Brainstorming thread - Multi-winner elections and how to handle NOTA

[Paul Bracco]

Colleagues,

There was a discussion in our original brainstorming thread about revamping at-large voting (possibly JC voting too?) away from approval voting.

I want to create a thread specifically to flesh this idea out. Specific things that I think need to be covered:

• Cumulative voting vs STV
• How to handle NOTA

I'd also welcome any ideas on how to structure things from the other members of the committee who are more knowledgeable about alternative voting methods.

I'd also ask that an informal discussion item for this topic be added to the rolling agenda.

For reference, a previous committee's (2010) STV proposal is at the very bottom of this document.

Sincerely,

Paul Bracco

Vice Chair - Libertarian Party of Virginia

[Secretary LNC]

I am going to push back on STV for a second (though I am a supporter of RCV etc etc)

This is way too complex unless we have a robust electronic system, but more importantly, it cannot be adequately hand recounted (so much would depend upon the skill and experience of the Tellers and Secretary and we are not all mathmeticians) if someone challenges.  This will lead to great suspicion.

Cumulative CAN easily be audited.  

With officers I have been thinking of a potential compromise with those who say you cannot have the information to strategically vote on subsequent rounds.... what about first round as normal.  If it has to go to a second round, THAT will be an IRV ballot.  So no matter what it is a maximum of two rounds for officers.

I like STV, but I can tell you that I hope I have been considered a decent LNC Secretary - my god, I can't imagine having to oversee that if it were not electronic.  

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

--
Committee members, download the Proposal Form here: https://tinyurl.com/2024Bylaws-SubmissionForm
---
You received this message because you are subscribed to the Google Groups "Bylaws Committee 2024" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bylaws-committee...@lp.org.
To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/d450095b-3537-4a78-af82-2e5ed142ed08n%40lp.org.

[Mike Seebeck]

I will be addressing NOTA in an upcoming proposal that I haven't written yet. Short form is we have a huge hole on that.

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAGiA9Wk_y%3D-JXVbkPAMpB97uqC1Quwhh%3DZfVMhFkxX3hGkEpbg%40mail.gmail.com.

[Secretary LNC]

We certainly do and there are very strong feelings on two opposing views LOL.  I just want it resolved, though I certainly have a preference.

One thing I want everyone to keep in mind - and do with this as you may - I think most of us want electronic balloting at some point.  We have to make our rules and bylaws ready for that transition.  I am not aware of any program that would satisfy both of the poles on this issues:

1.  One side wants NOTA not to be eliminated naturally.

2.  One side wants NOTA to mean no one, not some people plus NOTA.

There is probably another side, enough for a D&D die.

I am not sure there is any software that can do what we want.  I am in fear that we may have to choose to between NOTA and electronic balloting (in multi winner races), and with Libertarians, NOTA will win that fight every time. 

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAGasu4TpDR7SVGsee%3D1p2JVJa8QcrakMjvT1L59Z_pHm2-tVpA%40mail.gmail.com.

[Mike Seebeck]

My proposal will not assume electronic balloting as we don't have that yet. Programmable ballots can manually override any NOTA eliminations if the software is worth anything. 

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAGiA9WmcSjLhJUpSDH6d_WJABom7tVERP_LyiTk8HpgG247OOw%40mail.gmail.com.

[Rob Latham]

Riffing off of Mr. Seebeck's comment, I'd like to see a ranked choice voting solution that does not assume electronic balloting.

I understand that Cambridge, Massachusetts hand-counted STV / PRCV elections for decades before involving computers.

I like hand-counting as a transparency measure. 

On occasion, that's taken a few days for Cambridge, MA vote-counters accomplish, but my sense is that our tellers can process a STV / PRCV election involving about 1064 ballots sooner than that. I'd love to see the process videotaped as well, so that it can be analyzed, and hopefully emulated.

Wonder if someone could make a records request to the City of Cambridge ... or get the process out of a Cambridge, MA library ... for study and possible adaptation?

That said, I'd like the paper ballots for delegates to use to be machine-readable/scannable. That way, preliminary results could be shown quickly, but the real result will be once the audit/canvass is completed. (And if those ballot images can be scanned, anybody can audit the election.)

