Steven Hill
Editor's Note: The following commentary originally published on the DemocracySOS Substack and has been republished on IVN with permission from the author.
Ranked ballots are one of the wonders of modern representative democracy because they solve real democratic dilemmas that other methods can’t solve. Even a Party List Proportional method offers a voter only a single-choice and thereby replicates one of the worst defects of the truly antiquated “plurality wins all” method. Here’s what I mean.
In this previous DemocracySOS article, we reviewed the recent Democratic primaries for governor in Wisconsin and US Senate in Michigan, in which both winning candidates prevailed by an eyelash with less than a majority of the popular vote. In the Wisconsin governor’s race, the winner had only 39.8 percent of the popular vote, meaning three out of every five Democratic voters preferred another candidate. If the supporters of the other candidates in these two races had been allowed to cast a runoff vote to determine the majority winner, a different candidate might have won. In truth, in both races the Democrats can’t be sure that the correct candidates – the candidates preferred by the most voters – actually prevailed.
Moreover, a large chunk of the Democratic Party voters wasted their votes on minor spoiler Democratic candidates who had no chance of winning – including some “zombie” candidates who had dropped out and were no longer even running for the nomination! Yet those voters couldn’t change their votes to a continuing candidate. In the 2020 Democratic presidential primary, more than 3 million voters nationwide cast ballots for candidates who had withdrawn from the race; millions more wasted their vote on a candidate who didn’t reach the 15% threshold used in Democrat presidential primaries to qualify for being awarded delegates. Zombie votes, wasted votes, spoiler candidates, non-majority winners…this is what the Democrats served up in these primaries, which not only is frustrating for voters but also undermines the legitimacy of the non-majority winners and undercuts their ability to beat the Republican candidates in the November general election.
If, instead of the “pick-one” plurality system, the Democrats had used an “instant runoff,” otherwise known as “ranked choice voting for single winners” which gives voters the option of ranking their favorite candidates and elect the candidate with majority support, all of these completely unnecessary negative defects – frustrating to candidates, parties and voters alike – would have been avoided. With IRV/RCV, if a voter’s first choice can’t win their vote automatically migrates to the voter’s second choice, and then their third choice, and so on.
Or look at the mess of our presidential elections. Since 1992, five out of nine presidential elections saw a candidate win with less than a popular majority. In two of those, the winning candidate had fewer popular votes than the loser. In 2016, if you added up all the votes of center-left candidates, including Democrat Hillary Clinton (48.2%), Green Party candidate Jill Stein (1.1%), add a good chunk of the Libertarian candidate Gary Johnson (3.3%) (since he said he wanted “nothing to do with the GOP”), and Evan McMullin (0.5%) who was a former Republican “Never Trumper,” Hillary easily would have crossed the 50% threshold and prevailed in enough states to win the electoral college vote. The same happened in 2000, with the Gore-Nader center-left split allowing George W. Bush to win with only 47.9% of the national popular vote. Using instant runoff voting within the 50 states would have changed the outcome of those two presidential elections and elected majority winners. Think of how history would have changed. No Trump, no Bush.
Or how about the 1992 victory for Democrat Bill Clinton with only 43% while independent Ross Perot had 18.9% and the GOP’s George HW Bush pulled 37.5%. If the Perot voters had been allowed to select an “instant runoff” choice, would Clinton have won? Bush and Clinton would have had to woo Perot supporters throughout the campaign; who knows which candidate would have done that most successfully.
We accept these bizarre election results as “normal,” as “just the way it is,” but it’s really a national travesty. We elect the most powerful political leader in the world, and in most presidential elections over the last 30 years our defective electoral method was not competent enough to ensure that the most popular candidate – the one preferred by a popular majority – actually won. If we saw this happen in another country on a regular basis, we would call it a Banana Republic.
“Majority winners” should be a sacred, inviolate democratic value
There is a simple operative principle at work here: if a candidate wins an election for a single-seat office and has garnered less than a popular majority, that means you can’t be absolutely 100% certain that the right candidate — the most popular candidate — won. Every time that occurs, it violates key democratic values.
