[EL] RCV in Maine and Third-Party Spoilers

Mon Dec 3 08:45:28 PST 2018

In exploring the ballot information that ME provides (they provide for each ballot the order of choices as well as when that choice was under-voted or over-voted), I can also tell that they took the 2nd choice for anybody that under-voted for their 1st choice and applied that in the second round as well.

By looking at that data (from the link below), you can also see examples of some oddities in how people treated the ballot. See images below (if you cannot see them, you can email me privately for them). Seems an electronic ballot might be the only way to prevent these errors (unless people were trying to communicate something with these odd patterns of choices).

Is anybody else working on this data? I may contact the state, but I cannot get the exact outcome they certified for the final result (earlier totals in the process work, but the final is off by just a few hundred).

[cid:image002.jpg at 01D48AF5.35282130][cid:image003.jpg at 01D48AF5.35282130]

Douglas R Hess
Assistant Professor of Political Science
Grinnell College
1210 Park Street, Carnegie Hall #309
Grinnell, IA 50112
phone: 641-269-4383

http://www.douglasrhess.com<http://www.douglasrhess.com/>

From: Jack Santucci [mailto:jms346 at georgetown.edu]
Sent: Friday, November 30, 2018 2:38 PM
To: David Lublin <dlublin at american.edu>
Cc: Election Law Listserv <law-election at uci.edu>; Hess, Doug <HESSDOUG at Grinnell.EDU>
Subject: Re: [EL] RCV in Maine and Third-Party Spoilers

Ditto on STV in US cities. Batch elimination of candidates below some threshold (e.g., 50 votes) was standard procedure in the 1910s-50s. I think (but don't quote me) that this threshold was as high as 2,000 in New York City.

Jack
Sent from my iPhone

On Nov 30, 2018, at 15:01, David Lublin <dlublin at american.edu<mailto:dlublin at american.edu>> wrote:
This is handled differently across countries. Ireland often eliminates multiple candidates who stand no hope of making the quota at the same time in their STV system. I'm not sure how they decide but think that the returning officers have some discretion. In Australia, they always report each transfer as each candidate is eliminated, which of course makes it easier to see how the preferences flowed. Even though it thwarts following preferences, the decision in Maine seems reasonable as no one would achieve a majority by elimination of only the last place candidate.

Best,
David Lublin

On Fri, Nov 30, 2018 at 2:30 PM Hess, Doug <HESSDOUG at grinnell.edu<mailto:HESSDOUG at grinnell.edu>> wrote:
Here’s a question about the Maine RCV election: Why do they eliminate two instead of just one at a time? Given how close the election was in Round 1 and the large vote for the losing candidate, you could have had (albeit very unlikely) Poliquin win with 51.7% of the vote if all of Hoar’s votes had transferred to him.

Granted, the probability of that happening in this situation approaches null, but you get the idea. Or am I missing something?

If they do check for that, or even if they don’t, it would be great to know how the votes transferred by each eliminated candidate instead of both together. I haven’t checked their other RCV races.

The table below is my arrangement of data from the Nov 26, 2018 file of certified results on the Maine SOS website. This means the totals are slightly larger than in Vladimir’s email, but the relationships are largely the same.

[cid:image001.jpg at 01D48AF4.7EA4F460]

Douglas R Hess
Assistant Professor of Political Science
Grinnell College
1210 Park Street, Carnegie Hall #309
Grinnell, IA 50112
phone: 641-269-4383

http://www.douglasrhess.com<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.douglasrhess.com_&d=DwMGaQ&c=U0G0XJAMhEk_X0GAGzCL7Q&r=2fMgMunsCtJpikIZXRvVAXXnXpXnW1DdeOa9_DBJVAg&m=K0zyd2Z4j71d02w628AD_m72Qvxaezv6R7sioWtTbJU&s=NT0GLzCbJIai0ZWyEKGwGSUWHu-a3BNQS-Fm0Pzkefw&e=>

From: Kogan, Vladimir [mailto:kogan.18 at osu.edu<mailto:kogan.18 at osu.edu>]
Sent: Monday, November 26, 2018 8:08 PM
To: Rob Richie <rr at fairvote.org<mailto:rr at fairvote.org>>
Cc: law-election at UCI.EDU<mailto:law-election at UCI.EDU>
Subject: Re: [EL] RCV in Maine and Third-Party Spoilers

Rob: Completely agree that off-cycle runoffs are horrible!

