[EL] RCV in Maine

Sun Nov 1 06:40:36 PST 2020

All,

I just want to touch upon the time it takes to tally RCV/IRV ballots in Maine. The problem is not the tabulation, of course- that is extremely trivial and can be done very quickly and accurately. The problem is aggregating the ballots, preserving chain of custody, and transporting them by secure carrier, which takes several days. If it goes to tabulation, we will not know who won for at least 4 days, probably longer. It is slower than “regular” ballots to get an official count, though it hasn’t impacted the time it takes to officially certify the results.

Sent from my iPhone

On Nov 1, 2020, at 5:37 AM, Rob Richie <rr at fairvote.org> wrote:

Up early. as I adjust to the new clock. I’ll take a bite at Sam's interesting question.

I would say this prospective outcome is consistent for whom these voters are against. Consider a situation like the one Sam describes, where there is a strong independent candidate who is able win the instant runoff in a congressional district, but not statewide. Let’s give the candidates names drawn from the Star Wars universe:
Candidate A - Darth Vader
Candidate B - Luke Skywalker
Candidate C - Hans Solo

In the single-choice, plurality voting method, suppose Darth Vader gets 40% in both the statewide vote and in the 2nd congressional district. The remaining 60% is split differently statewide (with Skywalker getting 32% and Solo 28%) and in CD 2 (with Solo getting 32% and Skywalker 28%). Under single-choice rules, Vader sweeps all three electoral votes. Under ranked choice voting -- assuming general cohesion between backers of Solo and Skywalker despite some historic tensions between the candidates -- Skywalker gets the two statewide electors and Solo the one elector from CD2.

What was consistent for these voters is that they blocked Darth Vader from winning any electoral votes. That’s a good outcome for them -- much fairer than a candidate opposed by a majority winning all of Maine’s electoral votes.

Lest one think that ranked choice voting only helps rebels against empires, by the way ,it all depends on the context. Putting this in terms of Republicans and Democrats, one can imagine Republicans being relieved to have ranked choice voting in place when a Libertarian or Constitution Party candidate is on the ballot, while Democrats will be glad to have it when Green candidates on the ballot. Looking forward, there is no way to know where fissures in a party's support may come from -- so 2021 is  a great time for policymakers to get a more reliably representative system in place before the next presidential election.

Rob

On Sunday, November 1, 2020, Samuel S. Wang <sswang at princeton.edu<mailto:sswang at princeton.edu>> wrote:
I depart from what all of you are writing about, but...

Come to think of it, RCV has an interesting quirk when it collides with a Maine-style allocation of electoral votes by Congressional district.

Under RCV rules, a citizen's vote could be construed in two ways at the same time. Imagine if a third-party candidate had an extreme concentration of support in one of the Congressional districts. Then she/he would advance to the final round in that district, but not at a statewide level. Then some citizens could have their vote count in support of that candidate for the district-level elector, but simultaneously be counted for a major-party candidate at the statewide level.

I do not think this will ever happen, but it is quite a novelty to imagine such a scenario.

-Sam

>>>

Prof. Samuel S.-H. Wang

Neuroscience Institute, Washington Road

Princeton University

Princeton, NJ 08544

Office: (609) 258-0388

Virtual office: http://princeton.zoom.us/my/samwang

Neuroscience: synapse.princeton.edu<http://synapse.princeton.edu>

Elections: election.princeton.edu<http://election.princeton.edu>

Redistricting: gerrymander.princeton.edu<http://gerrymander.princeton.edu>

________________________________
From: Law-election <law-election-bounces at department-lists.uci.edu<mailto:law-election-bounces at department-lists.uci.edu>> on behalf of Rob Richie <rr at fairvote.org<mailto:rr at fairvote.org>>
Sent: Saturday, October 31, 2020 11:00 PM
To: Kogan, Vladimir <kogan.18 at osu.edu<mailto:kogan.18 at osu.edu>>
Cc: Election Law Listserv <law-election at uci.edu<mailto:law-election at uci.edu>>
Subject: Re: [EL] RCV in Maine

Waiting for all ballots to come in before doing the ranked choice tally is important for hand tallies, but not with scanned ballots where running a ranked choice tally is essentially as easy as pushing a button.

It is a common approach in the California cities with RCV, for example, to run RCV tallies on election night two or three times, and then once a day until ballot scanning is finished.

This year, Maine should have more than 90% of its ballots centrally scanned within two or three days of the election, and it would be a very meaningful - and straightforward - exercise to see what  the instant runoff tally is at that time. Of course the final tally would remain determinative.

FairVote just released a  SurveyUSA poll of the state of the Maine RCV races this week, and presents its findings in a nifty interactive format at this webpage:
https://www.fairvote.org/maine_poll_october2020

On election night, we will be able to use that data and historic voting patterns to do a forecast of what the instant runoff results might be when all the ballots are counted. It’ll be posted at fairvote.org<http://fairvote.org>

Notably, ranked choice voting in our survey had no effect on the margin in the presidential race in a second congressional district for the one electoral vote to be decided there. Major party candidates are affected by ranked voting tallies based on the nature of the voters backing the third-party independent candidate along with their willingness to engage with the concerns of those voters.

