Executive summary

• In ranked choice voting (RCV) elections, candidates can benefit from being the backup choice of voters who rank another candidate first. As a result, several pairs of candidates have “cross-endorsed” each other in RCV contests; these candidate pairs encouraged their supporters to rank them first and second. 
• Cross-endorsements and/or joint campaign activities can promote campaign civility, prevent vote-splitting, and lead to better representation. These tools may boost the performance of candidates who participate, even if they don’t always translate into wins.
  • There have been 440 RCV contests since 2010. FairVote has identified 29 with formal cross-endorsements or joint campaigning. 
  • Of these 29, candidates who cross-endorsed won in 11 races.  
• The tangible impact of cross-endorsements is difficult to measure empirically, due to extraneous campaign variables like whether the cross-endorsing candidates were already closely aligned on policy, whether the candidates were already competitive, and whether local media or civic organizations also offered similar cross-endorsements. Instead, we present a series of case studies.
  • In Portland’s 2024 District 3 City Council race, candidates Angelita Morillo and Tiffany Koyama Lane cross-endorsed one another and actively campaigned together. Both won seats in their three-member district, and both were the most popular second choice among voters who ranked the other first.
  • Sarah Palin lost Alaska’s 2022 special congressional election. The regular election was held two months later and presented a rematch between the top three candidates. In this contest, Palin changed her strategy and cross-endorsed fellow Republican Nick Begich. When Begich was eliminated, 67% of his supporters’ ballots transferred to Palin, compared to 50% in the special election. This improved performance was not enough to defeat Democrat Mary Peltola. 
  • In New York City’s 2021 Democratic mayoral primary, after Andrew Yang and Kathryn Garcia campaigned together, Garcia’s share of ballots transferred from Yang supporters was 11 percentage points higher than polls suggested. 
  • In San Francisco’s 2020 Board of Supervisors race, two Chinese American candidates cross-endorsed each other. One of them won by 125 votes, with the help of support from the other. Their team-up may have helped the Chinese American community win representation in San Francisco.
  • In Oakland’s 2018 District 6 City Council race, a slate of female candidates campaigned together. The outcome of the race demonstrated that underdog candidates, especially when working together, can perform better under RCV than they might have under single-choice voting. 

Background

Candidates can cross-endorse each other by supporting one another’s bid for office. Similarly, candidates may campaign jointly by collaborating on campaign activities. This could mean speaking at the same campaign rallies, appearing at the same events, raising funds together, sharing advertisements and branding, or even incorporating under the same campaign organization.

This typically does not happen within single-choice elections, because voters can only pick one candidate. Bolstering one’s opponent is functionally equivalent to harming one’s own chances. 

Ranked choice voting (RCV), on the other hand, lets voters rank multiple candidates. In a number of RCV races, candidates have encouraged their supporters to rank an ally second. For example, in the 2018 San Francisco mayoral race, candidates Jane Kim and Mark Leno jointly released advertisements encouraging voters to rank them first and second. In a 2023 New York City Council primary, candidates Yusef Salaam and State Assemblyman Al Taylor cross-endorsed each other, donning matching suits and ties during their announcement. Salaam went on to win the race.

This strategy works because voters’ second choices are not tallied unless their first choice has been eliminated. As a result, if Candidate A endorses Candidate B as a second choice, that endorsement carries no electoral penalty for Candidate A. Candidate A is essentially saying, “if I can’t win, I would like my supporters to help Candidate B win.” This is referred to as “ballot transfer” or “transfer ballots”; if a voter’s favorite candidate gets eliminated, that voter’s ballot can count for (or “transfer to”) a backup choice. 

This dynamic can lead to more positive, issue-oriented campaigning. If Candidate A attacks Candidate B, Candidate A will drive potential backup support from Candidate B’s coalition away. If Candidates A and B align on one or more issues, it would be to their advantage to court each other’s supporters and find areas of consensus. To differentiate themself, Candidate A may highlight their unique appeal to voters, rather than arguing how Candidate B falls short. 

Cross-endorsing can also lead to more representative government. Under single-choice voting, votes are often split between candidates of the same ideological, demographic, or other affiliation, preventing some groups from winning seats (and discouraging newcomers from running). With RCV, votes for a certain bloc can consolidate around a viable candidate by the final round of tabulation – even if first-choice support is split between several candidates. Via cross-endorsing and joint campaigning, candidates can send a clear message to their supporters about how they align, and increase the chances that someone of their ideological or demographic group gets elected. That said, many coalition-building RCV strategies do not require a formal cross-endorsement; candidates may actively seek the support of voters ranking a competitor by simply highlighting areas where the two candidates share common ground. 

