California Bill to Use Instant Runoff Voting in Special Legislative Elections Does Not Pass Fiscal Committee

On May 27, California SB 1346 failed to pass the Senate Appropriations Committee. The bill would have made it possible for willing counties to use Instant Runoff Voting in special legislative and congressional elections. The bill had earlier passed the Senate Elections Committee.


California Bill to Use Instant Runoff Voting in Special Legislative Elections Does Not Pass Fiscal Committee — 19 Comments

  1. Does this mean that County A could use IRV while County B doesn’t in the same special election?

    How would this work?

    It seems more sensible to conduct all special elections by IRV and skip the extra expense of a primary.

  2. Except, as reported last week, Minnesota hasn’t seen the promised, mythical “cost savings” that IRV is suppose to bring.

    Which is probably why the _Fiscal_ Committee blocked it.

  3. Looks like lawmakers weren’t willing to take the risk of a Florida style election fiasco when it comes to their elections.

    IRV is the only election method that isn’t additive, i.e you can’t simply add the votes up to get the winner.

    The votes have to be centrally tallied, so if that is a congressional race, there might be several counties involved that would have to transport or transmit their votes all to a central counting place.

    So if one jurisdiction fouled up the count, it could mean re-counting all votes.

    And if the election was close, or if a candidate challenged it, the cost to do a manual count would be astronimical.

    The SF elections budget advised that if San Fran had to manually count their votes, it would cost around $1 million dollars. Extrapolate that to consider a congressional district.

    It might work in Ireland, where they only use it to elect the President, (a ceremonial slot), and where they use it once every 7 years at the most, and sometimes not that often! Sometimes the President runs opposed so that there is no election, but in US ballots, especially California’s ballots, its just not feasible.

    Why not promote Fusion IF your goal is to see a stronger third party or see that third parties impact the elections?

  4. Bills don’t always pass first time around — this had a good run for a first year. New uses in Oakland, San Leandro and Berkeley this year will make IRV all the more “normal” in California.

    Joyce McCloy doesn’t seem to realize that fusion wouldn’t address this particular situation at all, which is seeking a way to fill vacancies in one election rather than two. Fusion also does nothing to accommodate more than two __candidates__. It “works” best when no more than two candidates run, but run with more than a total of two party labels.

  5. The bill, as amended, only applied to special elections that were entirely in one county. Los Angeles County has 28% of California’s population so this would be especially useful in that county, which has lots and lots of districts that are not associated with any other county.

  6. The bill has a lot of flaws:

    The decision to use IRV or not would not be made until after the vacancy occurred, and the counties involved could take up to 35 days to decide whether or not to use IRV. There is no requirement that the decision be made based on the election capabilities of the counties involved, so counties could game the process based on which method would produce a better result for a particular candidate or party. That is, it would be a political decision, and not a technical decision.

    Since the timing of special elections is based on when the governor issues his proclamation, the delay while the counties fiddle would further delay a conventional special election. We could see even more of the games played by the boys and girls of Sacramento Junior High in confirming Lt.Governor Maldonado almost 6 months after John Garamendi resigned.

    A significant share of special elections are resolved in the special primary, and the special general is cancelled. So use of “instant” runoff would delay the result, cost more money, and be unnecessary.

    The bill would permit the defective form of California IRV where voters are restricted to expressing 3 choices. In most congressional special elections in California there have been more than 3 candidates of at least one party, so a voter would not be allowed to rank all of them before his ballot was exhausted and discarded.

    The bill violates the current constitutional requirement that there be a partisan primary election for the legislature, and that each party has the right to have a candidate on the general election ballot.

    Proposition 14 would resolve that conflict. Ironically, the sponsor of the IRV bill, Loni Hancock, is one of the most vocal legislative opponents of Proposition 14.

    Hancock’s district includes part of AD 16 where Audie Bock was elected in a special general election after trailing badly in the special primary. That election showed the value of a conventional runoff, which gives an opportunity for voters to carefully (re)consider the merits of the Top 2 candidates.

    Voters should approve Proposition 14.

    Then, if IRV has merit for special elections, then it should be made mandatory for all legislative and congressional special elections, and be conducted entirely by mail.

    Paper ballots can be utilized, with voters simply numbering their choices, like is done in Ireland, Northern Ireland, Scotland, and Australia.

