[EM] Historic opportunity in Arizona for Approval Voting

[Abd ul-Rahman Lomax]

Subject was: Re: [EM] Historical perspective about FairVote organization

At 12:32 PM 3/17/2013, Richard Fobes wrote:

>I agree that using better ballots and better vote-counting methods 
>in real situation -- using real data -- is essential for making real progress.

A very unusual opportunity for voting system reform has recently presented.

The Arizona House of Representatives just passed HB2518, which 
provides an approval voting method for use in municipalities which 
choose to adopt it.

For background, all Arizona municipalities but one hold nonpartisan 
elections. The exception is Tucson, which holds party primaries and 
then partisan elections. (A "partisan election" has the candidate's 
party affiliations on the ballot, and, also, provides for ballot 
placement for parties, based on petitions and continued election perfomance.)

The measure allows voters in a "primary or first election" to vote 
for more than one candidate and then the top two candidates go on to 
the second election.

>1 Be it enacted by the Legislature of the State of Arizona:
>2 Section 1. Title 16, chapter 4, Arizona Revised Statutes, is amended
>3 by adding article 8.2, to read:
>4 ARTICLE 8.2. OPTIONAL CITY AND TOWN APPROVAL VOTING
>5 16-559. City and town approval voting; requirements
>6 A. NOTWITHSTANDING ANY OTHER STATUTE, A CITY OR TOWN IN THIS STATE MAY
>7 BY ORDINANCE ESTABLISH AND USE A SYSTEM OF APPROVAL VOTING IN THAT CITY'S OR
>8 TOWN'S PRIMARY OR FIRST ELECTION. AN APPROVAL VOTING SYSTEM SHALL PROVIDE
>9 FOR THE FOLLOWING:
>10 1. THE VOTER IN THE PRIMARY OR FIRST ELECTION SHALL BE PERMITTED TO
>11 VOTE FOR AS MANY CANDIDATES FOR A SINGLE OFFICE AS THE VOTER CHOOSES TO
>12 APPROVE.
>13 2. THE TWO CANDIDATES WHO RECEIVE THE HIGHEST AND SECOND HIGHEST
>14 NUMBER OF VOTES IN THE PRIMARY OR FIRST ELECTION SHALL ADVANCE TO 
>THE GENERAL
>15 OR RUNOFF ELECTION FOR THAT CITY OR TOWN WITHOUT REGARD TO WHETHER ANY ONE
>16 CANDIDATE HAS RECEIVED A MAJORITY OF THE VOTES CAST FOR THAT OFFICE.
>17 3. THE BALLOT AND ALL OTHER VOTING MATERIALS SHALL CLEARLY INDICATE
>18 THAT THE VOTER MAY VOTE FOR AS MANY CANDIDATES IN THAT ELECTION 
>AS THE VOTER
>19 CHOOSES, AND THAT THE CANDIDATES WHO RECEIVE THE TWO HIGHEST 
>NUMBER OF VOTES
>20 SHALL ADVANCE TO THE GENERAL OR RUNOFF ELECTION.
>21 B. EXCEPT AS OTHERWISE PROVIDED IN THIS ARTICLE, CITY AND TOWN
>22 APPROVAL VOTING ELECTIONS SHALL BE CONDUCTED IN ACCORDANCE WITH THE
>23 PROVISIONS OF ARTICLE 8 OF THIS CHAPTER.
>24 16-559.01. Approval voting; charter; ordinance
>25 THIS ARTICLE DOES NOT REQUIRE A CITY OR TOWN TO ADOPT AN APPROVAL
>26 VOTING SYSTEM, BUT A CITY OR TOWN MAY AMEND ITS CHARTER IF 
>REQUIRED FOR THAT
>27 CITY OR TOWN TO ADOPT AN ORDINANCE TO IMPLEMENT AN APPROVAL 
>VOTING SYSTEM AS
>28 PRESCRIBED BY THIS ARTICLE.

Municipalities in Arizona have great flexibility in choosing their 
own voting systems. I have looked at four cities in Arizona, and I 
have found two systems in use. Tucson is the only municipality which 
runs partisan elections.

In Tucson, there is a regular party primary, and the winner of the 
primary advances to the general election. So Tucson runs five 
elections each election cycle.

In all other towns, that I've checked, there is currently top-two 
runoff. It is no easy to check small-town elections, some of those 
may simply be plurality. My experience in small towns is that often 
there is only one candidate anyway....

It is possible that municipalities could tweak this, or even do 
something different. Tucson has demonstrated that it can defy the 
state; the legislature attempted to prevent Tucson from holding 
partisan elections, and Tucson won in court.