See, e.g., https://electionbuddy.com/features/voting-systems/stv-voting/

As to voting receipts, see also

https://www.qrcode-tiger.com/qr-code-for-voting-system


Here's how a hypothetical delegate voting experience might work:

1. Delegate gets a link (ElectionBuddy, OpaVote, etc.) by email.

2. Delegate ranks candidates with a smartphone, laptop, or a voting station set up in the convention hall.

3. The delegate's cast vote generates a ballot indicating that delegate's selections for the delegate. The printed ballot can be anonymous as to the delegate (unless that delegate wants to share publicly how that delegate voted), but the delegate can check the accuracy of the printed ballot as compared to the vote cast through their smartphone, etc. through a QR code. Anyone reading that ballot can see how the candidates were ranked.

4. The delegate (or someone on the delegate's behalf) takes the generated ballot to the Secretary's desk/Tellers.

The software can conduct the initial tabulation, but Tellers would hand count the submitted ballots and report the results from the hand count.

One would hope that the results from the software and the hand count match.

[Secretary LNC]

You are missing something really important here.  They did not transfer fractional votes by hand.  I have been through this with RCV Colorado, read that first link very carefully.

"An important finding of this report is that Cambridge voters should be confident that the current "random draw" rules of transferring surplus ballots from elected candidates are unlikely to affect the results."

NOT fractional transfers.  I have been convention secretary, I am telling you that a system where we are expected to hand do fractional voting is not going to work.  There is idealism, and there is reality.  All sheerly electronic?  Sure.  But Libertarians will not trust that and there will be no realistic way for them to challenge because once the convention adjourns, it is too late to challenge.

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/2b447b80-0720-486c-becb-d01c9b512bffn%40lp.org.

[Secretary LNC]

Let me give you a real life example from my own work with RCV Colorado - this is a conversation from me and a representative talking about this very subject in writing legislation for Colorado to consider - notice the part I highlighted:

I made some minor updates. There is one additional step in multi-winner - where the surplus votes get transferred to those voter's next choices. It spares the voters from having to second-guess. If there were not a surplus transfer, they'd have to not vote their favorite if they thought that their vote would go to waste on someone who was going to win anyway.

How do those transfer votes get picked out? We leave that to the clerks from among a handful of fair methods. In a software tally election it is fractional for all who voted for a winner. (We never use the word fractional with voters...because many voters do not understand the word "fraction".) In a hand-count tally - I have had the "losing" side randomly pick the ballots to get transferred. That way they can't complain that the random ballots were done wrong. 

As far as the voters are concerned - they don't care. They care about how to use the ballot and possibly that the proportional results are not winner-take-all.

I still have to update the color graphics on the proportional crowd where everyone is green or yellow. I am thinking that I'll change green to grey...because no one thinks they are grey. I also need to create some graphics around the threshold calculation to drive home that the more seats, the more opportunity there is. 

Please let me know when there is an even where it would make sense for me to present this material.

Thanks much!

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

[Secretary LNC]

With that in mind, if we want STV, I think we are going to need to consider something like that - the "losers" randomly pick the ballots to transfer if it is done by hand.

Adding things that are going to make conventions more complicated (we already don't finish business) is IMHO not a good plan.  If I have to choose between streamlining to actually finish our business and trying out voting methods I would rather see in the public sphere - I am going to pick streamlining.  I have been on that other side of the desk.  As has Mr. Seebeck.  It gets ugly.

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

[Secretary LNC]

(correction the discussion was about an explanatory graphic, but the concepts are all from extensive work in Colorado that RCV Colorado is doing to prep the laws on the books)

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

[Sylvia Arrowwood]

Is "eliminated naturally" the same as "totally eliminated"? 

Not sure on that term.

Think NOTA is a great option. But IMO  "none" means "not one of".  So

would lean towards No. 2 group.

---

From: "Mike Seebeck" <mike.s...@gmail.com>
To: "bylaws-committee-2024" <bylaws-com...@lp.org>
Sent: Friday, February 10, 2023 1:46:37 PM
Subject: Re: BYLAWS-COMMITTEE Brainstorming thread - Multi-winner elections and how to handle NOTA

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAGasu4SuV%3DkQ4hh1qBxxtGi2xR6KycNr1fdZYkt_-nYa3RKqoA%40mail.gmail.com.

