UHERO Insights

What Ranked-Choice Voting Would Mean for Honolulu Elections

October 16, 2026

Colin Moore Insights

INSIGHTS ARE PRELIMINARY MATERIALS CIRCULATED TO STIMULATE DISCUSSION AND CRITICAL COMMENT.
The findings and recommendations are those of the author(s) and do not reflect the views or official positions of UHERO, the University of Hawai’i, or any funding or supporting organization.

By Colin Moore

@misc{moore2026,
  author  = {Colin Moore},
  title   = {What Ranked-Choice Voting Would Mean for Honolulu Elections},
  year    = {2026},
  month   = {October},
  howpublished = {\url{https://uhero.hawaii.edu/what-ranked-choice-voting-would-mean-for-honolulu-elections/}},
  note        = {UHERO},
  url     = {https://uhero.hawaii.edu/what-ranked-choice-voting-would-mean-for-honolulu-elections/}
}

Key Takeaways

Honolulu voters will soon decide on Question 5, a proposed Charter amendment that would change how the City elects its mayor, prosecuting attorney, and City Council members by introducing ranked-choice voting (RCV).

The proposal would shift the final decision in all contested City elections to November. Under the current system, candidates can win outright in August or advance to a November runoff. Under Question 5, races with two to five candidates would go directly to the November ballot. If six or more candidates run, an August election would reduce the field to five candidates for November.

In Honolulu, turnout is substantially higher in November than in August. Across the five election cycles from 2016 through 2024, turnout averaged about 40 percent of registered voters in August and 58 percent in November. Research on election timing also suggests that voters in November general elections tend to more closely resemble the eligible voting population in terms of race, age, and partisanship.

RCV would change how voters choose among candidates, but Honolulu already uses a runoff designed to produce a majority winner. The current runoff system and the proposed RCV system address many of the same problems, including vote splitting and the possibility of electing a candidate without broad support. The main difference is when voters express their preferences. Under the current system, August voters narrow the field and November voters choose between the finalists. Under the Charter proposal, November voters would rank several candidates on the same ballot.

Research suggests that the effects of RCV itself are likely to be modest. Studies find some evidence that RCV encourages candidates to seek support beyond their core constituencies and that voters perceive campaigns as less negative. Evidence on turnout attributable specifically to RCV is mixed, and scholars do not consistently find large or lasting changes in who runs for office or in policy outcomes. Because Honolulu already uses a runoff rather than simple plurality voting, the differences produced by RCV may be smaller than much of the national debate suggests.

Voter education would be important for effective implementation. Most voters in American jurisdictions using RCV report understanding how the system works, but citizens would face more candidates to evaluate and a ballot with more opportunities for confusion or mismarking. Clear ballot design, multilingual voter education, and public outreach would help voters navigate these changes.

Why Honolulu Is Considering Ranked-Choice Voting

Honolulu voters will decide this November whether to change the way the City runs its nonpartisan elections for mayor, prosecuting attorney, and City Council. Although the proposed charter amendment is usually described as a ranked-choice voting measure, it would also change when most City elections are decided. Under the current system, some races are decided in the August election, while others advance to a November runoff between the top two candidates. Question 5 would make November the final election in every contested race.

The 2024 mayoral race illustrates the difference. Four candidates ran, and Rick Blangiardi won a majority in August, ending the contest before the November general election. Under Question 5, there would have been no August mayoral election. Instead, all four candidates would have appeared on the November ballot, where voters would have ranked them in order of preference (State of Hawai‘i Office of Elections, 2024).

Ranked-choice voting (RCV) provides a way to make that kind of November election possible while still seeking a majority winner. If three, four, or five candidates appeared on the November ballot under a simple plurality system, the candidate with the most votes could win with far less than 50 percent. A traditional runoff could prevent that outcome, but holding one after November would require an additional election. RCV instead allows voters to rank the candidates and uses those rankings, if necessary, to determine which candidate receives majority support among active ballots (Eggers & Bouton, 2024).

RCV is already used statewide in Alaska and Maine and in cities including San Francisco, Oakland, and Portland, Oregon. New York City uses it in municipal primaries. Hawai‘i law also requires RCV for federal elections held outside the regularly scheduled primary or general election and for special elections to fill vacant county council seats (National Conference of State Legislatures, 2026; HRS § 11-100).

