[EL] Single-Transferrable Vote

Charles H Stewart cstewart at mit.edu
Fri Jan 25 16:30:00 PST 2019


Amused at the people not from Cambridge opinining about the people of Cambridge.



Sent from my Verizon, Samsung Galaxy smartphone


-------- Original message --------
From: Jack Santucci <jack.santucci at gmail.com>
Date: 1/25/19 6:12 PM (GMT-05:00)
To: Douglas Johnson <djohnson at ndcresearch.com>
Cc: law-election at uci.edu
Subject: Re: [EL] Single-Transferrable Vote

Doug (and interested parties),

Cambridge's random-transfer approach was grandfathered in (the Plan E charter provision of the MA General Laws). It was known as the "Cincinnati method" and used in the city of that name for 16 elections (1925-55). I think there was also a Boulder method, but I can't remember the details. There are a lot of methods. This Boulder LWV presentation by Celeste Landry explains why random transfer isn't ideal but not awful either (see slide titled "Nuts and Bolts of Cambridge STV"): www.lwvbc.org/docs.ashx?id=357521<http://www.lwvbc.org/docs.ashx?id=357521>

My original point was about strategic coordination. I should have been clearer. "Strategic coordination" means voters doing what it takes to win. This usually involves some party or party-like entity, as voters don't coordinate all on their own.

In an RCV world -- single- or multi-winner -- "party-like entity" and "strategic coordination" can mean a preference-swapping deal along the lines of what we saw in Maine's Democratic gubernatorial primary or the 2018 San Francisco mayoral race. When I see an election with a lot of exhausted ballots (e.g., Oakland 2010), it suggests to me that elites failed to coordinate.

All electoral systems involve coordination. Consider single-seat plurality districts. Now consider such an election without parties or party-like entities. How would the "Duvergerian" top two emerge? Someone or something has to tell voters how to vote. STV works the same way. As does vote-for-a-party PR, which is why I offered the Belgium example with its D'Hondt allocation.

