Testamentary Transferrable Voting

[Brian Langstraat]

Below is a voting system that I have been thinking about recently that may work well.

For now, the voting system is named Testamentary Transferrable Voting (TTV).  Basically, election results are calculated using voter choice with transferred votes predetermined by the candidates.

Testamentary is defined as “bequeathed by a will or testament”.  Testament is defined as “something that serves as tangible proof or evidence, a statement of belief, or a written directive providing for the disposition of one's property after death”.

TTV is similar to Plurality Voting (PV), Single Transferrable Voting (STV), Instant-Runoff Voting (IRV), Ranked-Choice Voting (RCV), and Asset Voting (AV).

Below is order of events that occur in TTV:

1. Individuals apply (i.e. signed petitions) to be Potential Candidates (PCs) with the Election Organizer (EO) by a specific date [Election Day – 6 weeks].

2. The EO determines whether each applicant meets the legal requirements to be PCs and publishes a list of all PCs that are accessible to all of the applicants and the general public by a specific date [Election Day – 5 weeks].  There is a PC that will be designated as a “Write-in”.

3. Each PCs create a Testamentary Transferrable Ranking (TTR).  The TTR is a disposition of the PC’s received votes should they have the least votes (similar to AV, but prior to the election) or excess votes with multiple winners.  The TTR will list all other PCs in ranked choice order (similar to STV, IRV, and RCV, but determined by candidates).  PCs can negotiate with other PCs and the general public to create their TTR through legal political means such as promised policies and appointments.  However, the final TTRs will be private and sealed, so that a PC cannot modify their TRR to gain an unfair strategic advantage.

4. Each PC (including each hopeful write-in individual) gives the EO their sealed TTR by a specific date.  Each PC may give the EO a new sealed TTR that overrides the old TTR until the end of the due date [Election Day – 3 weeks].

5. The EO unseals all of the TTRs and determines whether each PC meets the legal requirements to be candidates on the election ballot.  Requirements include ranking all other PCs.  Each PC that fails to meet the requirements to be a candidate is removed from the unsealed TTR of the passing candidates.  Then, the EO publishes the final TTRs of all candidates that are accessible to all of the PCs and the general public by a specific date [Election Day – 2 weeks].  There will be one “Write-in“ candidate on the election ballot.

6. The EO distributes ballots with candidates listed, and each voter is allowed to privately vote for only one candidate on the ballot (similar to PV) by a specific date [Election Day].

7. The EO collects the ballots and counts the valid votes for each candidate.  If a candidate has more than the winning threshold of the total votes (winning threshold = total votes / (number of winners + 1) + 1), then they are a winner.  For multiple winners, each winner’s excess votes beyond the winning threshold are transferred per their TTR to the designated candidate (similar to STV).  For multiple and single winners, the candidate with the least votes is eliminated and their votes are transferred per their TTR to the designated candidate (similar to IRV).  Votes continue to be transferred from winning or eliminated candidates per their TTR ordering until the every winner has more than the winning threshold.  Any ties are broken by determining which tied candidate has the most high ranking votes, then by determining which candidate needs to be eliminated to result in the final winner with the highest number of votes.  Any transferred votes to the “Write-in” candidate are given to the legally eligible individual (who had submitted a sealed TTR) with the most votes.

8. The EO declares the winner(s) by a specific date [Election Day + 1 day].

Pros:

The act of voting using TTV is the same as PV, thus it does not require any change to most ballots and voting machines (in the United States).  However, there would need to be changes to ballot access and vote counting laws.  Determining the winner(s) is fast and deterministic (unlike STV, IRV, RCV, and AV).

Winners must be determined by the majority (single winner) or winning threshold (multiple winners), since every valid vote is used throughout determination of the election results.

Voters can choose their favorite candidate with minimal reasons to strategically vote.  A voter’s favorite candidate could change depending on the content of each candidate’s TTR.  Candidates will not be disappointed by the alternative votes of rogue supporters.