What Question 5 Would Change

The precise election procedure would depend on how many candidates file for an office. Under Honolulu’s current rules, a candidate is elected in August if only one person files or if one candidate receives a majority in a race with three or more candidates. If there are two candidates, both proceed directly to November; if three or more run and no one wins an August majority, the top two advance to November (City and County of Honolulu, 2025, § 13-116).

As Table 1 shows, Question 5 would replace that system with different rules based on the size of the field. An uncontested candidate would still be elected in August. With two to five candidates, all would proceed directly to a ranked-choice election in November. With six or more candidates, an August ranked-choice election would narrow the field to five for the November election (Honolulu Charter Commission, 2026).

The proposal applies to the mayor, prosecuting attorney, and all nine City Council seats. It could take effect no earlier than 2032 and only after state law authorizes counties to use RCV in county elections.

How Ranked-Choice Voting Works

In a November ranked-choice election, voters could rank up to five candidates in order of preference. They would not have to use all five rankings. A voter could rank all five candidates, rank only two or three, or vote for just one. Voters could stop ranking candidates at any point without invalidating the ballot, although the ballot would stop being counted if all of the candidates they ranked were eliminated.

The count begins with voters’ first choices. If no candidate has a majority, the candidate with the fewest votes is eliminated. Ballots counting for that candidate then count for each voter’s next-ranked candidate still in the race, if there is one. The process continues until a winner is determined. In every round, each person’s vote goes to their highest-ranked candidate who is still in the race.

Figure 1 illustrates the process using a hypothetical five-candidate election. Maya begins with the most first-choice votes but not a majority. As candidates with fewer votes are eliminated, the preferences of voters who initially supported those candidates come into play, and Alex eventually reaches 51 votes and wins.

Figure 1. How votes move in a ranked-choice election

Hypothetical election with 100 voters. When no candidate receives a majority, the candidate with the fewest votes is eliminated and ballots for that candidate transfer to each voter’s next-ranked candidate still in the race.

The example also demonstrates an important feature of RCV. The candidate with the most first-choice votes—in this case, Maya—does not necessarily win. Alex wins by combining first-choice support with support from voters whose preferred candidates were eliminated. Each ballot counts for only one candidate in each round, but the ability to rank multiple candidates allows those voters to help determine the final outcome.

Candidate Choice and the Runoff Comparison

Much of the national debate over ranked-choice voting compares it with plurality elections, where the candidate with the most votes wins even without a majority. Honolulu, however, already uses a runoff system designed to produce a majority winner. Because RCV addresses many of the same concerns, adopting it would represent a smaller change here than in a city using plurality voting (Eggers & Bouton, 2024).

Vote splitting and strategic voting

Both RCV and traditional runoffs help address a problem known as vote splitting. This occurs when candidates appealing to similar voters divide their support, allowing another candidate to win. Suppose Candidates A and B receive 30 percent each, while Candidate C receives 40 percent. Under plurality voting, C wins—even if everyone who supported A or B would prefer either of them to C. A traditional runoff addresses this problem through a separate election, giving voters another opportunity to choose between the two finalists. RCV instead uses voters’ rankings to make that choice without a second election. If a voter’s preferred candidate is eliminated, their vote counts for their next-ranked candidate who remains in the race, if there is one.

RCV also reduces the pressure to vote strategically compared with plurality voting. Someone who prefers Candidate A but believes Candidate B has a better chance of defeating Candidate C can rank A first and B second rather than deciding whether to vote for B instead. If A is eliminated while B remains in the race, the voter’s ballot then counts for B. This gives voters less reason to abandon their first choice simply because that candidate seems unlikely to win. Strategic voting does not disappear under RCV, but voters generally have less to gain from it than under plurality voting, particularly when they are uncertain about how competitive the candidates are (Eggers & Nowacki, 2024).

The proposed Charter amendment would change how and when these choices are made. In races with three or more candidates, the current system allows August voters to decide the outcome or narrow the field to two finalists. Under the proposal, November voters would choose from as many as five candidates by ranking them on a single ballot.

Choosing between the finalists

When a race advances to November, Honolulu’s current runoff system gives voters additional time to evaluate the two finalists. That opportunity may be especially useful when the original field is large or voters know relatively little about the candidates (Eggers & Bouton, 2024). Under the proposed RCV system, voters would also have more time to evaluate candidates in races that could otherwise end in August, because the decision would move to November. They would, however, be evaluating a broader field rather than focusing on two finalists.