HB2518 has the support of the majority of the Republican Party in 
Arizona. The vote in the House was, Clay Shentrup's analysis:

>Republicans:
>Yes: 31
>No: 4
>Not voting: 1
>
>Democrats:
>Yes: 0
>No: 22
>Not voting: 2
>
>So 86% of Republicans voted yes. So what are prospects in the senate?
><http://www.azleg.gov/MemberRoster.asp>http://www.azleg.gov/MemberRoster.asp
>
>17 Republicans
>13 Democrats
>
>86% would be 14 Republicans. JUST ENOUGH to pass.

I did more research. This bill was introduced as HB2569 in last 
year's legislature. See
https://groups.google.com/d/msg/electionscience/QTpV0Qr3Tfo/POWiz0m15NcJ 
for coverage of HB2518 committee actions in the 2013 session, and
https://groups.google.com/d/msg/electionscience/QTpV0Qr3Tfo/X9fgzrTPOoEJ 
for the Judiciary Committee in 2012.

Three Democrats approved this bill in the committee vote in 2012 
(which was unanimous, 7-0). Two of them left the legislature, the one 
that remained (Hale), voted in the first committee meeting in 2013, 
No. And then in the second, Yes. And then on the floor, No again.

I have been unable to find any coherent criticism of this bill. What 
was semi-coherent, *one blogger*, I addressed in
https://groups.google.com/d/msg/electionscience/QTpV0Qr3Tfo/eaDRR4WLDAEJ 
(which was written just before the previous two linked posts).

I'd like election methods activists to realize the potential here.

1. If this passes, Phoenix might adopt it. Phoenix is a much larger 
city than any in the U.S. that have adopted IRV. It's quite possible 
that the *momentum* that FairVote has been claiming for IRV could 
vanish in the dust of this. If, say, Mesa, Arizona, adopts the 
procedure, the population of towns using Approval Voting would pass 
that of all those using IRV (in the U.S.) Mesa is the home town of 
the representative who introduced the bill, Mesa is just south of 
Phoenix, part of the Phoenix Metropolitical Area.

2. This method wtih a two-stage election is actually a highly 
sophisticated election method. It's been shown to be very strong 
through simulations, in terms of Bayesian Regret. Range in general is 
improved, in the presence of "strategic voting," by a two-stage election.

3. It appears that this is *not* a runoff-if-no-majority system, but 
the language does seem to call the second election a runoff. It is 
possible that a municipality, implementing it, could decide to 
consider the primary complete if a majority is found, but given that 
the "runoff" would be held with the general election, the system, 
with the understanding that there is *always* a second election, will 
not add much expense. The runoff is not a special election, the 
primary is. And cities in Arizona can decide to use mail balloting.

4. Approval in a primary system is a fix for the Center Squeeze 
problem that top-two runoff normally suffers (like IRV). Even better 
would be Bucklin, which would be a next reform, that allows Ranked 
Approvals. Bucklin is essentially Instant Runoff Approval, and the 
mandatory runoff, if kept, would address certain complaints about 
Bucklin. Bucklin would encourage additional approvals.... (as it did, 
historically.)

5. The passage of this bill in the Arizona House is the best news 
I've seen *ever* as to U.S. voting systems.

6. We need to encourage contacts in Arizona, and especially 
progressives, i.e., Democrats and Greens, etc., to support this 
legislation. It is *not* what that lone Democratic blogger called it, 
some Republican trick. (It's *conceivable* that some Republican or 
other thinks it would advantage their party, but, as written and as 
it would be applied, it would not.)

The monolithic opposition of the Democratic Party in the floor vote 
is mysterious. There was no maintained opposition in committee by 
Democrats, but some wavering. It's looking like "party discipline" 
was enforced, i.e., some Democratic minority leader made a decision 
and made it important for all Democrats to vote together. But ... why?

Had there been this kind of system in, say, Florida in 2000, Gore 
would have won, it's nearly certain. Approval, in general, does fix 
the spoiler problem, to a degree. (Bucklin, i.e., Instant Runoff 
Approval, would do it better.) But this bill does not apply to 
partisan primaries and elections. It wasn't designed for that. If it 
were to be applied to party primaries, as in Tucson, the election 
would have to be ordinary one-round Approval. Very simple, and 
definitely an improvement.

There has been some confusion with a failed Proposition in Arizona 
that attempted to require all *partisan* primaries to be "open," 
i.e., unified, all voters vote in it, not just party voters. It 
*only* applied to primaries where the candidate's party is listed the 
ballot, and, except for Tucson, that is only for state elections. (I 
think it didn't apply to federal elections.) *That* was a power grab, 
possibly, or maybe just stupid.