[EM] Fwd: how goes American PR?

[robert bristow-johnson]

so Jan, i heard that you were for "keeping plurality voting over IRV in 
Fort Collins".  is that true?  do you continue to feel the same way 
about FPTP vs. IRV?

On 12/3/11 3:37 PM, Jan Kok wrote:
> The US President's power is huge. He can veto bills passed by
> Congress, and he can start wars. And now, de facto, can even order
> assassinations of US citizens.
this is more of an issue of (constitutionally) how much power a 
president or other nation's head-of-state should have.  whether a 
single-seat office holder has a lot (or too much) power or not, doesn't 
change the notion of how that person should be elected in a democracy 
and, particularly, a democracy that makes room for more than two viable 
parties and for independent candidates.

> So, it's important that we have good single-winner election methods to
> make the best possible choices of winners for single-winner offices.
it is regardless of how much power that single winner gets.  even for 
the official Town Clown, why award the office to the loser?  or the 
2nd-place winner?

> IRV/RCV is a poor method. It can make poor choices of winners, such as
> in the 2009 Burlington, VT mayor's race, and is more complicated than
> other methods.
it *has* made poor choices.  doesn't mean that it always had.  IRV 
doesn't do too bad when it elects the CW (and all the CW needs to attain 
is a place in the final round, then the CW is also the IRVW).  but the 
same argument can be used for the Electoral College vs. the popular 
vote.  it doesn't make much sense to keep the E.C. because most of the 
time it elects the winner of the popular vote which is deemed the 
measure of how well it works.  if that's the case, why not ditch the 
E.C. and just elect the popular vote winner?  same for IRV vs. Condorcet.

>   The complexity makes it difficult to sell to voters,
> some of whom are _extremely_ resistant to change.
yeah, but tax laws are complex too.  and we continue to pay taxes with 
complexity in the code (and some are _extremely_ resistant to that 
also).  some people tell me that Condorcet is more complicated than 
IRV.  i disagree but in any case reject the notion of adopting "simple" 
laws that are unfair eschewing those that have more subtlety but are 
naturally fairer.

>   The complexity also
> makes it more expensive to count the votes,
not really.  the scan/count machines can do fine with the ranked ballot, 
whether it's IRV or not.

IRV is harder to hand count.  but Condorcet would be even more laborious 
to hand count, i think.

>   and makes IRV elections more vulnerable to fraud.
all elections are vulnerable to fraud if there is corruption in official 
places and the rule-of-law is diminished.  the only manner that IRV is 
*more* vulnerable w.r.t. other methods is that it is not precinct 
summable.  it's harder to fix an election that covers many voting places 
if the results from the individual polling places cannot be tabulated 
and reported independently from each place.  now you can still do that 
with IRV (the media gets a copy of the same thumb drive that the 
precinct clerk takes to the central counting place), but the auditors in 
the media might not be able to easily check the overall results unless 
the method is precinct summable.

>   And when IRV gets rejected or repealed, as
> it has in several places, it poisons the well, making it harder to
> introduce other, better voting methods.

and that, i fully agree with.

> So, as long as there are people pushing IRV, let the quibbling (about
> single-winner methods) continue!

i agree with that, too.

-- 

r b-j                  rbj at audioimagination.com

"Imagination is more important than knowledge."