    Since it is a special election, there is no problem in using a ballot limited to the one race being contested, which simplifies hand counting. Since all ballots will be returned by mail to a central location in each county, there is none of the concern that Joyce McCloy expressed regarding precinct level counting.

  7. IRV is no guarantee that we will fill “vacancies” with one election rather than two. If just one polling place, or one county screws up the vote, then the whole district would have to retally the votes. Every time there would be a correction in one jurisdiction, the totals would have to be retallied.

    Because of IRV’s complexities. there’s a bigger chance of a Florida style meltdown where a panel of judges gets to decide who wins.

    If the results were challenged and a recount called for? $1 million to do a manual recount of around 450,000 votes is alot (SF figures), How much would it cost to recount a congressional district?

    How about logistics? Which county would be the “Central location” for the ballot tallying center?

    IRV won’t save money.

    Minneapolis Council Members learn about IRV’s continuing costs:
    So far, instead of saving money and increasing turnout – IRV has added an additional $244,000 in costs each year, and in the city’s first IRV election, turnout was the lowest in over 100 years.

    The groups that want to help California “save” money with IRV are also the groups paid to promote IRV and who also sell their election consulting services related to IRV.

    Fair Vote provides election services. From their website:

    “The Election Services Group
    We perform your elections.

    Total solutions: We can assemble the necessary team to conduct all aspects of an election at any scale.
    Once you settle on the necessary parameters — election design, voter list, balloting technology and budget —
    we will conduct the entire election from start to finish, from voter education to certified election results.”

    Trust, yet verify. Don’t take my word about Fusion, just google this: fusion + working families success

  8. #5 When was the amendment made restricting the process to one county? The 5/5/2010 version does not. The legislative analysis doesn’t indicate any such limitation.

    IRV would make more sense if Proposition 14 was passed.

    1) It would make IRV comply with the constitution, which currently requires that a candidate from each party advance to the general election.

    2) It could be made mandatory for all legislative and congressional special elections, eliminating political gaming.

    3) Paper ballots and all mail elections could be used. Voters could rank as many candidates as they wish, by simply using a numeral in a box – as is done in Ireland, Northern Ireland, Scotland, and Australia.

    Minneapolis demonstrated it was feasible to count IRV votes by hand. And in a special election you could use a single ballot for the single special election, which greatly simplifies IRV counting.

  9. Irish elections and use of IRV (Single Transferable Vote)

    Joyce McCloy is inaccurate when he states that IRV is used only in the Irish Presidential elecitons.

    A Proportional Representation form of it is used in the Irish Parliamentary elections as well and seem not to have resulted in confusion or chaos.

    Its also used for the election of the Australian House of Representatives

  10. IRV and STV are not the same thing.

    Yes, IRV is STV with a single winner; but STV becomes less and less proportional and less and less effective the fewer winners it selects per election. Cutting it down to just one guts pretty much any advantages out of it, leaving only the downsides, and elections that will tend towards the same two-party-dominated results that we get now with plurality (thanks for bringing up the Aussie house, btw).

    So yes, *IRV* is only used in Ireland for presidential elections.

    No proportionality, no cost-savings, no improvements for third parties; there’s pretty much nothing to gain from using IRV.

  11. #8

    No, general elections under IRV would make Prop 14 redundant and unnecessary.

    Unless you mean to suggest using a “top-two IRV” (ugh, seriously, just go with STV… or, you know, top-two approval) instead of top-two plurality in the primary? That just *compounds* cost problems on top of each other, and then fails your first numbered point anyway.

  12. Minneapolis found manual counting and voter education to be expensive, but who knows how much new voting machines would cost.

    City council actions :: Korbel confirmed
    BY CRISTOF TRAUDES May 10, 2010
    “…the projected ongoing costs of RCV total about $242,000.

    Technology could be the savior here. There are machines that can count RCV ballots; however, none are certified yet by the state, and that certification isn’t expected unless more cities switch to RCV. And even then, while the city would save a projected $140,000 in RCV costs by being able to eliminate the hand count, the cost of technology is unknown…

    At least one council member, President Barb Johnson (4th Ward), was miffed by the study. She noted that RCV’s supporters had promoted the system by saying it would draw out more voters and cost less than a traditional primary-plus-general election system. Considering the study’s results and last year’s very low voter turnout, she said, “all of these things did not happen in our city.”