Creating TTRs encourages clear coalitions between independent candidates, small parties, and large parties while maintaining party unity even with candidate cloning.  Voters could vote for independent candidates and small parties without worrying about the spoiler effect.  Independent candidates and small parties have leverage to negotiate with large parties, especially in close elections.  After being ranked by prominent large party candidates, independent candidates and small parties would gain attention and potentially more support.  Candidates from large parties could highly rank other candidates from their party.  It is possible for a candidate’s sealed TTR to go against coalitions or their fellow party members, but this would likely be seen negatively by voters.

Cons:

The act of voting using other methods such as ranges or ranks may not work within TTV.

The results of TTV may confuse some voters or anger plurality (but not majority) winners

TTRs may confuse some voters.  TTRs would be heavily discussed in the news and debates.

Voters may feel like their vote is being traded by corrupt politicians.

Questions:

Does a voting system similar to TTV already exist?  If so, what is it called?

What are the positive or negative aspects of TTV?  Which voting system criteria does TTV pass or fail?

Is there a way to experimentally test TTV?

[Andy Jennings]

It sounds similar to SODA voting:
http://wiki.electorama.com/wiki/SODA_voting_(Simple_Optionally-Delegated_Approval)

and PAL representation:
http://wiki.electorama.com/wiki/PAL_representation

--
You received this message because you are subscribed to the Google Groups "The Center for Election Science" group.
To unsubscribe from this group and stop receiving emails from it, send an email to electionscience+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Brian Langstraat]

Andrew,

I appreciate that you pointed out voting systems that are similar to Testamentary Transferrable Voting (TTV) such as Simple Optionally-Delegated Approval (SODA) Voting and Proportional Accountable Local (PAL) Representation.

SODA is similar to TTV, but seems even more similar to Asset Voting.  I am a fan of Approval Voting, so its integration with SODA is appealing.  I thought that TTV using Approval Voting would cause some voter’s votes to be counted multiple times, but TTV could integrate Approval Voting by discounting transferred votes from voters who had already voted for the receiving candidate.  I like the list of preferences next to the candidate’s name on the ballot, but having the lists somewhere else on the ballot may make the actual voting section clearer.

TTV seems to be better than SODA in the following ways:

SODA has a more complex ballot for voters than TTV, especially the ambiguous preferences and confusing description needed for opting out of delegated votes.  The TTV ballots would be identical to ballots currently used for plurality voting.

TTV forces candidates to negotiate rankings with outcomes shown to the voters prior to election, while SODA gives the candidates the option of negotiating with votes after the election.  SODA negotiations will seem corrupt to voters when they have no certainty as to who their vote will eventually support.  SODA would result in even more negative campaigning with candidates fighting not only before the election but after it as well

SODA does not ensure that every valid vote contributes to the final election results, since pure approval votes for an unpopular candidate would be eliminated or a selected candidate may choose not to even delegate votes.  TTV does ensure that every valid vote contributes to the final election results, since candidates must rank every other candidate and cannot opt out of delegating votes.

Are there properties of SODA that seem better than TTV?

PAL is fairly similar to TTV for multiple districts.  Below is how TTV could use factions and multiple districts.

Factions:

TTRs may contain factions that are special groups of consecutively ranked PCs.  Factions are typically created around political parties or alliances.  When a candidate’s votes are delegated to a faction, each candidate within the faction gets an equal fraction of the votes at the same ranking with remainder votes delegated to the candidates with the highest ranking.  As candidates within a faction are eliminated or win, the original candidate’s transferred votes are re-delegated at the next ranking to the remaining faction candidates.

Multiple District Groups:

Only the residents of a district are allowed to be candidates for that district.  However, votes may be transferred by candidates between districts within a Multiple District Group (MDG) such as a state.  Voters may vote for any candidate in their MDG on the ballot in an election.  TTRs rank all of the candidates within each candidate’s MDG.  If a candidate has more votes than the winning threshold for the entire MDG, then they are a winner for their district.  When a candidate within a district is eliminated or has excess votes beyond the winning threshold for that district, the votes are transferred per the candidate’s TTR throughout the MDG.  Candidates with the least transfer votes are delegated with top priority.  The last candidate remaining in each district cannot be eliminated and is that district’s winner.  The last candidate will eventually be transferred votes to reach the winning threshold.

