Battle in Seattle: RCV vs. Approval Voting this November

So far approval voting has not lived up to its promise

Electoral systems are a democratic wonder that don’t get nearly enough attention by either experts or the public. It’s usually an eye-opener for most people when they find out that you can cast the same votes through different electoral systems – district elections vs. at-large vs. ranked choice voting vs. proportional representation vs. cumulative voting and more – and end up with completely different results, in terms of which candidates get elected.

That’s a mindblower. Suspend for a moment what you’ve heard about the impacts of campaign finance reform, or redistricting commissions or mail-in ballots or Election Day registration. Those are important components, but compared to the electoral system itself those are sides to the main show. The electoral system is the Big Top Tent of the democratic circus, while those other reforms are more like the booths where you purchase your cotton candy or toss rings to win the giant stuffed snake.

So it’s great news that this year a record 10 jurisdictions across the US will vote on whether to adopt ranked choice voting (RCV) in their elections, including the largest cities in several states, small towns, and the entire state of Nevada.

One of those cities is Seattle, where voters will have a choice this November over whether to adopt RCV or yet another method, called approval voting, to use in their primary elections for local offices. Voters will also vote on whether they want any reform at all, and if they vote ‘yes,’ then whichever of these two methods they select will be used to whittle down larger fields of candidates to the final two finishers, who will face off in a November runoff.

RCV vs Approval Voting

Which method would be better for Seattle elections, ranked choice voting or approval voting? That’s what voters need to decide. Advocates of both reforms are fiercely making their case to Seattleites.

Among advocates of the two methods, there is actually agreement on at least one thing: that the current method, called plurality voting in which all voters can cast one vote for their preferred candidate and the highest vote-getter wins, is the worst method of all.

But it sounds so simple, how could it be the worst?

It has to do with the specific defects of plurality voting, which results in spoiler candidates, split votes among like-minded voters, and candidates often winning with low percentages of the vote.

Imagine you are voting in a primary election and there are eight candidates running for the final two spots. Let’s say there are three moderate Democrats, three liberal to progressive Democrats and two Trump-MAGA Republicans. In this election, the three moderate Democrats win, respectively, 14%, 13% and 8% – 35% of the overall vote. And the three liberal/progressive Democrats win 14%, 11% and 8%, for a total of 33% overall. And the two Trump Republicans each win 16% for an aggregate of 32%. That means over two thirds of this electorate voted for Democrats, but because there were so many Democrats in the race who split the center-left vote, then two MAGA Republicans would make it into the top two, with one eventually becoming the “representative” for this heavily Democratic district.

This is not merely a theoretical consideration. This kind of “crapshoot casino” vote has actually happened in some Congressional districts in California when too many candidates from the same party ran. There is too much at stake to simply say “tough luck, too bad, better luck next time.” When an election method selects a single winner who is actually not supported by the vast majority of voters, that undermines the very fundamental spirit of democracy itself – the sense that the majority should prevail.

To prevent such split votes, other methods besides plurality can be used. Approval voting and RCV are two of those methods (and there are others, including STAR, Range and Condorcet voting).

But each of these methods have their own pros and cons. Nobel laureate economist Kenneth Arrow proved 70 years ago that all electoral systems have flaws. Nevertheless, some are better than others because they produce winners with fewer unintended consequences and that better ensures that the most popular candidates win.  

The flaws of approval voting

Approval voting has its strengths. Theoretically, it is simple for voters to use in that voters get to “approve” of as many candidates as they wish. So voters are allowed more than one vote, and election officials count up candidates’ totals and the highest vote-getters win. Simple.

But its simplicity is deceptive. The more you look into approval voting, the more you look under the tent, the more questions arise about its efficacy. It doesn’t really solve the spoiler candidate or split vote problems. Instead it kinda sorta turns it on its head.