From: Rob Richie [mailto:rr at fairvote.org]
Sent: Monday, November 26, 2018 9:00 PM
To: Kogan, Vladimir
Cc: law-election at UCI.EDU<mailto:law-election at UCI.EDU>
Subject: Re: [EL] RCV in Maine and Third-Party Spoilers

I thought Vlad's email offers an opportunity to think why sometimes voters "exhaust" ballots with ranked choice voting or don't come back for runoff elections -- with, on the latter, the average drop inf participation in federal congressional runoffs this year being 47% of the first round, with half the runoff winners in fact earning fewer votes in the runoff than they did in the first round.

Although it's perhaps fun to look at the new kid on the block (ranked choice voting), I thought i'd look at the old standards of runoffs through the eye of exhaustion - -specifically the Top Two elections in California this year.

See some numbers below that I generated from the California Secretary of State's website. As of Nov. 26, with votes still being tallied<https://urldefense.proofpoint.com/v2/url?u=https-3A__vote.sos.ca.gov_returns_governor&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=4XOeQxhf0TSSUpky2UH4GkXicW7q7zmSroDzQAaIB0s&e=>. you'll see votes for governor (a Democrat-Republican runoff), Secretary of State (a Democrat-Republican runoff), and Lt. Governor (a Democrat-Democrat runoff).

My comparison isn't with the June preliminary round ("primary"), but within the same contest with voter reacting to the choices being offered to them. With the D-R runoff in statewide races, the dropoff from governor to those races is small. For Secretary of State, for example. it is only 1.53%. But for the D-D runoff for Lt. governor (and to a large, if lesser extent for US Senate), the dropoff is much bigger - 16.24% in the case of Lt. governor, or nearly 1.9 million votes.

If you think about who is likely dropping off, it's almost certainly Republicans. If you generously use the Cox vote in the governor's race as a the baseline for Republicans, that means that more than 40% of Republican voters likely skipped the Lt. Governor race -- they indicated a "pox on both your houses."

That is actually higher than the percentage of Maine independents who didn't indicate a Republican or Democrat as a backup preference. And I'm actually not surprised by that dropfff - there often is a core of third party and independent voters who really do think some version of "the lesser of 2 evils is still evil" or "don't encourage them with a vote."

So, then, the question is whether those voters are "spoilers" or really "abstainers." Are they fundamentally different than the some 40% of Maine's eligible voters who didn't vote at all? I'd suggest not. Just like those non-voters, they were "in play" for the candidates, and the candidates in fact did do enough to earn the votes of almost two-thirds of the independent candidate backers. The fact that they didn't earn the rest is something they'll need to reflect on, just like they need to reflect on not inspiring so many potential voters to get to the polls in the first place.

Speaking of runoffs as an alternative to RCV, I trust Vlad will keep an eye on the upcoming runoffs in Mississippi and Georgia. The Mississippi runoff turnout in the US Senate race may end up being pretty high , given the stakes -- we'll see. The Georgia turnout for next months' runoffs for secretary of state and another statewide office... not so much. You can be absolutely sure that a winner with RCV in the GA Secretary of State race this year would have won with a much larger number of votes than will end up being the case with runoffs.

