[EM] piling on against IRV

[robert bristow-johnson]

On May 9, 2010, at 4:45 PM, Kathy Dopp wrote:

> Oops. Sorry. I see I'm wrong again (to save you the trouble Robert). I
> see you did say 55% of the original turnout on election day numbers,

which is the correct number to compare the the 93% turnout in IRV  
runoff.

> so I see your point is valid.  Sorry, I'm trying to go too fast and
> get onto other things.
>
> However, that does not alter the fact that the voters have the CHOICE
> to vote or not in the real runoff elections, unlike with STV/IRV.

they have the choice in *either* TTR or IRV.  and *both* can fail to  
arrive at the Condorcet winner who is the only unambiguous majority- 
preferred candidate.  but TTR runs the *extra* risk of electing an  
even *less* preferred candidate than IRV.  in a 3-candidate race, at  
least IRV will not elect the Condorcet loser, but we know that both  
TTR and IRV can exclude the Condorcet winner from the runoff but,  
because of reduced voter turnout, TTR might very well elect the most  
losingest candidate of the three (the candidate *least* preferred by  
the electorate).  at least IRV will avoid that.  that is why you are  
silly for claiming that it got better when IRV was repealed.  it only  
got worse.

even though his values are wrong (in my opinion) the anti-IRV U  
Vermont prof (Gierzynski) that wrote an oft-referred analysis of what  
was wrong with IRV, one thing he said is true (he thinks this is good,  
and i think it's bad): the repeal of IRV will create pressure (if they  
want to win) for people with more aligned political interest to unite  
behind a consolidated candidate.  it will push Burlington back to a  
predominate 2-party environment.  we'll get to choose between Dumb and  
Dumber, and if you ever dare to vote for Smart, then Dumber will get  
elected.

> Voters may decide not to vote in the runoff for many reasons, such as
> not caring which of the two finalists win, but they ought to have the
> right to participate and to have the opportunity to learn about the
> two finalists and make a decision if they want to

and you have NEVER shown that they were not allowed to with IRV.   
anyone who expressed an opinion about *either* Bob Kiss or Kurt Wright  
participated in the election between the two finalist.  no one  
prevented them from doing so.  93% chose to do so and less than 7%  
chose not too.  no one forced that decision upon them.  now if you  
want to complain about how IRV chose *who* would be those two  
finalists, i can join you in that complaint.  but TTR would do no  
better and you have done *nothing* but try to obfuscate to show  
otherwise.

> - ALONG WITH THESE
> OTHER RIGHTS THAT IRV/STV REMOVES:
>
> 1. IRV/STV remove the right cast a vote with a positive effect on a
> candidate’s chances of winning.

non-monotonicity.  we agree about that.  if, on their way to the  
polls, 744 Wright voters had changed
their minds and decided that they liked Kiss instead and changed their  
vote to Kiss, the result would be that Kiss would have lost and the  
Condorcet winner (Montroll) would have been elected.  so who's  
complaining?  and how would TTR fix that problem?  if the same thing  
happened with TTR, then it would be Montroll who would go to runoff  
with Kiss (the Republican would have been left out).

> 2. IRV/STV remove the right to participate in the final decision of
> who wins the election

no it doesn't.  no more than TTR.

> by eliminating voters’ ballots prior to the
> final counting round. The more candidates, the more voters are
> eliminated prior to the final counting round.

oh, so you're complaining about *who* it is that goes to the final  
round.  well i complain about that, too.  but between the two that  
make it to the final round, IRV places no obstacle for anyone to  
participate in that election.  all they have to do is, by election  
day, make a decision about who, in a contingency, they support and  
mark their ballots.  but IRV *does* screw up in how it advances the  
candidates to the final round, there is no disagreement from me about  
that (and there never was).  but where we disagree is that you seem to  
thing that TTR would have done better, and it would not have.  if it  
were TTR in 2009, the same two losers (the 2nd and 3rd preferred  
candidates) would advance to the runoff and our choice would be only  
between those two.

> 3. IRV/STV remove the right to have one’s votes counted equally and
> fairly with all other voters’ votes because only voters supporting  
> the
> least popular candidates as their 1st choice are assured of having
> their 2nd choice candidate counted when their 1st choice candidate
> looses.

so people whose 1st choice is not eliminated are complaining?  if your  
1st choice is left standing in the final round, what's the problem?   
if it's about the 2nd choice not counting before the final round, then  
that complaint is about the method that IRV has in evaluating and  
eliminating candidates and determining who advances to the next  
round.  fine, we agree that IRV sucks, but TTR would have done no  
better in Burlington in 2009.

