Donate

Ranked-Choice Voting Explained

| October 28, 2011
  • Share:
  • Facebook
  • Pinterest
  • Reddit
  • Email
Ranked-choir voting demonstration ballot

Ranked-choice voting demonstration ballot (Courtesy SFGov.org)

In November San Franciscans choose their next mayor through an electoral process called ranked-choice voting (RCV). Also known as “instant run-off voting,” voters will pick three candidates (instead of one), and rank them in order of preference, eliminating the need for a separate runoff election.

Although this is not the first time San Francisco has employed this system (RCV was first used in 2004 for a board of supervisors race and again in the 2007 mayoral race, when Gavin Newsom was reelected with 72% of the vote) it is the first time the city is using it to decide a competitive mayor’s race that has a large candidate pool. And many are waiting to see how well it all works out.

Not surprisingly, RCV has its discontents; some candidates stand to benefit more than others, and a variety of opponents have cried foul, calling the system too complex and undemocratic. Last year, the system survived a court challenge alleging that it violated voters’ rights.

The biggest issue at play is that you no longer need to be the top vote-getter in the first round in order to win the election; instead you need the most-combined first, second, and third choice votes. And that significantly changes the calculus. Factor in that there are a grand total of 16 candidates running for mayor this year, and you’ve got yourself a pretty thick electoral swamp to wade through.

So, how’d this all begin? Well, San Franciscans voted on it (ah, the irony!). In 2002, residents approved Proposition A, effectively amending the city’s charter. The RCV system’s actually been used in smaller San Francisco elections since 2004, but never for such a high-stakes contest. It used to be you just voted for one candidate, and if no one received over 50% of the votes, a December runoff election was held, and the winner was then determined by whoever got the most votes. RCV will now be used to elect all of the city’s major elected positions, including members of the Board of Supervisors.  But with Election Day just a few weeks away there still seem to be a whole lot of voters out there unfamiliar – or just flat out confused – with how RCV works. So, here’s our best stab at explaining the rules of the game:

First off, a quick preview of what the new ballot looks like. A good mock-up is available at San Francisco’s Department of Election’s site. The basic format is fairly simple: Three side-by-side columns (first-choice, second-choice and third-choice). Each column includes the names of all 16 candidates.

Round 1:  An Even Starting Line
From the pool of 16 candidates, you pick your first, second, and third choices. (Note: you don’t have to pick three; if you want, you can just pick your first-choice, or your first two choices, etc. It also doesn’t do you any good to repeatedly pick your first-choice three times – it’ll only be counted once.

If any candidate gets more than 50 percent (50% +1) of first-choice votes, that candidate is automatically elected. Game over. But, if no one receives that majority, we go to the second round.

So just for kicks (and because puppets are more fun than politicians), let’s pretend we’re observing a heated mayoral race on Sesame Street (Just ignore that streets don’t typically have mayors, or the fact that the candidates are made of felt and fur.). There are four candidates running, and a total of 24 voters casting ballots.

Cookie Monster (the clear frontrunner, of course, well loved for his oratorical gifts and promises of free pastries to every supporter) gets 10 first-place votes. Oscar the Grouch gets 8 first place votes (with strong support from the waste management industry and a large contingent of the generally disgruntled). Big Bird gets 4 first-place votes. And poor, impetuous Grover gets only 2 first-place votes. Since no one got more than 12 votes, there’s no clear majority, but we do have our first loser … so we move on to Round 2.

Round 2: The First Elimination
Grover, the candidate with the least amount of first-choice votes, is outta there! But (and here’s the part that seems to trip people up the most), for the two voters who picked Grover as a first-choice, their second-choice votes still count. Here’s how:

One of the voters who chose Grover picked Oscar as a second choice. So that vote goes to Oscar (who now has a total of 9 votes). The other voter in Grover’s small fan club picked Cookie Monster as a second choice. So, that vote goes to Cookie Monster (who now has 11 votes).

At the end of Round 2, here’s the tally:
Cookie Monster: 11 votes
Oscar: 9 votes
Big Bird: 4 votes
Still no clear winner (because there still are three candidates standing), so onto Round 3 we go!

Round 3: The Deciding Moment
Three candidates left, and Big Bird’s got the least amount of first-choice votes (only 4), so that oversized avian is done! Now, we look at the second-choice votes of those four voters who picked Big Bird as their first-choice. Remarkably, as it turns out, all four of Big Bird’s second-choice votes were for Oscar! That means that Oscar picks up four more votes, giving him (or it?) a final tally of 13 votes to Cookie Monster’s 11 votes. And thus, that grumpy, trash-dwelling green dude is the new boss in town.

