[EM] Linear summability

robert bristow-johnson rbj at audioimagination.com
Tue Apr 28 10:54:24 PDT 2020

> On April 28, 2020 9:26 AM Kristofer Munsterhjelm <km_elmet at t-online.de> wrote:
> On 27/04/2020 06.35, robert bristow-johnson wrote:
> > 
> > Now with this virtually Condorcet-compliant IRV-BTR method, that I am
> > actively lobbying both the Vermont state legislature and the
> > Burlington city council to adopt, would still be IRV so the
> > individual ballot records would still have to be securely
> > transported to the central tallying location for the kabuki dance we
> > call the "single transferable vote" to take place, but since it's
> > virtually Condorcet, the precincts can report the C×(C-1) subtotals
> > that can be summed and *if* there is a Condorcet Winner, we will know
> > who it is from the sums of the C×(C-1) subtotals.  (the IRV-BTR is,
> > in my opinion, *virtually* Condorcet-compliant because, since it is
> > STV, there can be no equal rankings of candidates on the ballot.)
> Yes, that's true.
> I wonder if it would be a good idea to point at this "speculative"
> precinct summability as a benefit of BTR-IRV compared to ordinary IRV,
> or if it would get in the way of the central point, which is that
> BTR-IRV avoids center squeeze.

Precinct summability is not the main point, and in my opinion, this "center squeeze" point *should* not be.

We do not want Condorcet because it somehow favors (w.r.t. IRV) the "center".  The Center is just another political interest and we do not want our election methods to explicitly favor any specific political interest over the others whether it's Left, Right, or Center.  Here are the reasons I use to sell this Condorcet compliance:

Two axiomatic principles:

1.  One-person-one-vote.  It doesn't matter if I really really really like my candidate over yours and you only marginally like your candidate over mine.  Your tepid vote for your candidate counts just as much (but no more) as my enthusiastic vote for my candidate.  This is because you and I have equal franchise.  This, of course, leaves out Score Voting as a consideration.

2.  Majority rule.  If more voters mark their ballots preferring Candidate A over Candidate B than the number of voters marking their ballots to the contrary, then Candidate B is not elected.  This was explicitly the failure of IRV in Burlington Vermont in March 2009.

Further desired properties:

3.  Avoiding spoiler effect.  If more voters express on their ballots that they prefer Candidate A over Candidate B, that preference should be unaffected by the presence of a third candidate C.  If Candidate C enters the race, Candidate B should not be elected if more voters express on their ballots a preference for Candidate A.

4.  Removing the burden of tactical voting.  This is what Fairvote and other RCV advocates keep saying: "Vote your hopes, not your fears."  We want to remove the burden of tactical voting from voters and allow them to vote sincerely.  Again, in Burlington 2009, there were about 1510 voters that marked the Republican as their first choice, the Democrat as their contingency choice, and the Progressive was their last choice.  We promised them that they could vote their hope (their conservative choice) without fear of electing their fear (the left-wing progressive choice).  That was an empty promise in 2009 and we still cannot get Fairvote and the IRV happy-talkers to acknowledge that.  If 1 out of 4 of those 1510 voters had insincerely marked the center candidate, their contingency choice, as #1 instead of their true first choice, they would have prevented their "fear" candidate from being elected.  This "Vote your hopes, not your fears" slogan should apply to everyone, not just those on the left.

BTW, both Score Voting and Approval Voting fail this tactical voting property at the get go.  I keep asking, and never get a satisfactory answer from proponents, how much should I score or approve my second-choice candidate (or third).  There *is* no satisfactory answer.

5.  Precinct summability.  This is for election transparency and integrity.  There is something "fishy" about having to opaquely transport the tally records to the central location where it is opaquely tallied in this kabuki dance called "STV".  A Condorcet-compliant method can have precinct subtotals reported *at* *the* *precinct* before the voting machines are transported securely back to city hall for the official aggregate tallying.

6.  Convenience to voters in removing the need to return to the poll for a runoff.  The legitimacy of an election is greater with greater voter turnout.  If less than half of the voters return to the polls for a delayed runoff, that runoff election carries less democratic legitimacy than if the entire crowd weighs in.  Making elections less convenient to the average voter decreases turnout. 

These are the six selling points that I am using to promote a Condorcet-compliant RCV method.  IRV-BTR is what I am trying to sell, because it has the simplest and smallest modification to the existing proposed RCV language.  Simpler and more understandable legal language has promotional value.  Something like Ranked-Pairs or Schulze would be a harder sell even though I think they would be better methods.  If there is no Condorcet Winner and if the Smith set is three candidates (the simplest case), IRV-BTR will elect the candidate with the most first-choice votes out of the three (also assuming the three top vote-getters are the Smith set).  While this would happen rarely, it's a natural way (requiring no additional language) to resolve a cycle, but RP and Schulze (using margins) might resolve it differently.  However, I am willing to live with that because, to me, what's more important than "which Condorcet?" is "GET Condorcet!"

> (As far as I know, the standard way to extend IRV to do equal-rank is to
> count the first preferences fractionally, but I don't know of any public
> elections that have been conducted by such rules.)

Any kind of fractional vote scheme is dead-on-arrival.  STV in any governmental election cannot have equal ranking because we would not know how to transfer the vote for two equally-ranked candidates or to count them if they bubble up to the remaining shared first-choice.

r b-j                  rbj at audioimagination.com
"Imagination is more important than knowledge."

More information about the Election-Methods mailing list