Imagine it is the 2024 Democratic nomination for president. You prefer President Joe Biden as your top choice, but you also would be able to begrudgingly accept Gavin Newsom or Amy Klobuchar as the party’s nominee. So you are tempted to “approve” of all three candidates, and in fact that’s what the voting instructions as well as approval voting advocates instruct you to do.

But here’s the problem. Your vote for either Newsom or Klobuchar could actually help one of them defeat your favorite, Joe Biden. Biden’s very smart campaign advisors quickly figure that out, so candidate Biden starts telling his core supporters: “Don’t approve of any other candidate, only approve of me. Otherwise, you’ll hurt my chances.”

The other candidates start figuring that out too, and they relay that message to their supporters. Voters start engaging in a strategic voting practice known as “bullet voting,” in which they only use one of their many “approvals.” So what started out as a system to override the defects of plurality voting starts tilting toward – ta daaaa! – plurality voting.

Compare that dynamic to ranked choice voting. With RCV, voters rank their candidates, 1, 2, 3, and your vote counts for your first choice as long as that candidate is in the race. Your vote does not count for your second, third or subsequent rankings until your earlier-ranked candidates have lost. It is a runoff system, in which your lukewarm-preferred candidates cannot help defeat your favorite candidate. So voters can feel free to rank as many candidates as they wish. RCV truly liberates the voter to express all of her or his preferences without fear of unintended consequences like spoilers or split votes.

And RCV prioritizes that winners must have both strong core support (indicated by being ranked first on many ballots) and widespread support (ranked second, third, etc. on many ballots). But with approval voting, candidates win by having broad support but don’t need a lot of core support; a milquetoast candidate could win even though few voters regard that candidate as their favorite.

For all these reasons and more, which I will explore below, I regard approval voting as a much weaker reform than ranked choice voting.

Approval voting in practice

This is not just a theoretical concern. While approval voting has not been used in very many public elections – just two cities use it currently, Fargo, North Dakota and St. Louis – in both of those cities you can see the troubling dynamics of “bullet voting” playing out.

For approval voting’s maiden election in Fargo in June 2020, nearly 19,000 voters cast 42,855 total votes for city commissioner, with two seats being elected. So voters used an average of 2.28 votes per person, which is not that different than before approval voting, when Fargo voters were allowed to vote for up to two candidates. In its analysis, FiveThirtyEight concluded “not too many people actually took full advantage of approval voting” by using too many of their approvals. In fact, many if not most voters bullet voted.

For Fargo’s elections in June 2022, an opinion poll estimated that 60 percent of voters bullet voted in the mayoral election conducted by approval voting. Despite there being seven candidates – meaning voters could vote up to seven times, once for each of those seven candidates – voters approved an average of just 1.5 candidates.

In St. Louis, we see a similar story. For its mayoral primary in March 2021 using approval voting, two finalists were selected from a field of four candidates. In the old pre-approval system, voters had a single vote. But in the approval voting election, despite voters having four approvals to use, voters approved an average of only 1.6 candidates. In its city council primary, two candidates qualified to go on to the general election for each seat, and according to an analysis by Alan Durning of Sightline Institute, voters approved an average of just 1.1 candidates per seat. The low approval rate of 1.1 per ballot corresponds to a bullet voting rate of 90 percent or higher. Another study found that 24% of voters said that the reason they approved of only one candidate was because they feared that voting for more than one would hurt their favorite candidate’s chances.

Fargo candidates quickly figured out this bullet voting dynamic, because it just makes practical sense when running under the rules of approval voting. So candidates began instructing their voters accordingly.

Incumbent mayor Tim Mahoney during his reelection campaign said “I would probably bet that every candidate says just vote once.” One of his challengers, Republican state representative Shannon Roers Jones, agreed saying, “It will be important for me to convey to people . . . that the most effective way to elect the person they care about is to vote only for that person.” Election administrator Mike Montplaisir, who runs Fargo’s elections, said that he also “voted for only one candidate — the candidate I wanted to win — because voting for anyone else is like taking a vote away from [my first choice].” He added, “You can game the system.”  

