Can Ranked Choice Voting Transform Our Democracy?

(Cesar Calderon/BGA)

Perhaps you’ve heard the buzz about something called ranked choice voting. Instead of marking their ballots for just one candidate, voters are asked to rank two or three or more, with the winner determined through multiple rounds of counting if needed. 

San Francisco does it. Minneapolis does it. In November, Maine became the first state to use ranked choice voting in a presidential election. There’s been some early talk of adopting ranked choice voting, or RCV, for municipal elections in Chicago, and several bills have been floated that would extend RCV to General Assembly and statewide races.

You can expect to hear more as RCV gains support across the country. Here’s a primer.

What is Ranked Choice Voting? How does it work?

Most of the elections in the United States are plurality elections, or “first past the post”: whichever candidate has received the most votes wins, irrespective of the vote share. This means that sometimes candidates can win without receiving a majority of the vote. 

Ranked choice voting  is an electoral system that allows voters to rank candidates by preference (i.e., first choice, second choice, and so on). The votes are counted in rounds based on the first choice listed on each ballot. If one candidate has received a majority of the votes (over 50 percent) after one round of counting, the race is over. If not, then a second round of counting commences. The candidate with the fewest first-place votes is eliminated, and their votes are reallocated to the voter's second choice candidate. These rounds continue until one candidate has a majority. This system is similar to a two-round or runoff electoral system, but in the case of RCV, voters only need to take one trip to the polls. (NCSL)

(Cesar Calderon/BGA)

What are the pros of such a system?

As the chasm between Democrats and Republicans has grown wider and ever more contentious, proponents of RCV have touted its potential for reducing negative campaigning. They argue that because the system incentivizes second- and even third-place rankings, candidates would need to appeal to a broader base of voters; denigrating an opponent and their platform (and, by extension, their voter base) is no longer a viable campaign strategy. This shift in campaign strategy towards prioritizing voters might also serve to lessen the influence of monied interests.

RCV is also framed as an antidote to the current two-party system because it allows more candidates to compete. In plurality elections, because the goal is simply to get more votes than anyone else, it’s possible for candidates with longer odds of winning to siphon votes from bigger party candidates (see: Ralph Nader for Al Gore, Jill Stein for Hillary Clinton). Without the stigma of supporting a “spoiler,” voters can feel free to vote based on their true intentions instead of on a calculation of who has the best likelihood of winning. In that same vein, fewer candidates would be discouraged from running for fear of splitting voters, allowing more third party, independent, and minority candidates to enter the fray. Supporters also believe that more choices that may speak to more voters could boost voter turnout and engagement.

RCV also eliminates the need for subsequent runoff elections, so voters need to show up only once. Proponents argue that this would boost voter turnout and eliminate the costs associated with administering a runoff election.

What are the cons?

Any change to the election system requires public education on the mechanics of the system, how it might change the way electors vote, and the system’s implications. It also requires more education — not to mention more work — for clerks and election volunteers. There’s no guarantee that financial costs would be lowered, especially at first: There are costs associated with increased education, and potentially the need to pay for more ballot pages, ballot transport to a central counting location, and new ballot counting equipment or software.

Opponents argue that RCV is too confusing for voters, and could result in more spoiled ballots if voters do not rank more than one candidate or if they incorrectly rank candidates more than once. The confusion may even discourage voters from participating, potentially rendering the “higher turnout” argument moot. Opponents also note that additional rounds of counting mean voters have to wait longer for election results, sowing frustration and distrust.  

Is anybody using this right now? How has it worked? 

A number of cities (including San Francisco; Berkeley, Calif.; and Minneapolis-St. Paul, Minn.) currently use RCV. (Fun fact: so do the Oscars, which has been selecting winners using RCV since 2009.) 

In 2016, Maine became the first to adopt RCV statewide after a citizen-initiated referendum passed. Dating back to the 1970s, nine of Maine’s last 11 gubernatorial elections were won by candidates who did not receive a majority of the vote. In November 2020, Maine became the first state to use RCV in a presidential election.

Alaska became the second state to adopt RCV after voters approved a ballot initiative in November 2020. Instead of closed primaries, Alaska will have all candidates on a single ballot, with the top four candidates moving on to the general election. Voters will then rank candidates in order of preference. The change will take effect for the 2022 election.

Nearly 75 percent of New York City voters backed an RCV referendum in 2019, and got their first look at the RCV system in a special election this February. They will use RCV for the upcoming June mayoral primary.

But not everyone is enamored with RCV. Massachusetts voters considered a citizen-initiated referendum this past November, but 55 percent of voters decided against adopting RCV for statewide elections. Voters in Aspen, Colo., actually repealed RCV in 2010 after voting to adopt the system in 2007. One reason for the change of heart may be RCV’s seeming irrelevance in city races. In each of the four municipal elections for which RCV was in effect, the candidate who was ahead after the first round of counting eventually won the runoff.