[Rob Latham]

I'm in favor of streamlining as well.

Aware of a voting method or rule -- the "Andrae method" -- that uses the transfer of whole votes rather than fractional votes. (The process is so short, I'll copy it in the postscript below.)

I'll leave it to the mathematicians to explain whether one method is markedly superior to another (e.g., Meek STV vs. WIGM STV), but my sense in comparing similar methods is that the gradation among them is pretty fine.

See also some defined terms/terms of art at: https://prfound.org/resources/reference/

In liberty,

Rob Latham

P.S.: 

Reference Andrae Rule

The Andrae method deals with whole votes, with random surplus transfers. It’s easy to understand and count, but both WIGM and Meek’s method do a fairer job of transferring surpluses.

Count multiple-seat elections as follows.

1. Initialize Election
   Set the quota to the total number of valid ballots cast for that office divided by one more than the number of seats to be filled, ignoring any remainder, plus one. Put the ballots in random order and sequentially number them to permit recounts. Distribute the ballots (D.1). If count is complete (D.2), continue at step C.
2. Rounds
   1. Defeat the hopeful candidate with the lowest vote tally. Break ties by choosing the tied candidate who is earliest in a predetermined random tiebreaking order.
   2. If count is complete (D.2), continue at step C.
   3. Distribute the defeated candidate’s ballots (D.1).
   4. Continue at B.1.
3. Finish Count
   If all seats are filled, defeat all remaining hopeful candidates; otherwise elect all hopeful candidates. Count is complete.
4. General Procedures
   1. Distribute Ballots. Drawing the ballots from the top of the source pile, place each ballot in turn on the top of a pile corresponding to the highest-ranked hopeful candidate on the ballot, keeping a tally of the votes in each candidate pile. When a candidate’s tally is equal to the quota, elect that candidate, and transfer no more ballots to that candidate. Place ballots that rank no hopeful candidate in a separate exhausted-ballot pile.
   2. Test count complete. If the number of elected candidates is equal to the number of seats to be filled, or the number of elected plus hopeful candidates is equal to or less than the number of seats to be filled, the count is complete.

Notes

This is a very simple example of a whole-ballot random-transfer STV rule. It gives fairly good results for relatively large elections. More complex variations on this theme are available (see Cambridge MA for an example) that avoid some of the drawbacks of this general method. However, if this rule as specified is not sufficient to the purpose (as when a rule is required for hand-counting small elections), we recommend using the Reference WIGM rule rather than a more complicated Andrae rule. The simple WIGM rule will be nearly as easy to count as a complex Andrae rule, and will give better results in almost all cases.

Putting the ballots in random order is not trivial, but it’s important to this method. A possible approach: place each ballot on one of six piles, depending on the throw of a die. Reassemble the piles and repeat, the more times the better. Note that this process can be carried out by multiple die-thowers working in parallel.

In step A, the ballots should be numbered as they’re first distributed, so that they can be re-ordered for a recount. Number the ballots in a distinctive place and/or with a distinctive color to avoid the possibility of confusion with the ballot’s rankings.

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAGiA9WkbTSu%2BWxOWssh%2BL%3DQcRXyxP5ADrBqkY152S8fp3BAT5w%40mail.gmail.com.

[Secretary LNC]

Well that can be done by hand.  It doesn't streamline anything however.  Time is our enemy.  Conventions already don't have enough time.  I don't know how to balance this interests - honestly - I am an RCV or die type of person but I also know the Cold Equations of our schedule and capabilities.

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAOKuKG%3Dd4PYBiWLwwWYF1vhOLhGNQ76UYzWbCq7VKH_ohRUyxw%40mail.gmail.com.

[Rob Latham]

Well, look what I found.

"Rules for counting ballots, and determining results."

The only occurrence of the word "fraction" was in a sentence that discusses "disregarding fractions." :-D

At section 9, pages 9 to 11.

https://www.cambridgema.gov/election/programsandservices/~/media/7161093D8E10447B8A13DB3F05BD395E.ashx

On Fri, Feb 10, 2023 at 3:01 PM Rob Latham <freeu...@gmail.com> wrote:

[Secretary LNC]

==(t) No canvass or count ofthe vote shall be made on the Lord's day.==

Well that rules out one day of the convention ;)

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

--
Committee members, download the Proposal Form here: https://tinyurl.com/2024Bylaws-SubmissionForm
---
You received this message because you are subscribed to the Google Groups "Bylaws Committee 2024" group.
To unsubscribe from this group and stop receiving emails from it, send an email to bylaws-committee...@lp.org.