Throughout US history, there have been 19 out of 47 (40%) presidential elections when the winning candidate won the office with less than a majority (50%) of the national popular vote. Out of these, five winners actually received fewer individual popular votes than their direct opponents, while the rest won with a plurality (the most votes but less than a majority). The lowest ever was Abraham Lincoln in 1860, pulling only 39.7% of the national vote in a very split field. But the regional divisions on the cusp of a civil war were so dramatic that Lincoln won 59% of the electoral college vote and the presidency. In 1876, Samuel Tilden won a 50.9% majority of the popular vote, yet lost a very controversial electoral college count – by a single vote -- to Ruther(fraud) B. Hayes.
It’s understandable why Americans in the 19th century put up with bogus election methods that couldn’t guarantee that the most popular candidate would win. Methods like instant runoff voting/ranked choice voting hadn’t been invented yet, and the challenges of election administration and potential corruption required a degree of systemic simplicity that prevented improvement.
But what is our excuse today? Why do we still put up with using an electoral method that sometimes elects the wrong winner and makes voters feel like they can’t vote for their favorite candidates, lest it turns out that vote for a “spoiler” candidate actually contributed to their “greatest evil” candidate winning? A vote for Nader, Perot, etc. helped elect George W Bush and Bill Clinton. But the use of ranked ballots would have changed all that.
No question, ranked ballots are a miracle of modern democracy. Let us count the ways that ranked ballots unequivocally improve our elections and our representative government, adding a unique enhancement that no other electoral system can provide.
1) More voter choice liberated voters.
Voters have the option of ranking their ballots, 1, 2, 3, and if their favorite first choice candidate can’t win, their vote automatically moves to their second choice, then their third choice, etc. What a gift this is – voters no longer have to worry about spoiler candidates or feel forced to vote for the lesser of two evils. Voters no longer have to strategically try to figure out which candidates are more likely to win, and can enjoy unprecedented sincerity in their choosing of candidates. In the 2000 presidential election, Ralph Nader voters could have ranked Al Gore second or third, ensuring that when Nader was eliminated from the tabulation because of too little support, their votes would not have been thrown out because they would have had the option of giving their vote to Al Gore. Gary Johnson’s Libertarian voters would have had the option of casting their runoff vote for Hillary Clinton or Donald Trump…Ross Perot voters for Bill Clinton or George HW Bush. No other major electoral system empowers voters like this. This aspect of ranked ballots offers an unqualified improvement for democracy and representative government.
2) Liberates candidates too.
More candidates are liberated to run without fear of playing spoiler, or without fear of splitting the vote with like-minded candidates. For example, multiple progressive candidates can all run in the same race and instead of competing against each other and splitting the votes of their own base of support, causing them all to lose, now that vote can coalesce around the strongest of their bloc’s candidates. The same is true of multiple moderate or conservative candidates. Like-minded candidates can even campaign together, much as Democrats Zohran Mamdani and Brad Lander did in New York City’s mayoral election, and like women of color candidates have done in Minneapolis, St. Paul and New York City’s city council elections. Previously those candidates were told “wait your turn, if you run you will split our vote.” But with RCV’s ranked ballots, all those candidates could run at the same time and actually support each other. Do you have any idea what a sea level change this is, during a time when politics has become a bitter, hack attack zero-sum game? Research and opinion polls show that voters love this feature about RCV.
3) Incentivizes coalition-building.
The typical “plurality wins all” election encourages candidates to campaign with a strategy of “you against me.” But RCV changes the incentives for how you win. Suddenly you have to attract second and third rankings from the supporters of other candidates. That means you need to be mindful of what you say about those candidates. If you attack them, their supporters will not rank you. Instead, you have to find some degree of common ground. In a San Francisco mayoral election, the two leading progressive candidates co-endorsed each other and did TV ads together, trying to encourage a progressive wave of voters and organizations, instead of “don’t vote for her, vote for me.” In Maine, two Democratic candidates for governor endorsed each other and ran TV ads together (see their ad here). Building a coalition with other candidates in turn results in another crucial feature that voters’ love….
4) Less mudslinging campaigns.
If attacking your opponents may well be a losing strategy, and if finding common ground and building coalitions with other candidates is a winning strategy, then the campaigns themselves become less about slinging mud and personal attacks and more about the issues. In one mayoral election in Oakland using RCV, a leading candidate saw a house with another candidate’s yard sign. In a typical plurality election, this candidate would have skipped that house. Instead, she rang the doorbell and told the homeowner what she had in common with the other candidate, and asked this voter for her second ranking. In Alaska, some candidates told a similar story. Less mudslinging, more coalition building. No other electoral system incentivizes this kind of positive behavior among candidates and the organizations supporting those candidates. We have seen this dynamic play out in election after election.
