U.S. District Court Upholds San Francisco's Version of Instant Runoff Voting

On April 16, U.S. District Court Judge Richard Seeborg issued this 19-page order, denying a request that San Francisco be prohibited from using its version of Instant Runoff Voting in this year’s city elections. The case is Dudum v City and County of San Francisco, C10-504.

The plaintiffs had criticized San Francisco for letting voters choose only a first choice, a second choice, and a third choice. San Francisco limits voters to just three choices due to limitations imposed by technology. Plaintiffs had argued that limiting the voter to only three choices violates voting rights. Plaintiffs were represented by the Nielsen Merksamer law firm, the same law firm that represents the campaign to pass California’s Proposition 14, the “top-two” ballot measure on the ballot on June 8, 2010.

Judge Seeborg is an Obama appointee. He is the first federal judge appointed by President Obama to have issued an opinion in any lawsuit covered by this blog.


U.S. District Court Upholds San Francisco's Version of Instant Runoff Voting — No Comments

  1. This ruling is no surprise to anybody who knows anything about constitutional issues in election law. The lawsuit was a legal shot in the dark to begin with.

  2. Although I disagree with the decision, it was predictable.

    There’s a reason that San Francisco and then Pierce County Washington adopted “instant runoff voting” but after adoption changed the name to “ranked choice voting”.

    “Instant runoff voting” is actually a misleading name, because it implies that the method achieves the same result as a real runoff.

    The system is really a tally and elimination scheme, retallying without revoting (reallocating), and repeating until a “majority” of votes are reshuffled into one pile.

    Instant runoff voting can provide a different outcome than traditional runoffs, and, compared to a system with a different-day runoff in close races, deprives voters of the time and opportunity to learn more about the top two contenders.

    In San Francisco,”majority” is of the “continuing” ballots, not a majority of all ballots:

    “If no candidate receives a majority of votes from the continuing ballots after a candidate has been eliminated and his or her votes have been transferred to the next-ranked candidate, the continuing candidate with the fewest votes from the continuing ballots shall be eliminated. All votes cast for that candidate shall be transferred to the next-ranked continuing candidate on each voter’s ballot. This process of eliminating candidates and transferring their votes to the next-ranked continuing candidates shall be repeated until a candidate receives a majority of the votes from the continuing ballots.” SEC. 13.102. – INSTANT RUNOFF ELECTIONS.(D) go to this link and type in the SEC. 13.102 in search box.

    Again, my hope is that third party advocates will look at other voting methods. Consider New York’s experience with Fusion Voting, and look at Oregon’s recent choice to adopt Fusion and not IRV. NY is proof that Fusion voting can strengthen third party voices in a way that is meaningful and can influence the behavior of the “majority” parties. Fusion is basically an endorsement process but shown on the ballot, is additive at the polling place, and requires no special voting software.

  3. #2: While Joyce McCloy says that she “disagree[s] with the decision”, she doesn’t explain why she thinks San Francisco’s version of IRV is unconstitutional but instead just talks about why she thinks IRV isn’t a good idea.

    There are lots of voting systems that are clearly constitutional, some of which are widely used, even though all of us here would probably agree using them is bad policy. For example, at-large first past the post “vote for N” elections, are constitutional (and, except when there is found to be racially polarized voting, legal under the Voting Rights Act), yet they enable a voting bloc with a slight plurality win all of the seats on a legislative body. Another example that came up during this case was Hawaii’s ban on write-in candidates.

    I won’t try to address most of her points about IRV and fusion voting as an alternative, because they really are off topic for this news item.

    However, her argument that IRV is a misleading name because IRV is “really a tally and elimination scheme” different than “a real runoff” election actually addresses one of the legal points in the case. The plaintiffs tried to argue that the different rounds of tallying were actually separate elections, concluding that voters whose ballots weren’t counted in the final rounds due to the three ranking limit were being denied a right to vote. Because the rounds of counting are not separate elections, no one’s right to vote is lessened because they choose to rank only candidates eliminated before the final round, just as in first past the post elections, no one’s right to vote is lessened because they choose to vote for candidates who lose.