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAOKuKGnma1UbdR_q3QMrxCiYSyF%3DLyyZ9kn63h%3DD92zT3OpVKA%40mail.gmail.com.

[Rob Latham]

Lol ... spotted that, too.

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAGiA9Wkpbis1jpwVn6ot0DnUzcjCE7O%3DbDHiXk3Dht1gBdbSXw%40mail.gmail.com.

[Secretary LNC]

But in short, an STV method that doesn't use fractional votes is doable in theory.  But the reality we MUST grapple with is we already don't have enough time.  Having a highly trained teller team (one problem with teller teams is it is volunteers that don't always show up, change their mind, etc - we are not paying people to do this) can likely do it in the same amount of time.

Much more doable is cumulative voting.  

With all that, I think I am repeating myself so will try to shut up, a skill I am not terribly good at.

___________________________________________________

In Liberty, Caryn Ann Harlos

LNC Secretary and LP Historical Preservation Committee Chair ~ 561.523.2250

[Rob Latham]

My sense is that adjusting the existing order of business would accommodate a hand count of RCV / PRCV elections to save time that has otherwise been spent on elections during the convention ... the count would happen off-stage, both during and outside of the business session. (Perhaps this requires more tellers, and late nights, and I suggest a dry run with mock RCV / PRCV elections at the 2024 LPNatCon.) Results could be announced when deemed appropriate.

There was an informal mock RCV election -- for the Chair's race, I believe -- at the 2010 St. Louis LPNatCon. If memory serves correctly, it seemed to predict the eventual outcome fairly well.

Here's the existing order of business:

1. Call to order 

2. Credentials Committee report 

3. Adoption of agenda 

4. Treasurer's report 

5. Audit Committee report 

6. Bylaws and Rules Committee report 

7. Election of Judicial Committee (in appropriate years) 

8. Election of Party officers and at-large members of the National Committee 

9. Platform Committee report 

10. Nomination of Party candidates for President and Vice-President (in appropriate years) 

11 Resolutions 

12. Other business  

Recall that the order of business at the 2022 LPNatCon was changed from the floor ... moving officer, at-large, and judicial committee elections ahead of the BandRComm and PlatComm reports. An adjustment like that could allow enough time to administer those elections using ranked methods.

So, here's a revised order of business that could accommodate hand-counted ranked voting methods.

Friday
1. Call to order 

2. Credentials Committee report 

3. Adoption of agenda 

4. Treasurer's report 

5. Audit Committee report 
6. Election of Party officers (using RCV)
7. Bylaws and Rules Committee report 
8. Election of at-large members of the National Committee (once results for Party officer elections are announced, speeches and balloting could break into the BandRComm report ... the BandRComm report could then resume once at-large speeches and balloting had concluded)

*Counting at-large election would take place in the interim

Saturday

1. Election of Judicial Committee (in appropriate years)
2. Platform Committee report 
3. Nomination of Party candidates for President and Vice-President (in appropriate years) 

4. Resolutions 

*Counting judicial committee election would take place in the interim (in appropriate years)

Sunday

1. Unfinished business

Anticipate that the BandRComm report may continue into Saturday. One could see other adjustments as well. For example, the national ticket election debate could be that Friday evening, in appropriate years. Party officer debates could be that Thursday evening during national ticket nomination years. The voting for the national ticket would start on Saturday, but may continue into Sunday.

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAGiA9WnEXzTfi%2B%3DJfmirH%3Daxzp4CsLDiDEdtrruznVr22vPhRA%40mail.gmail.com.

[Rob Latham]

Following up on Mr. Moellman's comment about NOTA as a concept and NOTA as a candidate, I am comfortable with both options ("sticky NOTA" vs. NOTA as a candidate). However, my sense is that "sticky NOTA" is going to require additional programming if electronic balloting is used and NOTA is treated as anything other than a candidate.