Under RCV, voters express their preferences among possible finalists in advance by ranking candidates. Their ballot continues to count as long as someone they ranked remains in the race. Ranking all candidates keeps the ballot active through the final round, but voters may leave some unranked. If all their ranked candidates are eliminated, the ballot will have counted in earlier rounds but will no longer influence the outcome. This is known as ballot exhaustion and has been documented in local RCV elections (Burnett & Kogan, 2015).

Under either system, the candidate who leads the initial vote count can ultimately lose as supporters of eliminated candidates help decide the outcome. But voters may view such a result differently depending on the voting system. In a survey experiment, Cerrone and McClintock (2023) found that come-from-behind victories reduced voter satisfaction more under RCV than under a traditional runoff.

Moving the Decision to November

One of the most direct effects of Question 5 would be moving the final decision in contested City elections to November. As Table 2 shows, turnout across the five election cycles from 2016 through 2024 averaged about 40 percent of Honolulu’s registered voters in August and 58 percent in November (State of Hawai‘i Office of Elections, 2016–2024).

These figures do not tell us how many additional votes the proposal would produce in a particular City race. Council districts cover only portions of the county, and some voters skip local contests. They nevertheless suggest that more residents would participate in the final choice of City officials if all contested races were decided in November.

Research on election timing provides strong evidence that when elections are held affects both turnout and the composition of the electorate. Studies of local elections generally find higher turnout when elections are held alongside higher-profile state and national contests, in part because voters are already mobilized and the additional costs of participating in a local election are lower (Berry & Gersen, 2011; Hajnal et al., 2022). Hajnal et al. (2022) also find that on-cycle local elections tend to produce electorates that more closely resemble the eligible population in terms of race, age, and partisanship. That research is generally consistent with Honolulu’s turnout pattern, although the comparison is not exact. Honolulu’s August election already coincides with Hawai‘i’s statewide primary, so studies comparing November elections with stand-alone municipal elections may overstate the likely effect here.

The 2012 Honolulu mayoral election illustrates what this change could mean. Former governor Ben Cayetano led the August election with 44.1 percent of the vote, followed by Kirk Caldwell with 29.1 percent and incumbent mayor Peter Carlisle with 24.8 percent. A fourth, relatively unknown candidate, Khistina Dejean, received just 0.6 percent. Carlisle finished 8,862 votes behind Caldwell, and both Carlisle and Dejean were eliminated from the race. November drew 92,305 more voters countywide than August, but only Cayetano and Caldwell appeared on the runoff ballot—and Caldwell ultimately won (State of Hawai‘i Office of Elections, 2012a, 2012b). Under Question 5, all four original candidates would instead have appeared on the November ranked-choice ballot. We cannot know whether the outcome would have changed, but the larger November electorate would have had the opportunity to consider the full field, including Carlisle.

What a larger electorate would mean for election outcomes more generally is less clear. Candidates in contested races who can now win in August would instead have to campaign through November and reach a broader electorate, including many voters who do not typically participate in the primary. That shift does not by itself tell us which candidates would benefit, and its effects on public policy are even more uncertain. Berry and Gersen (2011), for example, find large effects of election timing on turnout but much smaller effects on the policy outcomes they examine.

What Does the Research Tell Us About Ranked-Choice Voting?

The effects of moving elections to November are somewhat easier to anticipate than those of adopting ranked-choice voting itself. Research from other jurisdictions offers evidence on how RCV affects campaigns, who runs and wins, and voter participation. The findings vary across settings, but they help to clarify what Honolulu might expect and where implementation would require particular attention.

Campaigns and candidates

RCV can modestly change campaign incentives by giving candidates a reason to engage voters beyond their core supporters, since second- and third-choice rankings can matter. Studies also find some evidence that voters perceive RCV campaigns as less negative (Donovan et al., 2016; Donovan & Tolbert, 2024). How much these incentives affect campaign behavior likely depends on whether second-choice support is genuinely contestable. When transfers are largely predetermined by partisan, ethnic, or cultural alignments, the incentive to make broader appeals may weaken or even reverse (Buisseret & Prato, 2026).