- Rob Richie

CALIFORNIA GOVERNOR

Gavin Newsom (D)

7,213,464

John H. Cox (R)

4,481,280

Total Votes

11,694,744

CALIFORNIA SECRETARY OF STATE RACE

Alex Padilla (D)

7,385,637

Mark P. Meuser (R)

4,130,499

Total votes

11,516,136

Dropoff, votes

178,608

1.53%

CALIFORNIA LT. GOVERNOR RACE

Eleni Kounalakis (D)

5,542,766

Ed Hernandez (D)

4,253,319

Total votes

9,796,085

Dropoff, votes

1,898,659

16.24%

Dropoff of GOP Voters (est)

42.37%

(lt. governor dropoff divided by Cox vote)

On Mon, Nov 26, 2018 at 7:45 PM Kogan, Vladimir <kogan.18 at osu.edu<mailto:kogan.18 at osu.edu>> wrote:
I wanted to flag the election results in Maine’s 2nd Congressional District, where ranked-choice voting was used. Four candidates ran — one Democrat, one Republican, two independents. Here are the results:

Round 1

Round 2

Candidate Names

Votes

Percentage

Transfer

Votes

Percentage

Transfer

Bond, Tiffany L.

16260

05.73%

-16260

0

00.00%

0

DEM Golden, Jared F.

128999

45.48%

10232

139231

50.53%

0

Hoar, William R.S.

6753

02.38%

-6753

0

00.00%

0

REP Poliquin, Bruce

131631

46.41%

4695

136326

49.47%

0

Several things worth highlighting: (1) The “ballot exhaustion” rate was fairly low, about 2.8% of valid first-round votes; (2) the victor still did not win a majority of valid first-round votes; (3) by my calculation 35% (!) of voters who picked one of the independents as their #1 choice did not subsequently rank either the Democrat or Republican in a lower-ranked position.

I bring this up in response to the argument that RCV is a solution to the problem of third-party spoilers (who syphon off enough votes from one major-party candidate to lose him/her the election). Of course, all of those 35% may not have voted at all had the independent candidates not run; or perhaps they would’ve voted for Bond or Hoar even in the absence of RCV (assuming, of course, that Bond and Hoar would’ve still run had the election not been held under RCV…). But I thought this was worth flagging.

Note: This is not meant to be a general indictment of RCV or an argument that RCV is worse than the alternative.

Vlad

Error! Filename not specified.
Vladimir Kogan, Associate Professor
Department of Political Science
2004 Derby Hall | 154 N. Oval Mall, Columbus, OH 43210-1373
510/415-4074 Mobile
614/292-9498 Office
614/292-1146 Fax
http://u.osu.edu/kogan.18/<https://urldefense.proofpoint.com/v2/url?u=http-3A__u.osu.edu_kogan.18_&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=E9PE-hBrR59xkITyaLzqLfl2xYA6iLDLhy8UYxPRp9o&e=>
kogan.18 at osu.edu<mailto:kogan.18 at osu.edu>
Error! Filename not specified. at vkoganosu<https://urldefense.proofpoint.com/v2/url?u=https-3A__twitter.com_vkoganosu&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=Bcn8jGfoHw-0aLkLHassHRf5bLK_E2B7mhJuDwmCdks&e=>

_______________________________________________
Law-election mailing list
Law-election at department-lists.uci.edu<mailto:Law-election at department-lists.uci.edu>
https://department-lists.uci.edu/mailman/listinfo/law-election<https://urldefense.proofpoint.com/v2/url?u=https-3A__department-2Dlists.uci.edu_mailman_listinfo_law-2Delection&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=H2810E-P7HhBVnNlrbhXQoIdXuJqCkeAH6QhMeWvZwc&e=>