How does RCV affect the work of election authorities?

One important group isn’t sold on RCV: municipal clerks. Clerks are the front line of any election change. They must train poll workers on how ranked choice voting works, they must educate voters, and in some cases, they are tasked with counting ballots.

In a 2019 survey, municipal clerks in Maine were asked about the cost and administrative burden of running an RCV election; voter education efforts and voter competence with new rules; and how ranked choice voting worked overall in the state’s 2018 elections. Responses were received from 110 out of roughly 480 clerks in the state. Of those who responded, 41 percent were Republican, 38 percent were unenrolled (Maine does not use the term “Independent” for those who don’t declare a party), and 21 percent were Democrats.

Across the board, the clerks indicated they generally did not see major new costs associated with implementing RCV, perhaps because of the training materials produced and shared by the Maine Secretary of State’s office. But almost two thirds of the clerks said RCV increased their administrative burden. (Unlike the cost question, the answers to perceived burden were markedly partisan. About half of the Democratic respondents noted an increased burden, compared to 83 percent of Republican respondents.) 

The increased burden included programming and testing voting equipment before Election Day and training poll workers on how to answer voters’ questions (more than two thirds of respondents felt that voters didn’t understand RCV rules). Maine’s mixed ballot also complicated matters: Statewide and federal races are decided by RCV, while municipal and state legislative races are decided by plurality, meaning voters see two different voting methods on the same ballot. Some clerks have to use two different scanning methods to tabulate votes; clerks without access to RCV-capable counting machines had to count ballots by hand and then again with electronic tallying machines for the plurality races.

Overall, nearly 90 percent of the survey respondents did not want to continue using RCV in Maine. (Southern Political Science Association 2020 paper)

In an op-ed to the Colorado Springs Gazette and The Maine Wire, El Paso County, Colo. Clerk Chuck Broerman estimated that county-wide adoption of RCV would cost $340,000 upfront and another $70,000 per year. He estimated that statewide adoption would cost $2.4 million to $3 million upfront and $350,000 to $400,000 per year. Broerman noted that without state funding, the cost of such changes would fall to local taxpayers through increased property taxes. (The Maine Wire)

Why are we hearing so much about ranked choice voting lately?

The current contentious political climate — particularly the most recent presidential election — has led to many calls for an overhaul of the American electoral system, with RCV frequently mentioned as a potential alternative voting method. Former presidential candidate hopefuls Elizabeth Warren and Andrew Yang endorsed RCV. 

Across the U.S., more states and municipalities are considering turning to RCV. Voters in Austin, Texas, recently approved a measure allowing RCV for city elections. 

New York City will debut RCV for its mayoral election this June. 

Is there a chance Illinois or Chicago will adopt RCV?

Voters in Springfield passed a measure in 2007 to allow military and overseas citizens to fill out an instant runoff ballot for local elections. Because runoff elections are typically held shortly after the primary or general election, citizens living or deployed overseas might not have time to receive a second ballot and return it for their vote to be counted. A ranked choice ballot allows these voters to participate in a runoff election if one is required.

Statewide RCV has been proposed in the Illinois General Assembly before: then-Sen. Barack Obama (ever heard of him?) sponsored a bill in 2002, and then-Sen. Daniel Biss did so in 2017. Neither bill saw the floor. Sen. Laura Murphy sponsored a bill in 2019 (SB2267) and again this session (SB1785). Rep. Jaime Andrade has introduced a similar measure (HB2416).

How might RCV be beneficial for Illinois or Chicago? 

Supporters say RCV might have alleviated some of the nastiness of the 2018 Democratic gubernatorial primary. Six candidates competed for the nomination, and eventual winner J.B. Pritzker dropped a record-breaking $70 million on his campaign. Pritzker handily defeated incumbent Bruce Rauner in the general election.

Chicago also might have benefited from RCV in its 2019 mayoral election. The record 14 candidates spent over $30 million during the campaign. In the first round of voting, the two candidates who moved on to the runoff election received barely a third of the total votes cast for mayor. In other words, over 60 percent of Chicagoans voted for someone other than Toni Preckwinkle or Lori Lightfoot.

One intriguing contender was Amara Enyia, a political strategist backed by many young voters (not to mention Chance the Rapper and Kanye West). Her supporters feared that Enyia would not advance to the runoff election, and a vote for her would weaken the chances of a less progressive but more politically viable candidate. Despite a lot of buzz, Enyia received less than 8 percent of the vote. It’s possible that the “political viability” calculation her supporters had to make resulted in votes being funneled to other candidates instead. 

