Internet Explorer 11 is not supported

For optimal browsing, we recommend Chrome, Firefox or Safari browsers.

The Tempting Panacea of Ranked-Choice Voting

It's gaining in popularity around the country, touted as a way to restore civility and bipartisanship. But it's not a perfect solution, and it doesn't come without costs.

Sample of a ranked-choice ballot from the state of Maine. (Source: Maine Secretary of State Office.)
You may never have heard of Ray Blanton, the late governor of Tennessee, but the story of his political career offers some insight into the current condition of our electoral democracy.

Unfortunately, it isn’t a very pleasant story. Blanton, who served one term from 1975 to 1978, was a criminal. He illegally sold surplus state cars to political allies, extorted money for liquor licenses and at the end of his term took bribes from state prisoners seeking pardons, including 20 convicted murderers. He was hustled out of office by state authorities three days early and went to federal prison for 22 months.

But it isn’t Blanton’s misdeeds that have contemporary relevance. It’s the way he was elected. A four-term congressman and former state legislator of few discernible accomplishments, he entered a 12-candidate Democratic primary for governor and finished first with only 23 percent of the vote. He had little name recognition or support statewide, but he had a loyal constituency in his southwest Tennessee congressional district, where he had spent more than a decade doing favors for constituents. He was running in a landslide Democratic year, so he went on to win the general election easily.

Ray Blanton stands today as Exhibit A for some form of runoff voting. Only Tennessee’s no-runoff election rules allowed the victory of a blatantly unqualified candidate.

Nothing quite like that has happened in recent decades, but there have been cases in which the absence of a runoff procedure has unalterably changed the outcome. George W. Bush won the 2000 presidential election with 47.8 percent of the popular vote, narrowly behind his main opponent, Al Gore. A third-party candidate, Ralph Nader, won nearly 3 percent of the vote. Since Gore was more liberal than Bush, and Nader was to the left of Gore, a provision requiring a runoff between the top two candidates would likely have thrown most of Nader’s vote to Gore and handed him the election.

Donald Trump won the Republican presidential nomination in 2016 with 44 percent of the primary vote. Would he have reached 50 percent in a runoff? No one will ever know. But the experience has led to renewed pressure to create some form of runoff system to block the anomalies that occur under the current rules.

THIS IS FAR MORE THAN AN ACADEMIC ISSUE. On June 22, voters in New York City will go to the polls and cast their votes by ranked choice. They will be able to mark their ballots for more than one candidate — up to five, according to their order of preferences. If no candidate receives a majority, the votes of the candidate who finishes last will go to the voter’s second-choice candidate. This winnowing process will continue until someone has a majority. The candidate who comes in first in the initial balloting may well lose when the re-sorting process is completed. This system is often referred to as an “instant runoff,” but there is nothing instant about it. The result may not be declared for weeks.

What is happening in New York is happening in a small but growing number of American jurisdictions. According to the organization FairVote, some form of ranked system is now in place in more than two dozen American cities, with a population of more than 9 million voters. More than 400 ranked-choice elections have been held in the United States since 2004.

The one state that has used a universal ranked-choice voting system is Maine, and it did so primarily to forestall a Tennessee-style travesty. Paul LePage became the state’s governor in a three-way contest in 2010 with more than 60 percent of the voters opposing him, most of those quite vehemently. Unlike Blanton, he was not personally corrupt, but he was belligerent, inflexible and combative, keeping the state in a condition of turmoil throughout his tenure. He managed to win re-election in another three-way race in 2014, again with a minority of the vote.

Angry at LePage and determined to prevent future such results, Maine switched to ranked-choice voting for all state offices in 2016, and elected Janet Mills as governor two years later after a primary in which she reached a majority in the fourth round of ranked-ballot redistribution. Then she won a majority in the general election, the first Maine governor in 20 years to do that.

Maine will use the system again in 2022. Alaska will use it as well, following passage of a referendum in 2020. Earlier this year, 29 cities and towns in Utah made a similar decision to go to ranked choice. And in May, Virginia Republicans used the system to select their gubernatorial nominee, Glenn Youngkin.

For most of the jurisdictions moving to ranked choice, however, the purpose is not to prevent the election of unqualified or unpopular candidates, however desirable that might be. The goal of the promoters is to foster a substantial change in American politics as a whole — to restore a climate of civility, moderation and bipartisanship to a process that has lost those virtues over the past several decades.

Is that realistic? Well, there are several plausible arguments for why ranked choice could significantly improve things. Ranked choice might reduce the potency of negative voting, for example. In a conventional system, if I go to the polls in a multi-candidate race determined to mark my ballot against the one candidate I truly despise, I’m not making much of a contribution to good government. If I can follow up by casting a second-choice vote for someone better, I am doing something more constructive.

Or let’s say I greatly admire one of the several candidates, but realize she has little chance to win and don’t want to throw away my vote. I can mark for her and then cast a second-choice vote for a candidate I mostly like who has a decent chance to win.

THOSE BOTH SEEM LIKE DESIRABLE OUTCOMES, but achieving them doesn’t come without a price. A ranked-choice system, especially if it is handled without sufficient skill, can end up confusing the electorate and leading people not to vote at all, or casting a spoiled ballot that doesn’t count. Jason McDaniel, a political scientist at San Francisco State University, examined several studies and concluded that there was a noticeable drop in voter participation in a city’s first two ranked-choice mayoral elections, although it leveled off after that.

Much more worrisome, a study by the MIT Election Lab found that in Maine in 2018 there was a decline in public confidence with the new voting system and an increase, not a decrease, in negative ads and negative campaigning. The study did find a 6 percent increase in the vote for third-party candidates.

The jury is still out on these changes, of course. But there is no evidence so far that it is leading to the civility or bipartisanship that proponents are hoping for. It does block the indignity of a Blanton or LePage situation. But that misfortune can be prevented by a simple runoff between the top two candidates. This solves at least some of the problems without creating a complicated five-way ranking system.

One other alternative to a cumbersome ranking is “approval voting.” Voters simply choose the boxes next to any candidate they generally approve of and leave the other boxes blank. All the approval votes are added together and the candidate who has the largest number of them wins. It’s very simple to operate. Fargo, N.D., and St. Louis are already using this method. Tishaura Jones was elected mayor of St. Louis in April under an approval system. There is one troubling drawback, however: Approval voting has no way to account for intensity of feeling. If I am crazy about candidate A and mildly positive toward Candidate B, they both get the same benefit. It isn’t a very precise form of judgment.

THE REAL QUESTION about all forms of ranked-choice voting is not whether they can help to guard against occasional fluke results. They can do that. The question is whether they can make a dent in a troubled political system.

In some circumstances, maybe they can. If you have a rough consensus between two major factions and a militant minority that keeps getting in the way by electing unpopular candidates, ranked-choice voting can be a major step toward bipartisan agreement. Unfortunately, we don’t have that. We have two major parties, at least at the state and national level, that deeply distrust each other and are almost equally divided in strength. If you put 10 candidates in a primary and allow voters to rank them, it’s very likely you will end up with hostility on each warring side. Gridlock doesn’t disappear.

What this suggests is that our political problems run deeper than a dysfunctional voting system. If and when we solve some of those problems of general distrust, we might not need complex new voting procedures. Until we solve them, the potential of this reform seems limited.

It doesn’t hurt to be hopeful about ranked-choice voting. In the short run, though, it may not make much sense to hold great expectations.
Alan Ehrenhalt is a contributing editor for Governing. He served for 19 years as executive editor of Governing Magazine. He can be reached at
From Our Partners