This report presents a series of case studies about where candidates have cross-endorsed and/or campaigned jointly under RCV, and shows how candidates can benefit from marketing themselves as a second choice. 

Case Studies

Campaigning together was a winning strategy in Portland

Portland, OR’s City Council District 3 race (2024)

In 2024, Portland, OR used the proportional form of RCV for the first time. Proportional RCV works similarly to single-winner RCV, except that it elects multiple candidates per district. For example, Portland elects three city councilors per district, and each needs 25% of the vote to win. 

In Portland’s City Council District 3 race, several candidates indicated that they would rank each other. Notably, Angelita Morillo and Tiffany Koyama Lane campaigned together and encouraged their supporters to rank them first and second. At a candidate forum, Morillo told Koyama Lane, “I keep a kinship with you as another woman of color and… the values that you carry.” Reflecting on their joint campaign, Koyama Lane said, “by the end, we were door-knocking for each other, and our volunteers were doing the same.” 

Come Election Day, Morillo and Koyama Lane were both elected. The pair were in second and third place (respectively) in each round of counting, so neither received transfer ballots from the other’s supporters. However, their collaborative campaigning styles may have helped them win. 

Moreover, ranked ballot data reveals that many voters listened to the pair’s cross-endorsement. Morillo and Koyama Lane were both the most popular second choice among voters who ranked the other first. 43% of Morillo voters ranked Koyama Lane second, and 45% of Koyama Lane voters ranked Morillo second.

Republicans benefited from “rank the red” messaging in Alaska

Alaska U.S. House of Representatives general election (2022)

In 2022, a special U.S. congressional election marked the first use of Alaska’s new voting system, consisting of an open, all-candidate primary and an RCV general election. Republicans Sarah Palin and Nick Begich, and Democrat Mary Peltola, competed in both the special and regular general elections. Leading up to the special election, Palin discouraged voters from ranking other candidates. Palin condemned Begich as a “Republican in name only” and “full of bull,” and instructed her voters not to rank multiple candidates.  

Peltola received the most first-choice support (40%), leading Palin by nine points. Begich was eliminated, and when his supporters’ second choices were counted, enough of them (28%) preferred Peltola to Palin to help Peltola pass the 50% threshold. 

The special election loss may have convinced the Republican candidates that something had to change. Ahead of the regular general election, Palin told voters that RCV “is the rule of the road now. We have ranked choice voting, and… we have to rank the red” – meaning that Republican voters should rank Palin and Begich first and second. Begich also encouraged his voters to rank the red. 

The repeat matchup for the general election allows us to analyze whether their efforts made a difference. Once again, Begich was eliminated in the RCV tabulation, but this time, 67% of his supporters’ ballots transferred to Palin (compared to 50% in the special election) and 11% went to Peltola (compared to 28% in the special election). In other words, voters appeared to follow the “rank the red” directive. In addition, the share of Palin voters who ranked a second choice rose by 4 percentage points between elections.

Ultimately, Peltola won the general election. Peltola grew her first-choice support by about 9 percentage points, and still had enough support from Begich’s voters to push her over the 50% threshold. This is not an atypical result given Alaska voters’ reputation for independence and cross-partisanship. 

However, this example shows that voters listen to cross-endorsements. Simple “rank the red” messaging was enough to help Palin win over 16,000 more transferred ballots than she did in the special election. If Palin had been a more popular first choice, or if she and Begich had worked together more rigorously, the outcome might have been different. 

In 2024, Begich defeated Peltola in a rematch for the seat, and currently represents the district.

Andrew Yang’s cross-endorsement helped Kathryn Garcia win 33% more ballots in NYC

New York City Democratic mayoral primary (2021)

In 2021, New York City mayoral candidates Andrew Yang and Kathryn Garcia launched a series of joint campaign events the weekend before Election Day. Voters could see Yang shuttling from Queens to Manhattan in Garcia’s campaign van. Their partnership demonstrated collaboration in a city famous for bare-knuckle politics. Yang used the joint campaigning as an opportunity to endorse Garcia as a second choice. Garcia, however, did not reciprocate:

“Let me be very clear, I’m not co-endorsing. We are campaigning together. We are promoting ranked choice voting.” – Kathryn Garcia

One reason she may have refused to cross-endorse was her strong poll numbers; as one of the highest-polling candidates, Garcia knew that her supporters’ backup choices likely wouldn’t come into play. Yang, on the other hand, would likely be eliminated based on his lower poll numbers.