We know our friends at OpaVote don't like the use of NOTA in a ranked choice election.

So, for ranked elections, perhaps NOTA as a candidate ... that can be defeated ... is preferable. And the NOTA-as-a-candidate approach appears to satisfy Article 10.7.

That said, we appear to be leaving balloting for nominees for Pres. and Vice-Pres. intact, so NOTA would survive each round in that election.

If we're hand-counting RCV elections -- officers -- sticky NOTA would be more easy to administer (as compared to a proportional RCV election) ... it would act more like "None of the Remaining" and perhaps Article 10.7 should be amended to accommodate that.

On Thu, Feb 9, 2023 at 9:25 PM 'Paul Bracco' via Bylaws Committee 2024 <bylaws-com...@lp.org> wrote:

--

[Rob Latham]

I went back and reviewed cumulative voting.

Bottom-line, it's a semi-proportional voting method because its use may incidentally result in a proportional outcome.

It's not a preferential voting method either.

Its susceptibility to strategic voting also increases the possibility that a well-organized voting bloc could win all seats, rather than a proportionate number of seats.

That said, it has been court-ordered as a Voting Rights Act remedy in certain jurisdictions. But it's efficacy in achieving proportional outcomes has been mixed. 

Still, seems like we can do better with a preferential and proportional voting method, and the single-transferable vote appears to provide such a method.

I hope to draft a STV / PRCV proposal based on our 2010 proposal, but use whole ballots rather than fractions.

On Thu, Feb 9, 2023 at 9:25 PM 'Paul Bracco' via Bylaws Committee 2024 <bylaws-com...@lp.org> wrote:

--

[Ken Moellman]

Election methods:  

I support RCV for single-winner elections other than nomination for POTUS, and that's only because we get a lot of media attention on the POTUS race proceedings.  If that weren't a factor, I would support RCV for POTUS too. Within the context of media coverage, I think a good balance is to have POTUS be conducted as multi-round elections, and VPOTUS be conducted as RCV; that way people at home watching can see just how much faster RCV actually is. And using RCV for all elections where there is only one winner should speed up business, especially if we allow other business (platform amendment) to continue while a tally is taking place.  Instead of waiting for hours, we could be working on other things while the tally is being completed. 

I generally disfavor STV. STV transfers feel icky. It's like forced-sharing or something. But beyond that, calculating it is complicated without the assistance of a computer. 

I really don't like Approval because I can't assign weight (preference) other than a binary Yes/No.

I really started to like Cumulative Voting, and especially so for races where the number of choices is approaching the number of candidates/options. That's one area where Cumulative definitely shines, and especially if there are criteria that limit the number of votes that could be placed on any particular candidate. The criticism that allowing a minority faction to pile all of their votes onto one candidate is valid, but also not an actual problem. Is there a filter for "problematic" members of the LNC?  I know that on the last term of the LNC, at least a couple of people were considered problematic. But that's what the voters want. And it's easy to plan resources for that tally.

NOTA: I prefer NOTA as a candidate, rather than a concept, unless NOTA winning lets everything start over. The bylaw that says if NOTA wins POTUS then no one is our candidate is awful.

A completely different approach:  eliminate multi-winner elections. In the recent-past there have been calls to have the Judicial Committee elected like the Regional Reps instead of at-large. I generally favor this idea, as it moves that business off of the convention floor. Another option floated was to increase the number of regions (to 16, which is effectively a maximum of 15), and eliminate At-Large members entirely.  If these two were combined, there would no longer be any field races at all and we could just use RCV for all single-winner elections other than POTUS where we use multi-round elimination.

Trust: Overall, any system of voting we adopt needs to be able to be reasonably hand-tallied, if the convention body demands it.  It's a critical check-and-balance that helps build trust, which in turn helps people who "lose" accept that the election was fair and that their candidate just lost.

---

Ken C. Moellman, Jr.

Executive Director

Libertarian Party of Kentucky

To view this discussion on the web visit https://groups.google.com/a/lp.org/d/msgid/bylaws-committee-2024/CAOKuKGn_h-f4AaZU5DZy2dc9qNU1A%2BXGnitrz3QOPrnX9s8f1w%40mail.gmail.com.