5) Your vote becomes an expression of your personhood.
Imagine you are a gay Republican. In the current “plurality-wins-all” system, you probably will have to pick whether you want to vote for a gay candidate or a Republican candidate, quite possibly a MAGA candidate. With RCV, you can express who you are – rank a gay independent candidate first, a moderate Republican candidate second, perhaps a gay moderate Democrat third. Or how about if you are a moderate Democrat who supports gun owner’s rights and tighter control at the border. You could rank as your first choice a moderate Democrat with a sensible position on border policy, and a pro-gun moderate Republican candidate second. How about a Green Party supporter who strongly supports the Green New Deal. You could rank first a Green candidate for Congress and rank a progressive Democrat second and a pro-environment Republican third. RCV gives the voter a more expressive vote that allows you to define yourself. You aren’t locked into a narrow choice of candidates who don’t represent your fullness.
Interestingly, there are some political scientists who are supporters of a Party List PR system, and who criticize this “maximum expression” of ranked ballots. Some political scientists call it “vote leakage” (or inter-coalition “vote leakage”) when a voter’s vote transfers from one political party to another. They see this as a dilution of a political party’s ability to hold onto its votes. Instead of this being a feature that liberates voters to pick their favorite candidates and parties rather than the “lesser evil,” to these academics the votes don’t belong to the voters they belong to the political party who has received the primary allegiance of the voter. The idea that many voters are actually complex mixtures of political sympathies, and that in a mature democracy should be able to express that personal blend, is not valued. While I am fully aware of the desirability of fair and proportional representation along party lines, nevertheless I find this view to have anti-voter consequences.
6) More candidates of color elected.
RCV makes it easier for women and candidates of color to run for office and win. With RCV, women and minority candidates aren’t pressured to wait their turn, nor can they be automatically labelled as “spoilers.” Research finds that candidates of color are more likely to win in RCV elections, in part because voters of color tend to rank more candidates than white voters. And we have seen some remarkable results: women of color and people of color majorities have been elected to city councils as distinctive as Minneapolis, St. Paul, Salt Lake City, San Francisco and Oakland. Or look at the miracle of New York City: it’s ridiculous that it has not been more widely reported that, since New York City began using RCV in its 2021 elections, the number of women-of-color elected to the city council has increased from only a handful to 26 out of 51 seats. And the number of women overall has reached 31 out of 51 (61%). Yes, the city council for America’s largest and most dynamic city is majority women-of-color and one of the most racially and gender diverse in the country. New York City has had public financing of campaigns and fusion voting for many years, but not until New Yorkers started using RCV did they see such stunning results.
7) Less polarization.
Because of these features of ranked ballots, including more coalition-building and less mud-slinging, RCV meets the current moment more than other electoral systems. People can come together around the ranked ballots and its full range of choices, around being able to select more than one party or one candidate instead of the same old “you against me,” and not feel trapped into being taken for granted by a single party. PR proponents talk about the importance of “multiparty democracy,” which I agree with, but to the average person “more parties” might well sound like “more conflict.” Because even in a Party List PR system, a certain number of voters would still be stuck in the same old bind of potentially picking a single political party that does not reach the victory threshold and thereby wasting their vote. RCV will result in less polarization and the freedom to pick multiple choices instead of only a single political party or candidate. In states like Alaska, which used to be highly polarized between MAGA Republicans and Democrats, RCV has led to coalition-building within the legislature between moderate Republicans and moderate Democrats, who have formed a cross-partisan governing coalition (see “Alaska – of all places – shows what a bipartisan coalition could look like”). For reasons I truly do not understand, certain political scientists see this kind of bipartisan coalition-building as somehow undermining the cohesion and influence of political parties as vote-maximizing entities.
8) Ranked ballots can be used to elect both majority-winners and to elect proportional representation.
