Arizona Bill to Allow Approval Voting in Non-Partisan City Elections Advances

On March 11, the Arizona House passed HB 2518, which allows local elections to use Approval Voting. On March 19, the bill passed the Senate Elections Committee. The vote in the House was a partisan vote, in which Republicans generally supported the bill and Democrats opposed it. Here is a story that explains the bill and why it has substantial support. Thanks to Clay Shentrup for news about the bill.


Arizona Bill to Allow Approval Voting in Non-Partisan City Elections Advances — No Comments

  1. “Rep. Justin Olson, R-Mesa, said problems arise when there are perhaps eight candidates with a conservative point of view and just two who are considered liberals. With voters entitled to cast ballots for only one, Olson said that creates a “spoiler effect,’ where the conservatives split the vote and the runoff in what might be a conservative community ending up between the two liberals.”

    Yeah, God forbid a shithole like Mesa should elect a mayor who had any brains. Well, it’ll never happen because Mesa, like 98% of Arizona, is filled with ignorant morons. They’re not smart enough even to understand “approval voting.”

  2. Thanks for reporting this. It should be noted that an amendment has been made which also allows Instant Runoff Voting (IRV). It would be kind of a weird use, since it would have to be used to pick the two finalists for the runoff/general election. Maybe it could just be the winner and runner-up. Probably the language will be vague about that and leave it up to cities to decide.

    While some might see this as a Republican power grab, there are complex reasons why we at The Center for Election Science (a non-partisan 501(c)3) believe this will have enormous benefits. For one, even if they don’t win, minor party and independent candidates will see a more accurate reflection of their support. And multiple similar candidates can run without being spoilers for each other.

  3. For any ignorant MORONS on this list —

    Approval Voting — vote for 1 or more — who are thus *approved* / ACCEPTABLE by each voter.

    The candidate(s) with the highest total approval votes is/are elected.

    — should be ONLY for executive/judicial offices – 1 or more.

    Pending major head to head voting education.

    P.R. for legislative body elections.

  4. This kind of election system would go a long way towards having a more representative government. I really hope they pass this and that more localities in the country pass this sort of thing.

  5. Arizona city council elections are either by district, or at-large, typically with half (3 members) elected for a four-year term every two years. They also have an elected mayor.

    When three members are elected at-large, each voter gets three votes in the primary, with a majority of ballots cast needed for election. If more than three receive a majority, then only the top 3 are elected.

    If 2 candidates receive a majority, then the 3rd and 4th candidates, advance to the general election. If 1 candidate receives a majority, then the 2nd through 5th advance to the general election. If no candidate receives a majority then the Top 6 advance to the general election.

    Presumably, under an approval primary, the Top 6 would always advance to the general election.

    Arizona voters are already accustomed to voting for multiple candidates in an election, including for state representative.

    Typically, there might be between 5 and 9 candidates seeking election. There may be a couple who badly trail, and perhaps an incumbent who is elected in the primary. I doubt that for multi-member non-partisan elections, an approval primary will make much difference. Approval voting might move someone from 7th to 6th and on to the general, but that is unlikely to get them elected.

    For single-member districts and mayoral contests, 1 or 2 candidates is the norm, with 4 or more quite unusual.

    Conceivably an approval primary may change the result in races with 3 or more candidates, if everybody’s second choice can advance. This depends on whether voter’s rankings mean:

    1. Adam Adams – this is who I want.
    2. Bill Billings – OK, but I like Adams a bit better.
    3. Carl Carlson – I’ll move if he is elected.


    1. Adam Adams – this is who I want.
    2. Bill Bennings – Don’t really know anything about him.
    3. Carl Carlson – I’ll move if he is elected.

    If there are a significant number of voters whose evaluation is the opposite (that is, a polarized race between AA and CC), then plurality, conventional runoff, or instant runoff will produce the same result.

    Approval has the potential of identifying a consensus or compromise candidate. But if voters don’t know anything about BB, they probably won’t approve him, and the result under Approval would be no different than plurality, conventional runoff, or instant runoff. But this is reasonable. “No opinion” is not a neutral opinion or equivalent to “I guess he’s OK”, and BB is not really a consensus candidate.

    IRV for multiwinner races is complex, requires new ballot formats, requires drastic new behavior on the part of voters, and doesn’t produce better results.

  6. “Only a knee-jerk partisan could object to this legislation.”

    It passed the Arizona House with ZERO votes from Democrats. Presumably because it was proposed by a Republican and therefore it “must be bad”. 🙁

    The Arizona League of Women Voters were originally against the bill, apparently because their president (who is also the head of the AZ branch of FairVote) believes Approval Voting is “too susceptible to tactical voting”. Nevermind that election researchers widely consider Approval Voting to be one of the most resistant to tactical voting; the really shocking thing here is that the status quo is PLURALITY VOTING, which is known to be pretty much the worst voting system ever seriously proposed.