[Rob Latham]

On Saturday, February 11, 2023 at 11:00:00 AM UTC-7 ken.moellman wrote:

Election methods:  

I support RCV for single-winner elections other than nomination for POTUS, and that's only because we get a lot of media attention on the POTUS race proceedings.  If that weren't a factor, I would support RCV for POTUS too. Within the context of media coverage, I think a good balance is to have POTUS be conducted as multi-round elections, and VPOTUS be conducted as RCV; that way people at home watching can see just how much faster RCV actually is. And using RCV for all elections where there is only one winner should speed up business, especially if we allow other business (platform amendment) to continue while a tally is taking place.  Instead of waiting for hours, we could be working on other things while the tally is being completed. 

Am favorable to this idea.
 

I generally disfavor STV. STV transfers feel icky. It's like forced-sharing or something. But beyond that, calculating it is complicated without the assistance of a computer. 

Then don't vote for more than one candidate and there won't be transfers. Don't you feel cleaner already? ;-)

The Andrae method I mentioned in another email thread does not require the assistance of a computer, or the computation of fractions. The randomness of the transfers is a bug or a feature depending on one's perspective. Jury selection is supposed to be random.

Here's a link to the Andrae method of processing a proportional ranked choice voting election:

https://prfound.org/resources/reference/reference-andrae-rule/

I'd like to try it out for our multi-member bodies, at-large members and judicial committee.

Adopting proportional ranked choice voting also models the method for wider use in our municipalities and other legislative bodies, which would give Libertarians a better chance to be elected to those bodies in proportional relation to our support in our communities.
 

I really don't like Approval because I can't assign weight (preference) other than a binary Yes/No.

I really started to like Cumulative Voting, and especially so for races where the number of choices is approaching the number of candidates/options. That's one area where Cumulative definitely shines, and especially if there are criteria that limit the number of votes that could be placed on any particular candidate. The criticism that allowing a minority faction to pile all of their votes onto one candidate is valid, but also not an actual problem. Is there a filter for "problematic" members of the LNC?  I know that on the last term of the LNC, at least a couple of people were considered problematic. But that's what the voters want. And it's easy to plan resources for that tally.

NOTA: I prefer NOTA as a candidate, rather than a concept, unless NOTA winning lets everything start over. The bylaw that says if NOTA wins POTUS then no one is our candidate is awful.

Does not NOTA winning as a candidate let everything start over for Party Officer elections?

Noting the difference in how the bylaws treats NOTA for Pres/VP and Party Officers. I'm fine with the difference for now, but am keeping an open mind on treating them the same (although if nominations for Pres/VP gets re-opened, that could take time).

 

A completely different approach:  eliminate multi-winner elections. In the recent-past there have been calls to have the Judicial Committee elected like the Regional Reps instead of at-large. I generally favor this idea, as it moves that business off of the convention floor. Another option floated was to increase the number of regions (to 16, which is effectively a maximum of 15), and eliminate At-Large members entirely.  If these two were combined, there would no longer be any field races at all and we could just use RCV for all single-winner elections other than POTUS where we use multi-round elimination.

Am keeping an open mind on this, but leaning pretty strongly against it; these two bodies should be reflective of a cross-section of Party members (or at least national convention attendees). They are also the only vectors through which minority constituencies can get a voice.

 

Trust: Overall, any system of voting we adopt needs to be able to be reasonably hand-tallied, if the convention body demands it.  It's a critical check-and-balance that helps build trust, which in turn helps people who "lose" accept that the election was fair and that their candidate just lost.

To facilitate this, I'd like the counting process to be video-recorded and the ballots scanned. This would promote transparency, and the scanning of ballots also creates a body of data that election scholars and others could review.

Did the Andrae method yield a result that would have been different under the Meek method or the WGIM method (or other methods)? I'm not alone in being curious about that. Let's commit to a method, try it out, then adjust if needed.

Rob Latham

[Rob Latham]

Thinking through the logistics of how to randomize [second sentence of Step A of the Andrae method] 1000-1100 cast ballots printed on 8 1/2" x 11" pieces of paper ... a sufficiently-sized raffle drum would work. A standard laundry dryer (preferably no heat) -- and most convention hotels would have one, although we'd want to work out permission in advance or, failing that, bring our own -- would also work.