--
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
Rob Richie
President and CEO, FairVote
6930 Carroll Avenue, Suite 240
Takoma Park, MD 20912
rr at fairvote.org<mailto:rr at fairvote.org>  (301) 270-4616  http://www.fairvote.org<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.fairvote.org&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=WKCu8bELsho2NFoLB29O80gYRyPxgoPL42qz7KbNY7E&e=>
FairVote Facebook<https://urldefense.proofpoint.com/v2/url?u=https-3A__www.facebook.com_FairVoteReform&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=V2ehUZuBH3c1UGtAWqiTydat-CxTH3Y3r_xTmmT9VRs&e=>   FairVote Twitter<https://urldefense.proofpoint.com/v2/url?u=https-3A__twitter.com_fairvote&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=IqZ4oiS3SQc-YYnPMQD67mcbq0XYNRf4_MGif1ui-7s&e=>   My Twitter<https://urldefense.proofpoint.com/v2/url?u=https-3A__twitter.com_rob-5Frichie&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=ck_EeZrbaqtDGRldUppL-520yRRUy0Q1LhyFCOyDN_I&e=>

Thank you for considering a donation<https://urldefense.proofpoint.com/v2/url?u=http-3A__www.fairvote.org_donate&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=LtSHKPWztyfrWePHXcPDs5TZbjIDT_70t1Xh5GaoDgY&e=>. Enjoy our video on ranked choice voting<https://urldefense.proofpoint.com/v2/url?u=https-3A__youtu.be_CIz-5FnzP-2DW-5Fc&d=DwMGaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=toVAec3-Hd1lH-6OfgFj2QG2wvva5-kmJj0jrp32VbI&s=H82lrL8sdQFA8vEWUxOTiIidHoeA4-3pMANzPeZyPc4&e=>!
(Note: Our Combined Federal Campaign number is 10132.)
_______________________________________________
Law-election mailing list
Law-election at department-lists.uci.edu<mailto:Law-election at department-lists.uci.edu>
https://urldefense.proofpoint.com/v2/url?u=https-3A__department-2Dlists.uci.edu_mailman_listinfo_law-2Delection&d=DwICAg&c=U0G0XJAMhEk_X0GAGzCL7Q&r=2fMgMunsCtJpikIZXRvVAXXnXpXnW1DdeOa9_DBJVAg&m=K0zyd2Z4j71d02w628AD_m72Qvxaezv6R7sioWtTbJU&s=NFR636DZl3mX-wmXSIomVn5zvkhRm7j9j4MRMkzhuTc&e=

--
David Lublin
Professor of Government
School of Public Affairs
American University
4400 Massachusetts Ave.
Washington, D.C. 20016
http://davidlublin.com/<https://urldefense.proofpoint.com/v2/url?u=http-3A__davidlublin.com_&d=DwMFaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=ht72lsH5pHjOkY81Z8AqSl8AAHPg0bVpDLU6vILoyE0&s=0PUR9E5Aquv8SFyCV1XBe3108FmvXypnsRR_GIuPw4Y&e=>
<image001.jpg>
_______________________________________________
Law-election mailing list
Law-election at department-lists.uci.edu<mailto:Law-election at department-lists.uci.edu>
https://department-lists.uci.edu/mailman/listinfo/law-election<https://urldefense.proofpoint.com/v2/url?u=https-3A__department-2Dlists.uci.edu_mailman_listinfo_law-2Delection&d=DwMFaQ&c=HUrdOLg_tCr0UMeDjWLBOM9lLDRpsndbROGxEKQRFzk&r=xr_OjwGHtP-zw6I-DJj_MQ4cusLbiVT1bScGa0c8ZJo&m=ht72lsH5pHjOkY81Z8AqSl8AAHPg0bVpDLU6vILoyE0&s=b4Fjf3rhFt5rmbamg_uQXO7W7CPHmQp6kYK94z3bXHQ&e=>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20181203/e97fef0a/attachment.html>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image001.jpg
Type: image/jpeg
Size: 141790 bytes
Desc: image001.jpg
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20181203/e97fef0a/attachment.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image002.jpg
Type: image/jpeg
Size: 125403 bytes
Desc: image002.jpg
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20181203/e97fef0a/attachment-0001.jpg>
-------------- next part --------------
A non-text attachment was scrubbed...
Name: image003.jpg
Type: image/jpeg
Size: 158511 bytes
Desc: image003.jpg
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20181203/e97fef0a/attachment-0002.jpg>