Ranked-Choice Voting: Does Anyone Really Understand It?

Every local election since 2004 has allowed voters the option of ranking their top three choices for office. If a voter's top choice is eliminated from contention, they can at least rest assured their support went to a similar candidate. A common complaint is that this system confuses voters. Turns out it also flummoxes candidates, who are losing elections for not adopting a ranked-choice strategy.

This year, outgoing Supervisor Sean Elsbernd and his chosen successor, businessman Mike Garcia, approached the other moderate frontrunner in District 7, labor leader F.X. Crowley. Neither candidate wanted progressive school board President Norman Yee, the clear outlier among the three, to win. But Crowley refused to issue a ranked endorsement, on the grounds that Garcia's lagging campaign just wanted labor's help. Both camps watched the neighborhood go progressive, as Yee won by the razor-thin margin of 130 votes.

Garcia's votes, when distributed to the other candidates after he was eliminated, went to Crowley by a 2-to-1 margin, but not enough of Garcia's ballots had Crowley ranked to make the final difference. Had Crowley agreed to a ranked-choice strategy, "I believe he'd be supervisor," Garcia says.

In progressive hotbed District 5, voters had three prominent left-wing candidates to chose from. It was an ugly campaign — ousted incumbent Christina Olague was blasted for her vote to reinstate Sheriff Ross Mirkarimi, and groping allegations sank progressive Julian Davis — and one without much coordination of ranked-choice voting. That helped London Breed win.

And speaking of Mirkarimi, the city might have been spared much of the drama of the last year had ranked-choice voting been in the playbooks of candidates Chris Cunnie or Paul Miyamoto, either of whom might have beat Mirkarimi with help from the other. Instead, the two most similar candidates — in this instance, career law enforcement officers — finished second and third behind the outlier. Many Cunnie votes had no second- or third-place candidates marked.

These scenarios are sometimes used to argue that ranked-choice voting should be scrapped in favor of the old December runoff election. Mostly, it needs refinement: Voters here can pick only three choices; other jurisdictions allow them as many rankings as there are candidates. Nearly 4,000 votes in District 7 were thrown out because they had none of the remaining candidates marked; in last year's mayoral election, 52,000 ballots were so "exhausted." As long as San Francisco's Elections Department lacks the technology to weight votes based on rank, elected officials — like Yee, and like Mayor Ed Lee — will win with only a plurality. Especially if their opponents let them.

My Voice Nation Help
David_Cary like.author.displayName like.author.displayName 2 Like

In District 7 this year, 33% more voters decided the contest between the final two candidates in the last round, compared to the District 7 contest in 2000, which was also extremely competitive and close, but used a separate December runoff that created lots of exhausted voters.

There is a lot of misinformation in the media that can confuse voters, like talk about RCV weighting their vote, or describing abstentions as votes being thrown out, or concluding that if voting patterns don't match a simplistic, naive  political model, maybe the problem is with ranked choice voting.

I agree with Chris Roberts that improvements can be made to San Francisco's implementation of RCV, including allowing more rankings and a simpler ballot design that more modern voting equipment could support.  But please also recognize the improvements RCV already is delivering and be careful about underestimating the capabilities of voters.

diabolical_mdog like.author.displayName like.author.displayName like.author.displayName 3 Like

I think it's FANTASTIC that bungling political consultants (and candidates) who couldn't fathom a campaign that included second-choice endorsements were defeated! Hopefully this will be the nail in the coffin for campaigning as usual in SF; woe to those who ignore the lessons mentioned in this article, and hooray for less negative campaigns in the future!

steve.chessin like.author.displayName like.author.displayName like.author.displayName like.author.displayName 4 Like

Candidates can understand RCV when they want to, especially if they don't let their egos get in the way. Whether or not you are happy with the outcome, that's exactly what Jean Quan and Rebecca Kaplan proved in 2010, when they each told their respective supporters to rank the other one as second choice if they didn't want to see Don Perata elected as Oakland's mayor.

As for voters, the average voter does understand it, given that the vast majority cast valid ballots.

There are folks that don't like RCV, because they preferred the influence they could exert with those low-turnout December runoffs, as negative campaigning is much more effective in a one-on-one race than in a field of three or more candidates. But they are in the minority.