    The addition of IRV to the bill will supposedly help draw Democrat support, and has changed the AZ LWV from “against” to “neutral”. All this strains the comprehension of election method experts.

  7. #7 The LWV is “neutral” because the bill would give cities the option of choosing approval voting. Barbara Klein said that the LWV would favor the bill if approval voting was removed as an option.

    This would appear to misrepresent the League’s official position which is:

    “Supports giving Arizona voters the option of more choice among election systems.”

    Andrew Jennings was quite polite, and said that he looked forward to debating before city councils which option to adopt.

    Barbara Klein claimed that Approval Voting violated “One Man One Vote”. She quite clearly does not understand the concept. Before ‘Reynolds v Sims’ each voter (man) was able to make one mark (vote) on a ballot. But what the Supreme Court actually ruled was that where a voter lived in a State determined how much influence the voter had over who represented them.

    The Supreme Court wrote:

    “There is more to the right to vote than the right to mark a piece of paper and drop it in a box or the right to pull a lever in a voting booth. The right to vote includes the right to have the ballot counted. It also includes the right to have the vote counted at full value without dilution or discount.”

    For example, Arizona CD 6 has 232,072 voters; while Arizona CD 7 has 410,018. A voter in CD 7 has 43% less influence than his fellow Arizonan in CD 6.

    But Approval Voting does not have this problem. Each voter has an equal opportunity to cast an effective vote.

    IRV has more potential for voters to cast an ineffective vote, particularly IRV-3 which restricts the right of a voter to rank all candidates. Many voters in San Francisco simply don’t understand the system, either ranking the same candidate 1,2, and 3; or ranking the same candidate 1 and 3, but some other candidate 2nd.

    While IRV is described as analogous to a conventional runoff, it is not. In a conventional runoff the voters know which candidates they may vote for in the runoff.

    In IRV they have to be able to forecast the result. Voters are told that it is up to them how many candidates to rank. The example of Florida 2000 is given, where it is presumed that Nader voters would have voted 1. Nader 2. Bush or Gore. 3. Gore or Bush.

    But in real elections, cast votes like: 1. Nader 2. Buchanan 3. Nader


    1. Gore 2. Nader
    1. Bush 2. Nader

    While these voters might well have regarded Nader as being acceptable, they feel compelled to vote tactically to block Bush or Gore. While the voters may be expressing an opinion that Nader is acceptable (or they approve of his candidacy) that information is ignored by IRV.

    IRV-3 is susceptible to vote splitting. For example in the Supervisor District 10 race in San Francisco, there were about half a dozen candidates representative of one “faction” and two candidates representative of another “faction”. Transfers were largely between candidates of the two factions – though the two candidates of the minor faction really didn’t support each other. Vote splitting, and inability of voters to express enough preferences resulted in massive number of ballots being discarded as “exhausted”. It is only because the first faction greatly outnumber the second faction that a candidate of the first faction eventually prevailed.

    The fact that the vote-counting program does not coredump does not indicate that voters are not confused. It simply indicates the vote-counting software is robust. When voters cast overvotes in Florida 2000, the software had no problem detecting them as overvotes. This doesn’t mean that some voters were not confused.

    The testimony of the representative of association of counties was quite interesting. Before the house committee she simply noted that because cities will be required beginning in 2014 to hold their elections coincident with the even-year fall elections (August-November) conducted by counties, that there would be two types of elections. But she apparently recognizes that the election software already handles multiple votes for an office (including for the legislature).

    She was quite alarmed by the addition of the IRV amendment (and 11-page addition to an original 18-line) which she didn’t see until just before the senate committee hearing. She particularly noted that IRV requires a determination of the total number of valid votes before counting could begin. She also realized that the ballot format would be greatly complicated.

    My guess is that there will be a floor amendment to strip out approval voting in favor of IRV.

  8. Jim,

    I wouldn’t be too hard on Barbara. She’s the head of the AZ chapter of FairVote, as I understand it, and so she’s surely heard a tremendous amount of anti-Approval FUD from FairVote leaders. E.g. the kind of stuff I discuss here:

    I actually sent her a copy of “Gaming the Vote”, and she sent a very friendly reply. I think she’s more open-minded than we might be giving her credit for.

    Further, we received a pretty cordial email from FairVote director Rob Richie. I think this bill has demonstrated to him that the Approval Voting community has become serious, politically savvy, and pragmatic. Thus we’ve earned a little more respect from Rob. Further, he realizes that it isn’t a zero-sum game. In other words, a city that enacts Approval Voting wasn’t necessarily a loss for IRV. That community may have not accepted IRV anyway. And vice versa.

    It would take years of better behavior for me to really trust FairVote, but a compromise bill like this may be a start to more collaboration and less infighting between proponents of different voting systems.

    That’s me trying to keep a positive outlook.

Leave a Reply

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