It will be a helpful contrast to show how many votes are preserved in the instant runoff tally in the US Senate race in Maine compared to the huge drop in turnout we can be sure to see in Georgia when it has its US Senate runoff in January. Voting in Georgia is hard enough without having to ask people repeatedly to come back to cast ballots instead of just offering them a chance to provide their back up choices with a ranked ballot.

Rob

On Saturday, October 31, 2020, Kogan, Vladimir <kogan.18 at osu.edu<mailto:kogan.18 at osu.edu>> wrote:

I wanted to push back a little on Rick’s comments regarding RCV in Maine and the plan to wait until all towns get their votes in before tabulating the second round votes. Rick wrote: “This delay is not at all intrinsic to RCV. That process can be completed quickly, if not for the Secretary of State’s policy choice in Maine to wait in this way.”

I actually do think this approach is intrinsic to RCV. We could imagine that in a really close, multi-candidate election (e.g., a San Francisco mayoral election with nearly 20 candidates running), it may not be clear who the lowest voter-getter will be until all of the votes have been counted. Without knowing who came in last, we would not know who to eliminate first and whose first-round votes to redistribute. And even if the lowest vote-getter is clear, the differences in the vote counts of other candidates could also be closed, determining the order in which they are eliminated.

Of course, there will likely be little doubt that Savage ultimately comes in last in this particular election, but we could imagine different elections with different mixes of candidates playing out differently. Waiting until all of the votes are is the only prudent general policy that can work for all elections under RCV, it seems to me.

Vlad

Why Maine’s Senate Race Will Likely Not Be Officially Resolved Until A Week or so After Nov. 3rd<https://urldefense.com/v3/__https:/electionlawblog.org/?p=117884__;!!KGKeukY!g5VJxhRKE5d0Ae9rE5g2RgmGMk2dkXyww6FjmoO7mkPC7Rq7320IVTbOORLWydvT$>

Posted on October 31, 2020 8:32 am<https://urldefense.com/v3/__https:/electionlawblog.org/?p=117884__;!!KGKeukY!g5VJxhRKE5d0Ae9rE5g2RgmGMk2dkXyww6FjmoO7mkPC7Rq7320IVTbOORLWydvT$> by Richard Pildes<https://urldefense.com/v3/__https:/electionlawblog.org/?author=7__;!!KGKeukY!g5VJxhRKE5d0Ae9rE5g2RgmGMk2dkXyww6FjmoO7mkPC7Rq7320IVTbOOanDP0lu$>

Maine’s Senate race is considered close<https://urldefense.com/v3/__https:/www.politico.com/news/2020/10/30/the-independent-that-could-decide-the-senate-433629__;!!KGKeukY!g5VJxhRKE5d0Ae9rE5g2RgmGMk2dkXyww6FjmoO7mkPC7Rq7320IVTbOOVVrBUvP$>, and if so, there’s an additional reason it could take longer — maybe several days — to determine who has won. The reason is that Maine now uses ranked-choice voting (RCV), and there is an independent, Green New Deal candidate, Lisa Savage, polling<https://urldefense.com/v3/__https:/www.fairvote.org/maine_poll_october2020__;!!KGKeukY!g5VJxhRKE5d0Ae9rE5g2RgmGMk2dkXyww6FjmoO7mkPC7Rq7320IVTbOOQQKh1uk$> in the 4-5% range. If the margin between Susan Collins and her Democratic opponent, Sara Gideon is less than that when the initial count is completed, the Senate outcome would then be determined in the second round of the ranked-choice tabulating process, when Savage will be eliminated and the second-ranked choices on those ballots will be distributed either to Collins or Gideon.

The assumption is that most of Savage’s voters will rank Gideon as their second choice, since Savage is the furthest left candidate in the race. But Maine has structured that process, as I understand it, so that it won’t get to that next stage for nearly a week. I’ve been told (maybe someone has a story to link to for this) that the Secretary of State will not turn to the second round until the vote totals from every town in the state are in; since small towns take several days to finish completing that count, that means the RCV process won’t start until the slowest town has finished completing its tallies.

This delay is not at all intrinsic to RCV. That process can be completed quickly, if not for the Secretary of State’s policy choice in Maine to wait in this way.

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Rob Richie
President and CEO, FairVote
6930 Carroll Avenue, Suite 240<https://www.google.com/maps/search/6930+Carroll+Avenue,+Suite+240+%0D%0ATakoma+Park,+MD+20912?entry=gmail&source=g>
Takoma Park, MD 20912
rr at fairvote.org<mailto:rr at fairvote.org>  (301) 270-4616  http://www.fairvote.org
FairVote Facebook<https://www.facebook.com/FairVoteReform>   FairVote Twitter<https://twitter.com/fairvote>   My Twitter<https://twitter.com/rob_richie>

Thank you for considering a donation<http://www.fairvote.org/donate>. Enjoy our video on ranked choice voting<https://youtu.be/CIz_nzP-W_c>!

_______________________________________________
Law-election mailing list
Law-election at department-lists.uci.edu
https://department-lists.uci.edu/mailman/listinfo/law-election
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20201101/8a846cca/attachment.html>