Ranked choice voting comes in two different varieties – for electing single-winner offices, in which it functions as an “instant runoff” to elect majority winners in a single election; or for electing legislatures from multi-seat districts, using the “single transferable vote” method (also known as proportional ranked choice voting, PRCV) to elect political parties or nonpartisan groupings of like-minded voters via a system of proportional representation. By using ranked ballots in all elections – from executive offices like president, governor, mayor (instant runoff) to city councils and legislatures (proportional representation) – that allows voters to learn a uniform method in all elections and to lessen the chances of voter confusion. In regard to which PR method is best for the United States, no other major electoral method empowers voters with all the features of proportional ranked choice voting and ranked ballots, not even Party List PR because…
9) Ranked ballots would significantly improve Party List PR methods
Party List PR (PLPR) democracies would be wise to include ranked ballots in their own elections. Proponents of PLPR apparently have not noticed that one-choice PLPR shares one of the worst defects of plurality elections – if a voter picks wrong with their single choice, her/his vote gets wasted. For example, in Germany during a recent election 34 political parties competed but only five parties crossed the 5% threshold and won seats. Votes for the other 29 parties were wasted, resulting in 16 percent of the List vote going to losing parties. In Germany’s more recent 2025 elections, there were 29 parties and only six of them won seats. Approximately 20 percent of the List vote went to parties that fell short of five percent. Germany could have allowed those voters to rank more than one political party, and if their first choice can’t get elected, their vote would transfer to their next-ranked party. That would reduce the number of wasted votes and increase the proportionality of the elections.
Other List PR democracies, including Italy, Czech Republic, Slovenia, Slovakia and New Zealand have been plagued by similar spoiler dynamics that could be fixed by using ranked ballots. Without ranked ballots in the previous Israeli election using Party List PR, nearly 300,000 anti-Netanyahu votes were wasted on small parties that did not win representation, allowing his rightwing, religious bloc to barely squeak out a majority of legislative seats despite winning less than a majority of the popular vote. So, these kinds of “wasted vote” dynamics can have tragic outcomes. Party List-based democracies should improve their electoral systems by incorporating some version of the Spare Vote, which combines ranked ballots in a List PR system.
Ranked ballots, one of the wonders of modern democracy science
Ranked ballots are one of the wonders of modern democracy science and should be incorporated into nearly every election situation, at federal, state and local levels. Minnesota held a public election using ranked choice voting to choose its new state flag design. RCV should be used to elect names of stadiums or sports team, favorite restaurants, it should be used in opinion polls and in any situation in which you want to receive a true representation of majority voter sentiment without worrying about spoilers, wasted votes and too many voters feeling frustrated and disempowered by picking losers. Once I was with a group of colleagues following a political reform conference, and we selected our pitcher of beer by ranking our choices. Even Party List PR elections, whether “closed” or “open,” which together constitute the mostly widely used electoral system in the world, would be improved by allowing voters to rank their favorite choices.
Ranked ballots are an unequivocal improvement in representative democracy, and it’s puzzling to me why a handful of political scientists like to trash them. Some people believe ranked ballots are complicated, or make the voting process more complicated, but there is no research, based on actual, real-world elections, to support that view. In actual fact, there is a lot of bad anti-RCV research out there using questionable methodologies, as attorney Paul Haughey and I have exposed in our report, “Shaky Political Science.” The fact is, millions of voters in the US, as well as millions more in Australia, Ireland and other places, going back many decades, have been using ranked ballots and have handled the task of ranking candidates as easily as they handle the task of ranking their favorite flavors of ice cream or movies. In exit polls from New York City’s RCV elections in 2025, 96% of voters said their ballot was simple to complete, including at least 94% of each racial group; 76% said they want to keep or expand RCV to other elections in New York City.
Ranked choice voting has established an impressive track record across the country because it delivers better choices, better campaigns, and better governance. It has been used in a thousand elections by over 16 million voters in over 50 US cities, counties and states, at local, state and federal levels. Six Southern states – Alabama, Arkansas, Georgia, Louisiana, Mississippi, and South Carolina – use ranked choice ballots for their military and overseas voters so they can participate in their states’ delayed two-round runoffs, by having them rank their runoff choices at the same time as their first choice. Another magical use of ranked ballots.
If we used RCV on the national level, I have no doubt that it would result in a dramatic improvement in American politics. The attractive features that are on display in RCV elections in dozens of US cities and states are all ones that our national politics desperately need.