It is true that San Francisco could benefit from equipment that would allow voters to rank all the candidates. Fortunately, some of the money San Francisco is saving by not having to pay for runoff elections can be used to finance that upgrade.

Rob_Richie like.author.displayName like.author.displayName like.author.displayName like.author.displayName like.author.displayName 5 Like

Terry Reilly (masquerading this time in his comment as "Transfer"  and the person who posted the anti-RCV videos he touts) is, as usual, very selective in what he shares. For example, Corey Cook may report on ways RCV can work better, but he also told a Berkeley reporter this summer that RCV is an improvement to most voting systems used in the United States. He said: "The biggest advantage is that (ranked-choice) allows voters to express multiple preferences. It allows for the nuance that exists in voters to emerge on the ballot.”

Critics tend to look past the many ways that RCV is working well. In addition to the ChrisJ examples, consider the District 3 election in Oakland this year. In a majority-minority district, several African American candidates ran. Their combined vote total was more than 50%, but a white candidate who had spent the most money led in first choices. After the RCV tally, the strongest of the African American candidates was the winner.

Or take the Berkeley election for mayor. It was won on the first ballot by incumbent Tom Bates, who also won by a comparable margin in the two-person race in November 2008. But this year, the percentage of Berkeley voters at the polls who skipped the mayoral race was cut in half -- in other words, the RCV race and its greater mix of candidates engaged them a lot more than the prior runoff-type vote.

Finally, we had a great chance to see runoffs in action this year in the Top Two primary. There were several bizarre outcomes based on lots of wasted votes in the first round. For example, in congressional district 31, only two conservative Republicans advanced and one of course won, even though the district voted heavily for Obama and is majority-minority.June turnout was MUCH lower and far less representative of the electorate than the high turnout November electorate.

I agree that more rankings makes sense, but RCV in its current form is a big improvement over non-RCV systems. See some of the useful resources collected by San Francisco backers at http://www.sfbetterelections.com

ChrisJ like.author.displayName like.author.displayName like.author.displayName like.author.displayName 4 Like

I think the better question is whether ranked choice voting is doing a good job. Are minorities winning representation? Are more voters choosing the winner because elections are held in November when voter turnout is highest? The answer to both questions is a resounding yes. 9 of the 11 supervisors will be non-white, and turnout this November was a whopping 73%.

You can substitute any aspect of our government into that attention-grabbing headline and make another headline out of it.  Does anyone really understand the electoral college? Does anyone really know what the Assessor-Recorder does? Does anyone really understand all those city and state propositions that we have to vote on every year?

The point is that if someone wants to, they can understand them.  The information is out there. But if they don't want to understand, then they probably won't.


Ranked Voting is a modern day literacy test for voters. The "informational Costs" are just too high for the average voter - except for those that are deeply involved and invested in Ranked Voting. They understand it, and just cannot fathom how others don't. That's their problem - they just cannot believe people are not as smart as them. And if you say RCV is confusing, you are attacked as "not giving Voters enough credit" or "calling them stupid". Prof. Cook of USF, wrote in a recent article entitled THE TROUBLE WITH RANKED CHOICE VOTING and wrote "Anyone who offers anything short of full-throated support (of RCV) is branded a critic and dismissed with misdirections and ad hominem arguments." He also has pointed out that by his analysis that upwards of 10% of ballots had errors on them, particularly minorities, and elderly voters. Here's the link to his article - http://www.spur.org/blog/2012-01-06/trouble-ranked-choice-voting.

Other studies have shown that even those that "say" they understand RCV, do not accurately describe RCV - meaning, they really don;t know how it works.

A quick search of "Ranked Choice Voting Confusion" on YouTube generates a whole list of news reports. My favorite video viewed is this one from the UK, where their national vote on RCV (called Alternative Vote) failed with merely 30% support.  Watch it here: http://youtu.be/CZwOrySIIeY

David_Cary like.author.displayName like.author.displayName like.author.displayName 3 Like

@Transfer Actually RCV helps simplify things for voters.  The informational costs of voting with RCV are often less for voters compared to simple plurality voting or traditional runoffs.  With RCV there is less need to vote strategically, there is only one election, and no etcha-sketch campaigning.

Any skilled communicator can take something simple and make it seem hopelessly complicated.

San Francisco Concert Tickets
©2014 SF Weekly, LP, All rights reserved.