    “It is disturbing to me that we’re talking about an extra quarter of a million dollars for a system that was supposed to decrease our costs,” Johnson said.

    Find the report at

  13. IRV = THE method to elect Stalin/Hitler TYRANT extremists into single offices when the middle is divided.

    34 H-M-S
    33 S-M-H
    16 M-H-S
    16 M-S-H

    IRV fanatics are as math CRAZY as the old flat earth fanatics.

    IRV IGNORES most of the data in a place votes table — which means ZERO to fanatics.


    H 34–16–49

    S 33–16–50

    M 32–67–0

    Sorry – this is another age of FANATICS — as in the 1930s.

    NO reasoning is possible with FANATICS.

    See the BARBARIAN fanatics in Asia — at WAR with Western Civilization.

    P.R. and nonpartisan App.V.

  14. #8, the amendment restricting the IRV option to elections entirely within one county was not actually introduced because the bill died first. Narrowing the scope of the bill in this way was part of an effort to get the cost to the state down close to zero, which is about what it takes in California these days to get bills that are not part of some omnibus package out of the appropriations committees.

    #3, as already pointed out, the Irish have proven that you can count IRV at several locations. The mechanics of this at the precinct level would be inordinately difficult, but for two or several counties to coordinate at the county level is quite reasonable. In addition, there are multiple approaches to doing post-election audits down to the precinct level.

    Everyone, please keep in mind that for the last several years getting anything at all through the California legislature with a price tag over about $5 has been a Herculean effort. This bill was one of a couple hundred (literally) that died in Senate Appropriations yesterday and were buried in a mass grave called “held in committee”. This doesn’t mean no one was interested in them. It just means the state has no money.

  15. #11 We are talking about special elections.

    California currently uses an odd system for special elections:

    In the special primary, all candidates regardless of party are on one ballot, and all voters can vote for any candidate regardless of party. If one candidate receives a majority of the vote they are elected. If all candidates are from one party, the plurality winner is elected – it is treated as a party primary, with only one candidate qualifying for a special general election.

    Otherwise, the top candidate from each party, along with all independent candidates advance to the special general election (this is the simplified version).

    So you could have one candidate with 48.8%, another with 36.5% of the vote, a 3rd with 8.7%, and a 4th with 6.1%. If the 1st, 2nd, and 4th were from the same party, only 1st (48.8%) and 3rd (8.7%) would advance to the special general election, and 2nd and 4th (36.5% and 6.1%) would be eliminated.

    Under the IRV bill, if no candidate received a majority of the vote, then trailing candidates would be eliminated one by one.

    So by converting to IRV, a different winner might result. Whether IRV was used or not would be determined after the vacancy occurred, and would be determined by county officials in a period of 35 days, which would (a) delay the special election; (b) allow local political concerns to influence the decision.

    The bill would also permit use of 3-rank IRV, which would be especially bad if a party had more than 3 candidates.

    Under Proposition 14, special elections would be run similar to regular Top 2 elections, with all candidates appearing on the special primary ballot. If one candidate received a majority, he would be elected, otherwise the Top 2 would advance to the special general election. So a special election would be just like a regular election, except that there is a possibility of a majority in the election in the primary.

    So if you convert this to an IRV election, you aren’t as dramatically changing the format of the election, and are less likely to produce a different result.

    You won’t give local officials the opportunity to game the system by choosing a partisan or nonpartisan election. It would simply be the usual system on an accelerated time scale.

    But with a special election, there is really no need for complicated mechanical voting systems. A particular problem that Minneapolis had was that they had 5 races on one ballot (and they also had two actual STV multi-seat races). Since an IRV or STV counting involves shuffling the ballots, you can’t count one race until each of the others have been resolved, or you have to limit voter choice to simplify machine counting.

    But in a special election you only have one race and can use a paper ballot that lists all candidates, and voters simply number the candidates, 1, 2, 3, 4, etc. You can use all-mail elections (the most recent special legislative election had 73% by-mail voting). Since all completed ballots are returned to a central location within a county, it simplifies counting.