Looking ahead to 2023, Chicago could be in for another contentious election. Nearly three quarters of those who answered one recent survey said they felt Chicago is headed in the wrong direction, with only 16 percent saying Mayor Lightfoot is doing a good or excellent job. 

 A ranked choice approach in 2023 could force mayoral hopefuls to work harder to build broader coalitions of supporters, relieving Chicago voters from the rancor and confusion of another 14 mayoral candidates.

(Cesar Calderon/BGA)

So… does it work?

How do the pros and cons of ranked choice voting actually play out? Supporters argue that RCV reduces negative campaigning, allows more candidates — specifically third party, independent, and minority candidates — to compete, improves voter turnout, and lowers costs for both election administration and candidates. Opponents argue that educating clerks, volunteers and the public is an administrative and financial burden, and that RCV is confusing and even intimidating for many voters — possibly resulting in more spoiled ballots and lower turnout.

In practice, RCV has been a bit of a mixed bag:

There is evidence that RCV makes campaigns more pleasant. In a survey of voters and candidates in cities with both types of elections, those in RCV cities reported greater satisfaction with the choice of candidates, found information shared through campaigns to be more useful, and were generally more interested in the election. 

Researchers even pored over transcripts of debates from elections between 1988 and 2020 elections, analyzing over 400 videos from more than 70 municipalities.They found that since the introduction of RCV, debates have been more civil, with candidates using more positive language, avoiding negative or even neutral words.

That positive feeling didn’t necessarily translate to improved election participation. Researchers from the University of Missouri-St. Louis examined 96 elections from 2009 through 2015 to compare voter turnout rates in RCV and plurality cities. (They left out elections in presidential election years because turnout is historically higher, making it hard to isolate RCV’s impact on turnout.) The researchers used a differences-in-differences (DID) regression analysis, meaning they compared average voter turnout rates in RCV and plurality cities before and after RCV implementation. Many factors can affect turnout (education level of a city, weather, differing civic culture),  and the DID approach allowed researchers to isolate the specific impact of RCV. 

A common criticism of plurality elections is that voter turnout drops between a primary and runoff election.To compare voter turnout in cities with ranked choice voting vs. cities with plurality voting, the researchers first looked at elections before the RCV cities made the switch. They found that turnout rates for primary elections already averaged 7 percentage points higher in the RCV cities. After the switch, the difference averaged 14 percentage points, suggesting that RCV increases turnout in primary elections.

In general elections, though, the average difference dropped from 9 percentage points before the switch to 6 percentage points after. These analyses suggest that RCV elections improve voter turnout for primaries, but they don’t have much of an impact on improving turnout for general elections.

While RCV improves voter turnout in some cases, it isn’t necessarily improving turnout across all demographics. Dr. Jason McDaniel, a professor of political science at San Francisco State University, analyzed RCV’s impact on voter turnout rates by demographic. He examined San Francisco mayoral elections from 1995 to 2011 (the city switched to RCV in 2007). McDaniel found that RCV decreased turnout among Black and white voters, voters with lower than a high school education, and younger voters. He also found a correlation between incorrectly marked ballots and foreign-born voters and voters whose primary language was something other than English.

Early evidence suggests RCV has a positive effect on encouraging diverse candidates to run for office. Researchers — all of whom are affiliated with FairVote, an electoral reform non-profit — examined municipal elections in RCV cities (San Francisco, Berkeley, Oakland and San Leandro) and non-RCV cities (San Jose, Alameda, Richmond and Santa Clara) in California from 1995 to 2015. Candidates of color ran for office more often in RCV cities even before RCV. But they made up over a quarter of total candidates after the switch, more than 14 percentage points better than in non-RCV cities. 

While it’s important to get voters to the polls, it’s equally important that they fill out their ballots accurately so they are counted. In expert testimony in a court case challenging the constitutionality of Maine’s ranked choice voting law (a federal judge eventually tossed the lawsuit), Dr. Nolan McCarty, a professor of politics and public affairs at Princeton University, found that a significant number of ballots cast in RCV elections are non-factors. McCarty performed a statistical analysis of 98 RCV U.S. general elections held from 2006 to 2019. He found that on average, nearly 11 percent of votes cast in an RCV election are “exhausted,” meaning they aren’t counted toward determining the winner. McCarty considered ballots exhausted if the only candidates ranked on a ballot were no longer viable, an elector selected more than one viable candidate for a ranking, or an elector skipped rankings (e.g., they did not rank additional candidates beyond their top choice). In 15 of the 98 elections examined, more than 20 percent of ballots were considered exhausted. McCarty considered the argument that voter confusion diminishes with time and practice as voters get used to RCV, but did not find empirical evidence that ballot exhaustion rates declined with greater exposure to RCV. 