O.K., so in the San Francisco mayoral election, things might not be quite that simple (and all the candidates are probably going to have noses). But, hopefully this example does illustrate how a candidate can viably receive the most first-choice votes and still lose the election. Because there are 16 candidates in the real race, that same elimination process keeps going until one candidate emerges with the most votes.

One key to understanding the RCV instant runoff process is remembering that the number of elimination rounds is determined by the number of candidates running. So, in the case of San Francisco’s mayoral race: there are 16 candidates, thus, 15 elimination rounds to determine a winner. In the Sesame Street scenario, there are a total of 4 candidates, requiring three elimination rounds to determine the winner. Just think of it as last man/woman/puppet standing.  Take a look:

Round 1: Four candidates on the ballot with a total of 24 votes cast.

C. Monster     O. Grouch     B. Bird     Grover     Total votes
10 votes          8 votes         4 votes     2 votes    24

Round 2: Three candidates standing; Grover is eliminated and his votes go to Cookie Monster and Oscar. Remember that all 24 votes still count, but some have just been transferred to other candidates.

C. Monster     O. Grouch     B. Bird     Total votes
11 votes          9 votes         4 votes     24

Round 3: Two candidates left; Big Bird is eliminated and all four of his votes go to Oscar (because the people who voted for Big Bird as their first-choice picked Oscar as their second-choice.

C. Monster     O. Grouch     Total votes
11 votes         13 votes         24

With 13 votes to Cookie Monster’s 11, Oscar the Grouch is the winner!

Oakland’s 2010 Mayoral Election
Last year, Oakland used RCV to elect its mayor and witnessed a similar outcome: There were 10 candidates, and Don Perata, the clear frontrunner (who vastly outspent his opponents during the campaign), got 35% of the first-choice votes. That left Jean Quan in a distant second with only 24% of first-choice votes. But Quan – who anticipated this outcome and allied herself with other underdog candidates and their supporters – received far more second-choice votes than did Perata. And after all the elimination rounds, with second and third-choice votes factored in, Quan received 51% of the vote to Perata’s 49%.

The Critics
So is RCV a good thing? The jury’s still out. It really depends on who you ask.  (Oakland’s Mayor Quan, I’m guessing would say yes; Don Perata … not so much. Oscar the Grouch though, is definitely a big fan.)
Like pretty much everything in politics, the system’s got its strong supporters and staunch enemies.

Some of the big arguments from supporters of RCV:

  • It could save taxpayers millions by eliminating the need for local primaries and separate runoff elections.
  • It boosts electoral competition because candidates only have to raise money for one election per cycle, not two or three.
  • It gives underdog candidates a better chance and produces a winner that’s supported by a clear majority.
  • It discourages mudslinging and negative campaigning; candidates are now more likely to ally with each.

Opponents say:

  • It’s too confusing for voters and unnecessarily adds to the complexity of an already complicated ballot.
  • There is lots of room for technical error as election computers tally results through the use of a complicated algorithm.
  • It encourages less popular candidates to game the system by teaming up against the frontrunner. Is this is a fair or appropriate strategy? Depends who you ask.
  • It’s discriminatory to less educated or knowledgeable segments of the voting public who haven’t received sufficient instruction on how the system works.

Need a Visual Aid?
KQED TV’s This Week got a rundown straight from the San Francisco Department of Elections. Watch the video below:

Related

Explore: , , , , ,

Category: Politics

  • Share:
  • Facebook
  • Pinterest
  • Reddit
  • Email

About the Author ()

Matthew Green runs KQED’s News Education Project, a new online resource for educators and the general public to help explain the news. The project lives at kqed.org/lowdown. Reach Matthew Green at mgreen@kqed.org.
  • http://fairvote.org Rob Richie

    Frances Matthew does a nice job, but …Matthew Green could have been clearer…. Matthew, consider the name that is also used: instant runoff voting. Every voter has one and only vote. But if your first choice doesn’t make the runoff, you ballot can still count – for your next choice who still is in the race. It’s __not__ “combined first, second and third choice votes” – -that makes it sounds like all those rankings count at the same time, which is wrong.

    So this statement is wrong as well: “You no longer need to be the top vote-getter to win the election.” That’s silly. Of course you need to be the top vote-getter — but in the final round, not in the first round when the vote is split among several candidates. For example, when Jean Quan was matched against Don Perata last year, she was preferred by more people: one-person, one- vote comparison between the two and Quan wins. And she also was was the “top vote getter.” The only reason she wasn’t the top vote-getter in the first round is because the majority vote was split.