Evidence that RCV substantially changes who runs for office is also limited. Colner (2025), comparing American cities with and without RCV over three decades, finds an initial increase in candidates after adoption, but the effect fades over time and is concentrated among candidates who receive relatively little support. The study finds no consistent increase in the share of women or non-White candidates. A separate survey experiment finds little evidence that RCV itself changes voters’ support for candidates of color (Crowder-Meyer et al., 2024).

There is, however, some evidence that reforms incorporating RCV can affect which candidates win. A recent study of Alaska’s combined Top-4 primary and RCV system finds that moderate candidates were more likely to win after the reform, and the authors link those electoral changes to more bipartisan legislative coalitions and more moderate policy outcomes. But because Alaska changed both its primary system and its general-election voting rules at the same time, the study cannot isolate the effect of RCV itself (Wright et al., 2025).

Turnout

Whether RCV on its own increases turnout is much less clear than the effect of moving elections to November. An earlier study of San Francisco mayoral elections found lower turnout under RCV among Black and White voters, along with larger disparities by age and education (McDaniel, 2016). More recent studies using individual-level voter records find that living in an RCV jurisdiction is associated with higher turnout and find no evidence of lower turnout among voters of color, although these observational comparisons do not establish that RCV itself caused the differences (Dowling et al., 2024, 2025).

Voter understanding and ballot errors

Most voters appear able to use ranked ballots successfully, and surveys from New York City find high levels of reported understanding of RCV (Donovan et al., 2022). But voters do not all make use of the ballot in the same way. Across studies, voters with more education are more likely to report understanding how the system works and, in some cases, to rank multiple candidates (Donovan et al., 2019, 2022). Research from New York City similarly finds that voters who understand the instructions and receive encouragement from campaigns to rank additional candidates are more likely to do so (Wendland & Carman, 2023).

Ballot errors occur in conventional elections as well as under RCV. The relevant question is whether ranked ballots change how often errors occur, whose ballots are affected, and whether those errors prevent votes from being counted. Survey evidence documents some voter confusion, while studies of actual ranked ballots find higher rates of mismarking in areas with more lower-income households and lower levels of educational attainment (Atkeson et al., 2024; Pettigrew & Radley, 2026). These mistakes do not necessarily invalidate a ballot, but they can prevent some rankings from being counted or cause a ballot to become inactive before the final round. There is, however, evidence that voters may become more accustomed to the system over time. In New York City, the share of ballots rejected as invalid declined substantially between the 2021 and 2025 citywide RCV elections, although disparities associated with education persisted (Cormack, 2026).

For Honolulu, clear instructions and voter education would be important for helping residents use the new ballot successfully. The proposed amendment reflects that need by requiring the city clerk to provide multilingual voter education, public workshops, and online resources, alongside rules governing ballot design and tabulation (Honolulu Charter Commission, 2026).

The Practical Effects of Question 5

The research gives us more confidence about certain effects of Question 5 than others. Like Honolulu’s current runoff, RCV would help address vote splitting and could give candidates an incentive to seek support beyond their strongest supporters. It would also ask voters to use a more detailed ballot, with more opportunities for marking errors, and require election officials to administer a more complex counting process.

The larger change would come from moving the final decision in contested City races to November. In races with three to five candidates, all would appear on the November ballot, rather than having August voters elect a winner or narrow the field to two finalists. The decision would take place during an election in which substantially more Honolulu residents participate. Evidence that RCV itself produces large changes in who runs, campaign tone, or what governments ultimately do is considerably less certain (Donovan & Tolbert, 2024; Colner, 2025; Vishwanath, 2025).

For Honolulu, Question 5 would give November voters more choice among candidates and would likely increase participation in electing City officials. Making that system work well would depend on effective voter education, clear ballot design, and careful election administration.

Sources

Atkeson, L. R., McKown-Dawson, E., Santucci, J., & Saunders, K. L. (2024). The impact of voter confusion in ranked choice voting. Social Science Quarterly, 105(4), 1029–1041.

Berry, C. R., & Gersen, J. E. (2011). Election timing and public policy. Quarterly Journal of Political Science, 6(2), 103–135.

Buisseret, P., & Prato, C. (2026). Politics transformed? Electoral competition under ranked choice voting. American Journal of Political Science [online].

Burnett, C. M., & Kogan, V. (2015). Ballot (and voter) “exhaustion” under instant runoff voting: An examination of four ranked-choice elections. Electoral Studies, 37, 41–49.