Elections are like a board game, with certain rules and strategies that give advantage to the players that play the game the best. For approval voting, these strategic voting schemes are a core part of the system. Overall, bullet voting rates in those Fargo and St. Louis elections were likely between 50 and 90 percent. In comparison, RCV bullet voting rates have been between 10 and 40 percent. In New York City’s contested Democratic RCV primary for mayor in September 2021, which nominated current mayor Eric Adams, only 13 percent of voters bullet voted, according to a study by the City University of New York.

So this is a real defect of the approval voting system. And its proponents refuse to acknowledge it, which only undermines their advocacy. That’s not to say that using approval voting might not still have advantages over a plurality method. In the St. Louis mayoral election, which resulted in the election of St. Louis’ first black female mayor, some observers credited approval voting with preventing the splitting of the black vote. Still, this dynamic of bullet voting, and candidates encouraging such behavior from their voters, is something to watch closely and to gather more data about, as more elections are conducted using this method.

Approval voting disadvantages low-information voters

Voters and candidates who understand how in approval voting, choosing a second or third candidate can cause your favorite candidate to lose, will always have an advantage over those who are not aware of it, or are undisciplined in how they spread their votes. Like with plurality voting, voters still must strategically weigh whether it’s more important to help their first choice win — by withholding votes from compromise candidates —or to approve a “lesser evil” candidate to prevent their “greatest evil” from winning. But for those voters who do not understand these dynamics, they could inadvertently end up undermining their favorite candidate by simply following the instructions on the ballot or listening to approval voting advocates, who invite them to approve of multiple candidates.

In fact, just like with plurality voting, with approval voting you cannot even guarantee that the most popular candidates will win. Look at my earlier example, in which two Trump Republican candidates advanced to the runoff because too many Democratic candidates split the center-left vote. With approval voting, if all of those candidates were to instruct their core supporters to only approve of them, those two MAGA Republicans would still finish in the top two even though Democratic candidates garnered two thirds of the vote. Or if the supporters of some candidates are disciplined in how they use (or withhold) their votes, while other voters are not disciplined, distortions in outcomes could easily result.

So for many voters, they are stuck inside an “approval voting conundrum,” a kind of Catch-22, in which they might be damned if they vote for too many candidates; and unlike with RCV, they are not even allowed to express different levels of intensity for various candidates without potential blowback. Not only is that prejudicial against the uninformed voter, it also does not reflect the way most people think as they go about their lives. Most people rank things in their heads all the time, from their favorite movies to flavors of ice cream. And in most elections, they usually have some idea of which candidates are their favorite, which is their second choice, and so on. But with approval voting, that natural process is truncated.

Seattle has a choice between approval voting, which is largely untried in public elections and has exhibited strange characteristics and anti-democratic tendencies in the handful of races that are available to study; and ranked choice voting, which has been used for nearly 100 years in Australia and Ireland, and more recently in Malta and Sri Lanka to elect their presidents, and in Scotland to elect its local councilors (in multi-seat districts). Increasingly it is used in the US to elect candidates at local, state and federal levels, including some 53 cities and the states of Alaska and Maine (including blue cities such as New York, Minneapolis, and San Francisco and nearly two dozen red cities and towns in Utah), and six states where parties use it for presidential primaries while six other states use it for their military and overseas voters.

Some 11 million voters in the United States now live in RCV jurisdictions that in total have completed hundreds of RCV elections, and Americans have cast tens of millions of RCV ballots. To be sure, the consequences of using either ranked choice voting or approval voting in the primary election to select the top two candidates for a November runoff are not nearly as momentous as switching elections to proportional representation, which is the best method of all for a “multi-everything city” like Seattle. Three hours south down the I-5 Interstate corridor, Portland will be voting this November whether to adopt a method of proportional ranked choice voting to elect its city council.

Both of these votes in the Pacific Northwest are consequential. It’s time to vote, Seattle. Good luck in making your decision.

Steven Hill @StevenHill1776