Have a great evening,
Jack

On Fri, Jan 25, 2019 at 5:41 PM Douglas Johnson <djohnson at ndcresearch.com<mailto:djohnson at ndcresearch.com>> wrote:
>
> I am missing where the congressional apportionment math used in the back rooms of the Census Bureau relates to the question of how complicated a system the voters are open to using to decide how to mark their ballots. Of course we don’t ask anyone to vote using the Method of Equal Proportions. The only folks that need to learn it are the Census statisticians and Clark and Kim to predict and to check the work of those Census statisticians.
>
>
>
> Separately, thanks for the clarification on Cambridge. I did not realize that. Is there awareness in the city that the 2nd round winners are essentially chosen by random drawing?
>
>
>
> Bigger-picture, I think the Cambridge situation proves my point: if the fractional-transfer approach – which is the only way to avoid a potential random drawing election result – is too complicated for MIT voters, how does one support an argument it would be practical nationwide? Or are single transferrable vote advocates actually suggesting letting random draws heavily influence the makeup of Congress? (I know all the lawyers on this list are salivating at the business that would generate, but . . . )
>
>
>
> Regarding the Belgium modified-straight-party-voting system, I am not aware of any jurisdiction in the US, nor any opinion poll, that has found American voters have any interest in voting for a party list rather than for individuals, even in the Belgium modified system. (And of course party leaders like a party-ticket system where party leaders decide who gets on the ticket in what order).
>
>
>
> Actually, let me correct myself: Americans did change our election system nationally to straight-party-ticket voting for one office: the electoral college. Anyone think that’s a good example to cite when trying to enact an election system change?
>
>
>
> More seriously, I think the idea of replacing candidate names with party label voting is a complete non-starter in the US. It may be fun for academic papers to ponder, but I agree with Professor Mulroy (whose pro-STV article started this conversation) that: “Most PR systems have voters vote for parties rather than candidates, eschew primary elections, or foster the instability of parliamentary systems, making them unsuitable for America.”
>
>
>
> Doug
>
>
>
> Douglas Johnson, Ph.D.
>
> Fellow, Rose Institute of State and Local Government
>
> at Claremont McKenna College
>
> douglas.johnson at cmc.edu<mailto:douglas.johnson at cmc.edu>
>
> direct: 310-200-2058
>
>
>
>
>
>
>
> From: Law-election <law-election-bounces at department-lists.uci.edu<mailto:law-election-bounces at department-lists.uci.edu>> On Behalf Of Edelman, Paul
> Sent: Friday, January 25, 2019 11:35 AM
> To: John Tanner <john.k.tanner at gmail.com<mailto:john.k.tanner at gmail.com>>; law-election at uci.edu<mailto:law-election at uci.edu>
> Subject: Re: [EL] Single-Transferrable Vote
>
>
>
> The US has over 80 years of using the Hill method for apportionment.  Any guesses on  how many people understand it?
>
>
>
> Paul H. Edelman
>
> Professor of Mathematics and Law
>
> Vanderbilt University
>
> paul.edelman at vanderbilt.edu<mailto:paul.edelman at vanderbilt.edu>
>
> 615-322-0990
>
>
>
> From: Law-election <law-election-bounces at department-lists.uci.edu<mailto:law-election-bounces at department-lists.uci.edu>> On Behalf Of John Tanner
> Sent: Friday, January 25, 2019 1:30 PM
> To: Jack Santucci <jms346 at georgetown.edu<mailto:jms346 at georgetown.edu>>
> Cc: Election Law Listserv <law-election at uci.edu<mailto:law-election at uci.edu>>
> Subject: Re: [EL] Single-Transferrable Vote
>
>
>
> Well, the Belgians have had nearly 120 years to get used to the H’ondt system, so most of them probably understand its operation
>
>
>
> On Jan 25, 2019, at 2:16 PM, Jack Santucci <jms346 at georgetown.edu<mailto:jms346 at georgetown.edu>> wrote:
>
>
>
> Doug and Vlad et al,
>
>
>
> It sounds like Doug is describing a fractional transfer method, which is not what Cambridge (MA) uses. From a candidate with surplus, ballots are randomly chosen for transfer. Others on the list know these details better than I.
>
>
>
> To Vlad's point, how many Belgians can explain the mechanics of a D'Hondt seat allocation under that country's closed-list PR system? I suspect most cannot. Rather, a sufficient set of parties is happy enough with the system to avoid referring to "lottery effects" and other confusing mechanics. Parties simply tell their voters to vote for the party -- not unlike what we saw here, when many cities had STV.
>
>
>
> Jack
>
>
> On Jan 25, 2019, at 13:42, Kogan, Vladimir <kogan.18 at osu.edu<mailto:kogan.18 at osu.edu>> wrote:
>
> Doug raises a very important point. I think the educational challenge in terms of explaining the system to voters is only a piece of it. Imagine also the voter education challenge in actually learning enough information about a sufficient number of candidates to be able to rank them in a way that makes your vote effective. I think that challenge is impractically difficult for a significant minority of voters, especially for lower-salience down-ballot races, and this explains why we see such high rates of ballot exhaustion when IRV/RCV is used in local elections.
>
>
>
> An important, and I think woefully understudied, question is whether this affects certain voters more than others. The answer has obvious political and legal implications.
>
>
>
> From: Law-election <law-election-bounces at department-lists.uci.edu<mailto:law-election-bounces at department-lists.uci.edu>> On Behalf Of Douglas Johnson
> Sent: Friday, January 25, 2019 1:34 PM
> To: 'Rick Hasen' <rhasen at law.uci.edu<mailto:rhasen at law.uci.edu>>; 'Election Law Listserv' <law-election at uci.edu<mailto:law-election at uci.edu>>
> Subject: Re: [EL] Single-Transferrable Vote
>
>
>
> While single-transferable vote sounds nice in the simple description below, I encourage list readers to consider how it works in practice, while keeping in mind how hard states are working to remove obstacles to voter participation and causes of voter confusion with our existing very simple elections. Now swap that current simple system for this:
>
>
>
> If there are 5 seats open and 1,000 votes cast, then it takes 201 votes (1/5 + 1) to meet the “quota” that guarantees a win.
>
> Now imagine a candidate receives 287 first-place votes. The candidate only needs 201 (70% of 287) votes to win, so 86 votes, or 30%, of the candidate’s votes will be allocated to the voters’ 2nd choice. But not all of the 287 voters have the same 2nd choice. So for each of those 287 voters’ ballots, the 2nd choice receives 0.3 votes (30% of one vote). So a candidate who was listed 2nd on 95 of those 287 ballots would gain 28.5 votes toward the 201 vote target in the second round of vote-counting (and yes, we would now be announcing fractional votes).
>
>
>
> I suppose it is no surprise that such a complicated system is used in the city that is the home of MIT, but can you imagine the voter-education challenge explaining all of this anywhere else?
>
>
>
> -          Doug
>
>
>
> Douglas Johnson, Ph.D.
>
> Fellow, Rose Institute of State and Local Government
>
> at Claremont McKenna College
>
> douglas.johnson at cmc.edu<mailto:douglas.johnson at cmc.edu>
>
> direct: 310-200-2058
>
>
>
> _______________________________________________
> 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
>
> _______________________________________________
> 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
>
>
>
> _______________________________________________
> 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
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://webshare.law.ucla.edu/Listservs/law-election/attachments/20190126/2adcc6b3/attachment.html>


View list directory