Are there properties of PAL that seem better than TTV?

To unsubscribe from this group and stop receiving emails from it, send an email to electionscien...@googlegroups.com.

[Brian Langstraat]

Jameson Quinn's post "GOLD voting and the 1967 law for single member districts" on 5/23/2017 points out a voting method that is very similar to Testamentary Transferrable Voting (TTV), especially for mulit-winner races.

Geographic Open List/Delegated (GOLD) voting is described at http://wiki.electorama.com/wiki/Geographic_Open_List/Delegated_(GOLD)_voting.

Similarily, Open List/Delegated (OLD) voting is described at http://wiki.electorama.com/wiki/Open_List/Delegated_(OL/D)_voting.

OLD voting is basically the same as GOLD voting, but without the constraint of one candidate per riding, and with a slightly weaker elimination rule.

Comparison of TTV and GOLD/OLD:

1.  "Choose a candidate" is similar to TTV.  I like listing the "first three transfer preferences" on the ballot.

2.  "Choose a transfer method" seems unnecessary and complicates the ballot.  "Trust the candidate" essentially becomes "Trust the voters of your chosen candidate’s party", since coordinated candidates within parties would use the TTV Factions described earlier.

3.  For "Tally votes", allowing fractions of a vote seems unnecessary and complicates the counting, since votes in TTV are hopefully transferred to acceptable candidates. If not, then the voter should choose another candidate.

4.  The TTV Multiple District Groups (MDG) is pretty similar to GOLD/OLD vote transferring except without the elimination thresholds.  TTV MDG eliminates from lowest votes to highest votes for the entire MDG, while GOLD eliminates all but the top 2 in each district and OLD eliminates anyone with less than 25% of the votes in a district.

TTV seems to be slightly better than GOLD/OLD, since the voting and counting are simpler and the elimination of candidates does not intentionally discourage minor parties.

[Jameson Quinn]

Thank you, Brian, for pointing out the similarity between GOLD/OLD and TTV. You are quite right, and I'd read your TTV proposal with interest, and probably it influenced my thinking as I was designing GOLD. I respond to your specific points below. 

2017-05-23 12:02 GMT-04:00 Brian Langstraat <langstra...@gmail.com>:

Jameson Quinn's post "GOLD voting and the 1967 law for single member districts" on 5/23/2017 points out a voting method that is very similar to Testamentary Transferrable Voting (TTV), especially for mulit-winner races.

Geographic Open List/Delegated (GOLD) voting is described at http://wiki.electorama.com/wiki/Geographic_Open_List/Delegated_(GOLD)_voting.

Similarily, Open List/Delegated (OLD) voting is described at http://wiki.electorama.com/wiki/Open_List/Delegated_(OL/D)_voting.

OLD voting is basically the same as GOLD voting, but without the constraint of one candidate per riding, and with a slightly weaker elimination rule.

Comparison of TTV and GOLD/OLD:

1.  "Choose a candidate" is similar to TTV.  I like listing the "first three transfer preferences" on the ballot.

I agree; transfer preferences should be public, and the first few should be listed on the ballot. In GOLD/OLD, transfer preference rankings need not be strictly ordered, so instead of "first three preferences", I'd say "enough of the top transfer preference levels to include at least three candidates", but that's a detail.

 

2.  "Choose a transfer method" seems unnecessary and complicates the ballot.  "Trust the candidate" essentially becomes "Trust the voters of your chosen candidate’s party", since coordinated candidates within parties would use the TTV Factions described earlier.