> 4. In comparison with top‐two runoff elections, IRV/STV remove the
> right to elect majority winners.

*both* IRV and TTR runs the risk of removing the majority-preferred  
winner before the runoff, leaving us with a choice of voting for one  
of the two losers in the runoff.  but IRV will at least pick the  
lessor loser while TTR, with it's reduced turnout, runs the risk of  
picking the biggest loser.

> San Francisco had to eliminate its
> legal right to elect majority winners when it adopted IRV/STV because
> STV routinely elects winners with far less than 50% of the votes.
>
> 5. IRV/STV remove the right to a transparent, verifiable election
> process with a decentralized, simple counting process that can be
> easily manually counted and audited.

precinct summable.  definitely a preferred property and Condorcet  
(using the same ranked ballot that IRV uses) is also precinct  
summable.  IRV *could* be precinct summable but the number of piles  
grows quite large if there are any more than 3 candidates.  but the  
transparency and verifiability can still be accomplished in a *small*  
venue (like a small city or county, but not for a statewide or  
nationwide election) by, at the precinct level, handing each  
legitimately interested party (and media pool representative) a thumb  
drive that has the same ballot information that Burlington published  
and that we were able to use for our own computer programs to rerun  
the IRV or Condorcet or Borda or Bucklin or whatever that uses the  
same ranked ballots.  precinct summable is a good thing and we should  
seek to have it, but it is not precisely the same thing as transparent  
and verifiable.  you can still have the latter (in a small venue)  
without the former.

> 6. IRV/STV removes the right to have an economical election process.

well, if people weren't fussing so much about it, we could have  
recovered the onetime $10K costs in Burlington when 1 runoff was  
avoided.  nothing needed to be changed in the optical scan voting  
machines (because they only scanned and recorded the ovals and did not  
count anything in the IRV election), the counting software was free  
and sorta public domain and appropriated from Cambridge MA who still  
uses IRV.  the money (now wasted) was used in voter education and in  
training of the clerks/judges and was a onetime cost.  there was  
literally no extra money needed in Burlington for the "process".

> 7. IRV/STV removes the right to change one’s mind between the  
> primary
> and general election and to have time to get to know the candidates.

ah, now we're getting to the core truth (without the little lies and  
obfuscations).  that is a salient difference in values you have  
pointed out.  what you say is true except it's debatable that it's a  
"right".  and i think that is, at the core, what the IRV opponents in  
Burlington didn't like.  i see no reason that people can't be expected  
to make up their minds by Election Day.  it's what we normally require  
from voters anyway.

according to this "right" you espouse, then Condorcet or any other  
method that uses a ranked ballot and fully resolves the election on a  
single Election Day also violates that "right".  i guess, so also does  
FPTP with no runoff, also violates that "right" to change one's mind  
after Election Day.

> SURELY THERE ARE ALTERNATIVE VOTING METHODS THAT DO NOT MAKE THE
> ELECTIONS PROCESS LESS FAIR AND REMOVE SO MANY VOTER RIGHTS!!

when people start yelling, i stop listening.

> I am sure that Condorcet, Approval and other methods that can also be
> applied in PR elections would be far less problematic and destructive
> of voting rights and even be an improvement over plurality.

twisting this issue into one of "rights" is and always has been  
obfuscation.  the same rules applied to me as to those complaining  
that their "rights" were violated.  people younger than 18 can  
complain about "their rights" because different rules apply to them  
(the rules let me vote, but not them).  non-citizens can complain  
about their rights because different rules apply to them.  perhaps  
even i can complain about my rights to influence an election for an  
official in a town or county that i don't reside in (perhaps i own  
property or a business or something in that town) but i doubt i'll get  
many sympathizers.

the issue is what method correctly (or most correctly) reflects the  
will of the electorate without rewarding strategic voting swinging an  
election or punishing sincere voting.  twisting it into a diatribe  
about rights is just dishonest.  the rules that applied to Burlington  
residents that were anti-IRV are the same rules that applied to me.   
no one's rights were violated and it's a canard to inject such  
language into it.  i hope the pro-IRVers don't adopt the same  
dishonest tactic in complaining that their rights are now being  
violated because they no longer have IRV.

--

r b-j                  rbj at audioimagination.com

"Imagination is more important than knowledge."