Cerrone, J., & McClintock, C. (2023). Come-from-behind victories under ranked-choice voting and runoff: The impact on voter satisfaction. Politics & Policy, 51(4), 569–587.

City and County of Honolulu. (2025, January 16). Revised Charter of the City and County of Honolulu 1973 (amended 2017 edition). Department of the Corporation Counsel.

Colner, J. (2025). Running toward rankings: Ranked choice voting’s impact on candidate entry and descriptive representation. American Journal of Political Science, 69(3), 1010–1028.

Cormack, L. (2026). More chances, fewer problems? Revisiting ranked choice voting errors in New York City. American Politics Research, 54(3), 340–346.

Crowder-Meyer, M., Gadarian, S. K., & Trounstine, J. (2024). Ranking candidates in local elections: Neither panacea nor catastrophe for candidates of color. Journal of Experimental Political Science, 11, 117–134.

Donovan, T., Tolbert, C., & Gracey, K. (2016). Campaign civility under preferential and plurality voting. Electoral Studies, 42, 157–163.

Donovan, T., Tolbert, C., & Gracey, K. (2019). Self-reported understanding of ranked-choice voting. Social Science Quarterly, 100(5), 1768–1776.

Donovan, T., Tolbert, C., & Harper, S. (2022). Demographic differences in understanding and utilization of ranked choice voting. Social Science Quarterly, 103(7), 1539–1550.

Donovan, T., & Tolbert, C. (2024). Civility in ranked-choice voting elections: Does evidence fit the normative narrative? Representation, 60(4), 583–600.

Dowling, E., Tolbert, C., Micatka, N. K., & Donovan, T. (2024). Does ranked choice voting increase voter turnout and mobilization? Electoral Studies, 90, 102816.

Dowling, E., Tolbert, C., Micatka, N. K., & Donovan, T. (2025). Is ranked-choice voting associated with turnout across race/ethnic groups? Social Science Quarterly, 106, e70025.

Eggers, A. C., & Bouton, L. (2024). Ranked-choice voting: A practical research guide. Center for Effective Government, University of Chicago Harris School of Public Policy.

Eggers, A. C., & Nowacki, T. (2024). Susceptibility to strategic voting: A comparison of plurality and instant-runoff elections. The Journal of Politics, 86(2), 521–534.

Hajnal, Z. L., Kogan, V., & Markarian, G. A. (2022). Who votes: City election timing and voter composition. American Political Science Review, 116(1), 374–383.

Haw. Rev. Stat. § 11-100. Ranked-choice voting; application; procedure.

Honolulu Charter Commission. (2026, July 20). P130 CD3 (Question 5): Final amendment proposal. City and County of Honolulu.

McDaniel, J. A. (2016). Writing the rules to rank the candidates: Examining the impact of instant-runoff voting on racial group turnout in San Francisco mayoral elections. Journal of Urban Affairs, 38(3), 387–408.

National Conference of State Legislatures. (2026, March 23). Ranked choice voting. https://www.ncsl.org/elections-and-campaigns/ranked-choice-voting

Pettigrew, S., & Radley, D. (2026). Overvotes, overranks, and skips: Mismarked and rejected votes in ranked choice voting. Political Behavior, 48, 229–251.

State of Hawai‘i Office of Elections. (2012a). Primary election 2012: City and County of Honolulu final summary report. https://files.hawaii.gov/elections/files/results/2012/primary/cch.pdf

State of Hawai‘i Office of Elections. (2012b). General election 2012: City and County of Honolulu final summary report. https://files.hawaii.gov/elections/files/results/2012/general/cch.pdf

State of Hawai‘i Office of Elections. (2024). Primary election 2024: Statewide summary report. State of Hawai‘i.

State of Hawai‘i Office of Elections. (n.d.). Registration and turnout statistics. https://elections.hawaii.gov/resources/registration-voter-turnout-statistics/

Vishwanath, A. (2025). The effects of ranked choice voting on substantive representation. Quarterly Journal of Political Science, 20(3), 409–437.

Wendland, J., & Carman, E. (2023). New or “normal” election? Understanding ranking activity in New York City’s first ranked choice voting election. Social Science Quarterly, 104(4), 591–604.

Wright, G., Reilly, B., & Lublin, D. (2025). Assessing Alaska’s Top-4 primary and ranked choice voting electoral reform: More moderate winners, more moderate policy. Journal of Political Institutions and Political Economy, 6, 59–84.