The point of the "Trust your party's voters" in GOLD is for voters who explicitly do NOT trust their candidate to be willing and able to use their own judgement and resist pressure to rank entrenched party insiders / powerbrokers higher than they deserve. Note that this is the default option for the most common kinds of vote, so that unengaged voters will not inadvertently give their voting power to such party insiders, while engaged voters will still have the power to give their voting power to whomever they want. 

 

3.  For "Tally votes", allowing fractions of a vote seems unnecessary and complicates the counting, since votes in TTV are hopefully transferred to acceptable candidates. If not, then the voter should choose another candidate.

I agree; I've removed that option from the GOLD page.

 

4.  The TTV Multiple District Groups (MDG) is pretty similar to GOLD/OLD vote transferring except without the elimination thresholds.  TTV MDG eliminates from lowest votes to highest votes for the entire MDG, while GOLD eliminates all but the top 2 in each district and OLD eliminates anyone with less than 25% of the votes in a district.

TTV seems to be slightly better than GOLD/OLD, since the voting and counting are simpler and the elimination of candidates does not intentionally discourage minor parties.

TTV is better for minor parties than OLD, which is better than GOLD (which is in turn still far better than FPTP). For major parties, GOLD is better than OLD which is better than TTV. In a national legislature with hundreds of seats, I believe that GOLD would lead to 3-5 parties and a handful of independents; OLD would lead to 4-6 plus that same handful; and TTV might well lead to 7 or more parties. In the mydemocracy.ca poll, 58% of Canadians said they preferred a few big parties over many small parties.¹ I think that 3-5 parties is a happy medium between FPTP's 2-party domination and extreme fragmentation. Unlike FPTP, it would allow enough dynamism for new parties to grow and old ones to subside naturally; unlike TTV, it would tend to avoid tiny parties wielding disproportionate power by holding the few votes necessary to make up a majority.

¹ When asked to choose between “Having many small parties in Parliament representing many different views or having a few big parties that try to appeal to a broad range of people”, 58% of Canadians said they preferred a few big parties

 

To unsubscribe from this group and stop receiving emails from it, send an email to electionscience+unsubscribe@googlegroups.com.

[Brian Langstraat]

"Testamentary Transferable Voting" was my first posted Topic on this Group forum (and any forum on the internet).  Now, it makes sense that I did not find GOLD/OLD in Wikipedia when writing my earlier Posts.  I am glad that TTV has inspired investigation of other "candidate delegated" voting methods.  Your description of GOLD/OLD is much clearer than my description of TTV.

Perhaps, "Single Delegated Full-List Transferable Voting" would be a more precise name than "Testamentary Transferable Voting".

In GOLD/OLD, transfer preference rankings need not be strictly ordered, so instead of "first three preferences", I'd say "enough of the top transfer preference levels to include at least three candidates", but that's a detail.

I think that "first three preferences" in order is better, so voters would know (or be reminded) how their vote would be transferred.

In TTV, all candidates (including "Write-in") are consecutively ranked (even in Factions), so the "first three preferences" would be specific candidates instead of listing parties.

The point of the "Trust your party's voters" in GOLD is for voters who explicitly do NOT trust their candidate to be willing and able to use their own judgement and resist pressure to rank entrenched party insiders / powerbrokers higher than they deserve.

In TTV, the Testamentary Transferable Ranking (TTR) would be the most important piece of information that most voters would have about each candidate.

The TTR would clearly display each candidates political allegiances, so voters would have to consider how much pressure applied by party insiders affects to each candidate.

For example, an anti-establishment candidate that highly ranks establishment candidates would lose some credibility.

I think "Trust your party's voters" would further entrench party insiders and give voters an excuse for not carefully examining their favorite candidate's allegiances.

I agree; I've removed that option from the GOLD page.

Fractions of a vote is a neat idea, but kind of defeats the point of "candidate delegated" voting methods with simple "FPTP" ballots and counting.

For major parties, GOLD is better than OLD which is better than TTV.

I think that TTV is philosophically better for proportional representation, but GOLD is realistically better for likelihood of legislatures accepting it and avoiding legislative gridlock.