    It is true that IRV doesn’t always produce the same winners as two round runoffs, but there are good reasons for that. One is that IRV doesn’t use rankings to simulate a two round runoff, but to simulate the multi-round runoff procedure used at conventions and other meetings. Except when all the voters are gathered together in one place for an extended period of time, it’s clearly impractical to use multi-round runoffs, though they are the ideal method to guarantee that a winner has majority support. One compromise is the two round runoff system, which has simple ballots but doesn’t always end up with the best two candidates in the runoff (e.g., the 2002 French Presidential election). IRV is a different compromise.

    Another reason for the sometimes differing results is that while reducing the number of candidates for the runoff can have positive effects in allowing voters and the media to pay closer attention to the runoff candidates, when they are reduced to just two there is the negative effect of promoting mudslinging (if there’s no third choice, a campaign can only gain by attacking opponents). IRV supporters generally believe that the advantage of reducing mudslinging outweighs the disadvantage of the lost opportunity for additional scrutiny, or that other issues (such as cost and turnout) are more important.

    A two round runoff is not a gold standard to which all other single-winner election methods should be compared, and it’s not always even what people mean when they talk about runoff votes.

  4. “Although I disagree with the decision, it was predictable.”

    Why is the decision on constitutionality wrong? It’s not enough to dislike IRV.

    “Again, my hope is that third party advocates will look at other voting methods. Consider New York’s experience with Fusion Voting”

    Fusion has a rather mixed track record in New York. It has given the WFP some influence, but the Liberal Party was hardly a shining example of how fusion is supposed to work. Furthermore, the Democratic Governors of New York make very large contributions to the WFP, which would seem to undermine their independence.

    San Francisco’s election system is nonpartisan, so fusion wouldn’t work unless you made it into a partisan system.

    As for two-round runoffs, it’s the voters’ responsibility to find out what they need to know about the candidates by Election Day, not the date of the runoff. Giving voters “more time” comes at the price of a more expensive election with lower turnout.

  5. Just because something is a bad idea, doesn’t mean it is, or needs to be made, illegal.

    And #3: IRV “reduces mudslinging” now too? This can be disproved–just like the “better for third parties” claim–by looking at Australia. They’ve used IRV for years, and their political discourse is as mud-filled as the US.

  6. IRV = THE method to elect Hitler and Stalin clones — when the *middle* is divided.


    34 H–W–S
    33 S–W–H
    16 W–H–S
    16 W–S–H


    IRV – H beats S 50-49 Take Cover.

    Approval Voting – (add 1st and 2nd place votes) – W wins.
    Condorcet (head to head) – W wins
    Bucklin (add 1st and 2nd place votes) – W wins

    H Hitler clone
    S Stalin clone
    W Washington, George clone

    IRV IGNORES most of the data in a Place Votes Table — regardless of all math MORON judges.

    P.R. and A.V. — otherwise get ready for CIVIL WAR II due to the EVIL folks being elected with IRV fake majority *mandates* from Hell.

  7. San Francisco should tabulate the number of ballots that are discarded due to system-induced exhaustion (a ballot with 3 distinct choices that is deemed exhausted even though there are two or more continuing candidates), and then consider those votes in determining whether additional candidates can be excluded.

    For example, if the votes for the Nth-place continuing candidate plus the number of system-induced-exhausted votes is less than the number of N+1th-place continuing candidate, the Nth-place candidate can still be excluded.

    But if at any point in the count, there is a possibility that a voter was thwarted by a deficiency in the system to be able to fully express his preferences among the candidates and that this deficiency could have changed the ultimate result then counting should stop.

    The remaining candidates would then be placed on a runoff election ballot also conducted using a ranked-choice ballot.