In Maine’s 2018 election, McCarty found higher rates of ballot exhaustion in towns with more elderly voters or more voters without a college degree. This was especially true when looking at “truncated” ballots, in which a voter ranks or choses fewer candidates than allowed (ranking only a first and second when asked to rank their top three, for example). (The Maine Wire)

In another study, researchers examined four RCV municipal executive elections (three in California and one in Washington). All four asked voters to rank three candidates, used the same ballot design and election technology, and had publicly-available digital images of ballots cast. 

Supporters of RCV say  one of the biggest benefits of RCV is that it rewards only the candidates that have majority support. But in the four observed elections, none of the winners won a majority of votes cast in the election — only a majority of those that made it to the final round of counting. This means a significant number of ballots cast were exhausted before the deciding round. The rate of exhausted ballots ranged from 9.6 percent in San Leandro to 27.1 percent in San Francisco.

Again, a common theme across the four elections was voter error, either not submitting a complete ballot (i.e., ranking fewer than three candidates) or mistakenly ranking the same candidate more than once. More than a quarter of Pierce County voters ranked only one candidate in a four-candidate race. Almost 12 percent of San Leandro voters ranked the same candidate more than once. In San Francisco, 5 percent of ballots had duplicate rankings — even though San Francisco has had ranked choice voting since 2004. This suggests RCV is still confusing to voters, even after substantial public education and experience.

Hype vs. hope

Given how polarized our politics have become, ranked choice voting is an intriguing alternative to the status quo. It has devout believers and vocal detractors, but the rest of us might have trouble choosing a side. That’s because RCV doesn’t have a meaningful track record — and of course it won’t have one until more jurisdictions adopt it. 

So for Chicago or Illinois (or anyone), the choice is either wait and see or take a leap of faith. In making that decision, they should consider these factors:

Cost: Ranked choice voting is supposed to save money, largely because it eliminates the need for runoff elections. So far, municipalities that have adopted it are spending a lot of money on voter education (the jury is also still out on whether these efforts are effective) and ballot counting (machines, ballot transportation, central counting location rental and security). They’ll need a longer time horizon to see if there are real cost savings.

Learning curve: The rates of spoiled and exhausted ballots, even in municipalities that have experience with RCV, suggest that the electoral change is confusing. In study after study, minority, low-income and less educated voters are more likely to cast ballots that are spoiled or not counted in the deciding rounds of vote tabulation. Those rates haven’t been high enough to change the outcome of those elections, but they raise concerns that historically disenfranchised groups are disproportionately harmed by the change.

Effect on voter turnout: RCV does appear to increase turnout in some elections, but it doesn’t seem to draw more low-income or minority voters as claimed and may actually widen the access gap further. Again, this could improve with experience.

New candidates, new ideas? Ranked choice supporters say the plurality system discourages challengers who might offer different ideas — particularly women and minority candidates — from entering the political fray for fear of siphoning votes from more familiar and more “viable” candidates. 

Too often, voters are forced to choose the candidate they find least objectionable instead of the one whose policies speak to them the most. RCV promotes diverse representation by giving voters more choices of candidates who more accurately reflect their identities and interests. 

Analysis of California elections found that women (and women of color) were more likely to win elections in municipalities that had switched to RCV. Still, the winning candidate has to garner the broadest base of support or appeal to the most people. While a wider range of views may be championed in an RCV race, it doesn't necessarily mean that candidates who support emerging policies win more often. Candidates may actually take more centrist stances or advocate for incremental change in order to build bigger coalitions. Some voters may view those types of candidates as more of the same.

Changes take time. So do multiple rounds of counting. It will take time for municipalities to develop new election protocols (e.g., how ballots are counted or transported) and train poll workers on the system change. They’ll also need to work harder to educate voters on how to cast their ballots. And they’ll need to prepare the public for longer waits to learn election outcomes, which may take longer to certify. 

The switch is easier for cities than for states. Implementing RCV statewide would require officials to accommodate varying voting systems used by individual election authorities, as well as coordinating the collection and transport of ballots for central counting. Municipal elections are less likely to require confusing mixed ballots, on which voters are asked to vote in both RCV and plurality races. (Confusing ballot designs yield more voting errors, the impact of which falls disproportionately on low-income and minority voters.)

The takeaway, for now: Ranked choice voting carries some lofty aspirations. It aims to encourage candidates that better reflect our diverse populace, to reward coalition-building instead of partisanship, to build a more engaged electorate and to move our government toward a more truly representative democracy. Oh, and to save some money, too. 

It hasn’t fully delivered on its promises. It’s gaining momentum because our current system is far from perfect, too.  No election system is a silver bullet. But ranked choice voting is worth watching closely — and with an open mind.