters on how to utilize the new voting system. Switching to RCV elections can cost jurisdictions millions of dollars and quickly negate any potential cost-savings of not needing to conduct a separate runoff election. Furthermore, expensive software and equipment are required to run RCV elections and process RCV ballots.

New York City spent $15 million on an education campaign to explain RCV to its voters, and the system was still plagued by issues and voter confusion.
Maine spent nearly half a million dollars on equipment and software to conduct its first RCV elections in 2018.
Nevada anticipates it will cost $3,200,000 to switch to RCV and cover expenses required for “voter outreach and education, increased ballot stock costs, personnel expenses, equipment, software and programming costs for voting machines and updates to training materials.”

Delayed Election Results

While statistical support for how RCV delays election results is limited, the added steps required to process and tabulate RCV ballots certainly does not speed up the counting process.

As NBC reports, “Some recent ranked-choice elections have taken several days to sort out a winner, but advocates stressed that was due to how and when officials chose to count ballots, not the ranked-choice process itself. Last year, the winner of New York City’s Democratic primary for mayor was called two weeks after primary day. And in 2018, it took over a week to determine who won Maine’s 2nd District race.”
The Foundation for Government Accountability highlights that in the Minneapolis’ 2009 mayoral race, the election office estimated that, for a turnout of 70,000 voters, tabulation of RCV ballots would take 37 eight-hour shifts with 102 election officials working.

Administrative Errors

The added steps required to conduct RCV elections creates greater opportunity for administrative errors. Worse, these errors can easily go undetected if not caught early. Undetected errors have led to situations where a jurisdiction certifies an election only to later discover a fatal error in the result allowed the wrong candidate to be declared the winner. For obvious reasons, this adds to election controversy and undermines public confidence in election results. Even if an RCV election is run smoothly, voters might still be hesitant to accept the official results due to a potential undetected error.

On December 28, 2022, the Alameda County Registrar’s office announced it had made a major algorithm error that resulted in inaccurate tallies of ballots in Oakland, California’s RCV elections, leading to the wrong candidate being certified as the winner of the Oakland Unified School District 4 race. The error was discovered after members of the California Ranked Choice Voting Coalition informed the Registrar that they had independently run the county’s election data through their own RCV tally software and obtained a different result than what was certified on December 8, 2022, (one month after the election) as the final outcome. On January 10, 2023, the Alameda County Board of Supervisors in Oakland, California voted unanimously to seek a recount of several RCV races in the Nov. 8 election because of the error.
In 2021, New York City’s first RCV election was marred with controversy after it was discovered that 135,000 test ballots were counted in the reporting results. Worse yet, this error wasn’t discovered until after the 11^th round of tabulation. The New York City Board of Elections released a statement explaining the discrepancy was caused by the inclusion of test results that were done before primary day and had not been cleared from the system before the first round of tabulation was conducted. The 135,000 test ballots had been counted as real votes before the error was identified.

Click here to download a copy of CEC’s ranked-choice voting issue brief.

Footnotes

Burlington then re-adopted RCV for city council elections in 2021. See Josh Altic, Burlington, Vermont, voters approve a measure March 2 to use ranked-choice voting for city council elections, BallotPedia News (Mar. 2, 2021), available at: https://news.ballotpedia.org/2021/03/02/burlington-vermont-voters-approve-a-measure-march-2-to-use-ranked-choice-voting-for-city-council-elections/. ↩︎