  8. #7 I agree with you on the continuing vs. total ballots issue, but if your solution calls for a runoff election *after* the instant runoff election (which, remember, is after the primary elections), I start to think that there’s got to be a better way. I think approval or score voting is that way.

  9. Dale – don’t get stuck in the Winner Take All box. Proportional representation is what’s needed to represent independents and third parties. Within the confines of Winner Take All, instant runoff voting is fine.

    Note that fusion voting does nothing for handling more than two candidates. It’s smoke and mirrors to present it in this context.

  10. Joyce McCloy was singing a different tune about this lawsuit when it was filed. But it was a slamdunk loser, just like the lawsuit she was excited about trying to block instant runoff voting in Minneapolis.

    Joyce suggests she supports runoff elections because of the “second look.” But in her home state of North Carolina, runoffs are used frequently for statewide and congressional elections. Those state and federal runoffs have had big drops in turnout, plunging by more than 90% between rounds of voting. Candidates also have to raise and spend big dollars for the runoff round, and delayed runoffs mean more money for the voting machine companies that make most of their profits from servicing elections. That’s why Democracy NC and other leading civic groups there have supported IRV over runoffs.

    The form of IRV adopted in NC actually does directly simulate a runoff — e.g., the top two candidates advance to the second round, and everyone else is eliminated. That final round is likely to retain far more voters than the delayed runoffs Joyce supports.

    As to Australia, the median number of candidates in house races in 2007 was seven, and no race had fewer than four. So there were a lot of voter choices. I’ve seen absolutely no evidence that any voter didn’t cast a sincere ballot because of alleged flaws with IRV. The major parties dominate because, as Peter suggests, it’s still a winner-take-all system.

  11. As far as I can tell, Dudum did not a “court case” he lost a “motion for preliminary injunction.” Still, it would appear based on the 19-page motion-denial document that his chances in a lawsuit might be poor.

  12. #8 This was in reference to the elections in San Fransisco which doesn’t have a primary.

    Why should a voter be prevented from choosing who his supervisor is simply because he favors some lesser known candidates? In a conventional runoff, he would have the option of voting. In San Francisco he may be prevented from doing so unless he guesses right. There have been elections in San Francisco where no candidate had support of 50% of the voters even counting all 3 preferences.

    So occasionally they would have to go to an actual runoff.

    As for use in primaries, Fairvote in its amicus brief the 9th Circuit in the Washington Top 2 litigation recommended the use of ranked voting in the primary to determine which two candidates would advance to the general election. In such a form, 3 choices might well be adequate. Or you could simply use my adjusted form, and on occasion have more than 2 candidates on the general election ballot.

  13. #10 When was the last runoff in a North Carolina congressional race?

    In Australia voters not only are permitted to rank all candidates, but they are required to do so. How could anyone expect preferences to be sincere when it is unlawful not to vote nor vote all preferences?

    In senate races, political parties are permitted to lodge two below-the-line tickets. The reason for doing so under STV is to game the system.

    San Francisco apparently argued that voters would be confused if they could rank more than three candidates. Since the San Francisco city&county charter requires that new voting systems that they acquire support full preferences, it now appears that they regard the voters as part of the voting system and that they regard new voters as an acquisition.

  14. #11 The judge didn’t understand the argument being made.

    He said that presumably Minneapolis had at one time permitted full preferences and then reduced them. The court case there had been before implementation. Limiting preferences to three is especially unwise for multi-seat STV elections are used in Minneapolis.

    He made the same mistake with regard to Cambridge.

  15. Perhaps some appellate judges will be informed that IRV does NOT treat all later choices equally —

    i.e. ALL 2nd choice votes are NOT treated equally.

    ALL 3rd choice votes are NOT treated equally. Etc.

    See the EQUAL in the EQUAL protection clause of 14th Amdt, Sec. 1 — way too difficult for MORON lawyers and judges to understand ??? Duh.

Leave a Reply

Your email address will not be published. Required fields are marked *