    It was the same rule In the old runoff system. Being the” top vote getter” in the first round didn’t mean you automatically won. (In fact, in the last 14 runoffs in SF, 6 of the first round leaders lost.) You needed 50% plus one. RCV is the same: if you don’t have 50% plus one of first choices, you hold instant runoff rounds to determine the real top vote getter.

    Finally, consider that Ireland just held yet another RCV election for president this week. Seven candidates, no talk of “spoilers,” as we have in our “plurality-wins” presidential system. No need for the candidates to raise big bucks for a second runoff.

  • David Cary

    It is not nearly as complicated as Matthew Green indicates. I’ll bet he could condense his explanation to less than half, and it would be clearer. Some key points about RCV in San Francisco:

    1. The candidate with the most votes wins — every time.

    2. In each contest, every voter has one vote, but three choices for how that vote is counted.

    3. In each round, your vote counts for the your most preferred candidate that is still in the running.

    4. It is not a new ballot layout. It is the same kind of ballot that San Francisco voters have been using for the last seven years. Every city office that is elected with RCV, including city-wide offices and the mayor, have been elected at least once with RCV already.

    5. Most voters don’t have to know how votes are counted in order to vote effectively. Knowing how the votes are counted can help in some cases, but the beauty of RCV is that simple voting is usually very effective voting.

    6. A computer can tally the entire mayoral election in less than a minute.

    7. One of the biggest sources of voter confusion about RCV is the misinformation being given to voters. Sometimes it is unintentional misinformation, and sometimes it is very deliberate, carefully crafted disinformation.

  • http://www.cfer.org Steve Chessin, President, Californians for Electoral Reform

    The second opposition argument:
    “It could take too long for election computers to tally results (up to several days) because of the complicated algorithm.”
    is just plain wrong. It takes seconds to run the algorithm; what takes days is processing absentee and provisional ballots, and that applies to all elections, not just RCV ones.

    It takes several days to determine the outcome of an election, whether RCV or simple plurality, only if the election is close. Consider the November 2010 election for Attorney General. The outcome of that election wasn’t known for days because we had to wait for all of the absentee and provisional ballots to be processed, and that wasn’t an RCV election. If an election isn’t close, we know right away who the winner is going to be, even if all the absentee and provisional ballots haven’t yet been processed. That is true about plurality elections, and that is also true about RCV elections. The difference with RCV, of course, is you can’t tell if an election is close until you run an RCV tally; just looking at first choices isn’t enough. That’s why it took so long to determine that Jean Quan beat Don Perata, because the election was close; Perata’s 9% lead in first choices was misleading.

    The reason why some people think it takes a long time to tally RCV elections is that until recently the Department of Elections had a policy to not run the first semi-official RCV tally until the Friday after the election. That policy was recently changed to run the first semi-official RCV tally the Wednesday afternoon after the election. That’s good as a first step, because there is no reason why a semi-official RCV tally can’t be run election night, as San Francisco’s Voting Systems Task Force recommends. At least this year we’ll only have to wait a few hours, rather than a few days, to know if an RCV election is close or not.

  • tony santos

    I believe the segment on IRV was inadequate because you failed to have an opponent, such as myself, speak on behalf of the opposition; also, the story had several inaccurancies; the system does not produce a majority winner and as the program suggested and showed IRV is confusing to those who have a problem with the english language. Both Mr. Chessin and I have had the opportunity to speak about IRV; he for and me against. It is noted there was mention of IRV reducing negative campaigning; just look at is occurring in SF. There is as much mudslining as in any other campaign and there certainly was a lot of negative campaigning in my race for Mayor last year. Finally, why wasn’t there any mention of the communities who have used IRV and then repealed the system? The system has been repealed in many communities after one use of IRV. Burlington, Vermont, Aspen, Colorado, Cary, North Carolina, the state of North Carolina for Judge races and Pierce County, Washington. Further both England and Fort Collins, Colorado voted overwhelmingly against adopting IRV. Hawaii has also voted to table the discussion about introducing IRV. Our group has played a role in many of these communities..If IRV was a great system, I can assure you these communities would not have rejected the system. Again Mr. Richie the opportunity for you to debate us is still on the table. Quit ducking us. You have my email address and can accept our offer to debate the subject any time you deem it appropriate. Tony Santos, former Mayor City of San Leandro, Calif.

  • Chris Jerdonek

    “The biggest issue at play is that you no longer need to be the top vote-getter in the first round in order to win the election.”

    But this was always the case. In the old two-election November-December runoff system, the person winning the December runoff didn’t necessarily have to be the top vote-getter in the first-round November election.