Yang’s endorsement proved to be consequential; when he was eliminated, 45% of his ballots transferred to Garcia – helping Garcia win 33% more ballots than she otherwise would have. These votes saved Garcia from elimination in the seventh round, where she edged out Maya Wiley by 1.4 percentage points. In the next round, Garcia lost to Eric Adams by less than one point.

This election provides empirical evidence that voters listen to cross-endorsements.¹ The day the pair began campaigning together, polls indicated that 47% of Yang’s ballots would transfer to Adams and only 34% to Garcia. In the actual election, 39% of Yang’s ballots transferred to Adams, and 45% went to Garcia. Meanwhile, Wiley’s second-choice support among Yang voters hardly changed between the polls and election; it increased from 16% to 18%. This suggests that just a weekend of joint campaigning was enough to convince more of Yang’s voters to choose Garcia as a backup than Adams. 

On Election Day, Garcia was indeed the most popular second choice among Yang supporters; 25% of Yang’s first-choice supporters ranked Garcia second, and 37% of Yang’s first-choice supporters ranked Garcia in their top three. 

Interestingly, Yang did not benefit as much as Garcia. Only 14% of Garcia’s first-choice supporters ranked Yang second, and 27% ranked him in their top three. Two other candidates – Scott Stringer and Maya Wiley – were a more popular second choice among Garcia’s voters than Yang. Perhaps this is because Garcia did not explicitly endorse Yang in the same manner that he did for her. 

Though Garcia did not win, the pair demonstrated how cooperation and coalition-building are beneficial under RCV.

Cross-endorsement likely led to Chinese American representation in San Francisco

San Francisco Board of Supervisors District 1 election (2020)

In San Francisco’s 2020 Board of Supervisors District 1 election, candidates Connie Chan and David Lee cross-endorsed each other and campaigned together, which ultimately helped Chan win. Both credited their partnership to their dedication to Chinese American constituents, who make up 45% of District 1.

“We have one thing in common, that one thing is that we both want to fight for our Chinese American community.” – Connie Chan and David Lee

Both Chan and Lee came to the race with a long history of advocating for Chinese Americans. Chan, who emigrated from Hong Kong at 13, began her career as an interpreter with the San Francisco Bar Association’s Volunteer Legal Outreach, and as a community organizer in the Bay Area. Meanwhile, Lee served for twenty years as executive director of the Chinese American Voters Education Committee. 

Ultimately, Chan secured the seat. Chan received the most first-choice support – perhaps aided by a rare endorsement from the Democratic Party, as well as an endorsement from then-Supervisor Sandra Fewer. When the six-way race came down to Chan, Marjan Philhour, and Lee, Lee was eliminated in the penultimate round. 39% of Lee’s ballots transferred to Chan, granting her a 125-vote victory over Philhour. 
Of note, this actually means that more of Lee’s supporters preferred Philhour to Chan as a backup choice. At the same time, Lee and Chan’s joint campaigning was almost certainly decisive in a race decided by less than half a point.

Indeed, a significant number of Lee’s first-choice supporters ranked Chan as a backup choice. Almost half (47%) ranked Chan in their top three; 30% followed Lee’s advice and ranked Chan second. A significant chunk of Chan’s first-choice supporters ranked Lee, too: 45% ranked Lee in their top three, and 28% ranked him second. RCV allowed Lee’s supporters to vote for him without worrying about “spoilers,” and without forfeiting the opportunity to weigh in between frontrunners Chan and Philhour.  

This election is a reminder of both the value and limitations of cross-endorsements. These offer just one more piece of information for voters who are considering the candidates in a race. In a close race, it can be enough to make a difference – even if only a minority of voters seem to follow the candidates’ cues. 

An all-female slate campaigned in Oakland

Oakland’s City Council District 6 election (2018)

In 2018, the Oakland City Council District 6 race saw a slate of female candidates campaign together. 