    Instead of hiring 100s of election judges to sit around in polling places for a typically light turnout, you hire a similar or smaller number to serve as counting judges.

    Most special elections have relatively light turnout, and voting is quick since there is only one race on the ballot, so you pay election judges to sit around knitting.

    Because you would standardize the process and don’t require special technology, you don’t have to let county officials decide whether or not to use IRV, which may be politicized.

    The reason that there was controversy over the schedule for the special election in SD 15 was because Democratic legislators were trying to delay the election until after the budget was determined. The 2/3 requirement, is 2/3 of the legislators elected. There are 40 senators. 2/3 of 40 is 27. If one seat is vacant, then 2/3 of 39 is 26.

    So Proposition 14 would create an election system that is consistent with IRV. If one can vote for any candidate without regard to their party preference, it is reasonable to be able to rank any candidate without regard to party preference.

    You produce a majority winner, but you produce it quicker, which is of extreme importance in filling a legislative vacancy. While there is value in letting voters reconsider their choice in a conventional runoff, I believe that in the case of special elections, resolution in an expeditious fashion is more important.

  16. To get a sense of how much you can trust Joyce McCloy, consider her post this week at her “throw-lots-of-mud-and-hope-something-sticks” anti-iRV blog:

    In it, she says Sunnyvale (CA) has “scrapped IRV for mayor” after “using it,” and goes through several other alleged “repeals.”

    But… Sunnyvale never used IRV! And the rule it changed was an __internal__ one for the seven-member city council to use IRV to pick a mayor from its own members. One of the councilors she quotes tried to correct her on the site, but to no avail. She also didn’t change it after I explained her errors and how they reflect her approach that “if a URL says it, it must be true.” See:

    I assume the “Sunnyvale repeal” will join the “Georgetown University repeal” (oops, Georgetown uses IRV), “Aspen repeal” (Aspen will use IRV next year unless voters approve a charter amendment that may or may not be on the NOvember ballot) and so on.

    Here she goes after Minneapolis, even though the majority opinion (both among voters acording to surveys and among city leaders) is that IRV was a success last year. Consider this quote from former elections director Patrick O’connor, who oversaw the implmentation of IRV: “I have had the great fortune to be a small part of what could easily be considered the most significant civic exercise in the history of Minnesota government: the implementation of the first Ranked Choice Voting election in Minneapolis and in Minnesota. We proved that it could be well administered, quickly and accurately counted, and that voters had little problem with the concept.”

  17. Candidate/Incumbent rank order lists of replacements during a term of office for legislators — who MUST be at 100 percent 24/7.

    Think 7 Dec 1941 and 911.

    NO special elections are needed.

  18. What happened to IRV in Burlington, VT — one of the centers of so-called progressive politics ???

  19. #16 40% of voters in Minneapolis voted for

    1) Rybak; 2) blank ; 3) blank

    Rybak had 74% of the vote, the 2nd place candidate 10%, and it only took several weeks to figure that out. I suppose “quickly” meant that it didn’t take as long as the 2008-2009 Senate “race”.

    The survey of “randomly selected” voters showed 54% of registered voters voted in the Minneapolis election. Actual data from the election showed 19% voted. So how do we explain this?

    a) In Minneapolis, voter participation is highly skewed toward those wealthy enough to own personal telephones. Over 1/2 of the top quintile who own personal phones voted, while only about 10% of the phoneless underclass that comprise fully 4/5 of Minneapolis voters went to the polls.

    Someone who is unfamiliar with a telephone keypad may not understand the concept of 1,2,3. Older, whiter voters who have used dial phones may have an advantage, since the distance the dial travels is proportional to the magnitude of the number.

    b) Some poll respondents may have lied about their voting. So let’s say you pick 6 people at random, and asked them if they voted. 3 say they did. 2 of those lied. You ask them about their voting experience (you ask a different set of questions who didn’t vote). Is someone who didn’t vote, but lied about it, going to give you an accurate assessment of how helpful the election judges were, or whether they ranked multiple candidates (which was also over reported), or it was easy to vote?

    c) The sample was self-selected. Phone numbers were dialed at random and voters asked whether they wished to participate in the survey. Those who voted, may have been more willing to participate. But what if those who liked IRV were also more willing to participate? And those who did not vote may try to rationalize their non-participation.

Leave a Reply

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