Candidates Natasha Middleton, Marlo Rodriguez, and Mya Whitaker –  all running against mayor-endorsed incumbent Desley Brooks – were united by their interest in revamping a struggling Oakland. The bloc put out a press release a month before Election Day encouraging voters to rank the three of them. The trio stated:

“We are united in our belief that District 6 residents need a change in leadership at City Hall. It’s time for a new District 6 councilmember. And we are united in our confidence that each of us is qualified to serve and would best represent the district on the issues that matter most to District 6 residents.” 

Ultimately, none of the three candidates won the seat. Although they were the bottom three finishers, their voters were aligned and consolidated round-by-round behind the stronger candidate(s). In other words, these candidates were able to pool their support rather than split it. 

During the RCV tabulation, Mya Whitaker – the last of the three eliminated from contention –   closed the gap with incumbent Brooks from 11 points to 5 points. 36% of Whitaker’s vote total came from transferred ballots, compared to 8% for Brooks and 15% for fellow candidate Loren Taylor. 

This example demonstrates how underdog candidates can perform better under RCV than they might have under single-choice voting, specifically by avoiding vote-splitting. RCV allowed the slate’s supporters to rank their favorite candidate first, and also support the coalition as a whole via backup choices. This may have helped those candidates perform better than they would have under single-choice voting.

Recommendations and conclusion

The question of whether (or how much) candidates benefit from cross-endorsements is difficult to answer empirically – there are too many factors that influence voting behavior to isolate the impact. For example, apart from the Yang/Garcia and Begich/Palin examples, we do not know how voters might have ranked the candidates prior to cross-endorsement. In addition, candidates who cross-endorse are more likely to be ideologically similar, so voters within their ideological bloc may be inclined to rank them #1 and #2 regardless of the cross-endorsement. 

In each of our case studies, however, many voters appear to have followed the guidance of candidates who cross-endorsed and/or campaigned together. This may have helped some candidates win, or at least improve their performance.  

Kathryn Garcia received more transferred ballots than expected after a weekend of campaigning with Andrew Yang. Sarah Palin won 67% of Nick Begich’s supporters’ transfer ballots after lightly embracing “rank the red” messaging – 17 percentage points more than she did in their first matchup. Meanwhile, Connie Chan’s 125-vote margin suggests that winning 39% of ballots transferred from David Lee was critical to her victory, and may have been fueled by their cross-endorsement.

To candidates running in RCV races, the message is: Cross-endorsing and/or joint campaigning is one way of coalition-building and performance boosting. It can provide a useful cue to supporters, and help someone from your ideological or demographic bloc get elected. Our analysis shows that voters listen to cues from candidates, making it likely that voters will heed your cross-endorsement. At the very least, it will not hurt your chances. 

There are, of course, many coalition-building RCV strategies that do not require a formal cross-endorsement. Most commonly, candidates may actively seek the support of voters ranking a competitor first by highlighting the areas where the two candidates share common ground. 

As a final note, cross-endorsements and joint campaigns only regularly happen in RCV races.² For candidates and voters to reap the benefits of cross-endorsements (e.g. civil campaigning, better representation), policymakers must first adopt RCV in their jurisdiction.

Footnotes

1. For this case study, we can look at how polling changed after joint campaigning began. We do not have this data for the other case studies.
2.  This refers to single-winner contests. Slates sometimes form multi-winner elections with first-past-the-post voting. They help a candidate campaign to a wider range of voters and influence whether other winners represent their ideology. However, multi-winner RCV elections carry more incentives than first-past-the-post multi-winner elections, since zero-sum situations still appear in multi-winner races. Imagine a two-winner race where a political opponent is all but guaranteed to win one seat. Plurality winner elections discourage joint campaigning, since endorsing an opponent takes away from your support for the only available seat. RCV elections, by contrast, let candidates cross-endorse without compromising their chances.

About the authors

Ben Fitzgerald was a Research and Policy intern with FairVote. He worked with FairVote through a partnership with Haverford College’s Center for Peace and Global Citizenship (CPGC). He is currently pursuing majors in philosophy and computer science at Haverford College.

Rachel Hutchinson is a Senior Policy Analyst at FairVote. Rachel conducts data and policy analysis in support of FairVote’s advocacy mission. She is an experienced researcher and writer in the electoral reform space. Her policy expertise has been leveraged to service state partners and draft legislation, as well as featured in outlets like Real Clear Politics and NPR.