Editorial/LTE submitted to Aspen Times
Jan Kok
I recommend Range Voting or Approval Voting as better alternatives to
the current runoff system used in Aspen elections.
The existing runoff system usually picks the same winner as the
previous single-round vote-for-one method (called Plurality Voting)
would. Separate runoff elections cost more, and voter turnout is poor.
Runoff elections also have a nasty habit of eliminating broadly
acceptable, moderate candidates and electing off-center candidates.
For example, if 34% of voters prefer Left, 32% prefer Middle, and 34%
prefer Right, Middle would be eliminated, and the second choices of
the Middle voters would determine the runoff winner. But note that
most Left and Right voters probably prefer Middle as their second
choice. Thus, Middle may be preferred by as much as 66% of the voters
over either Left or Right. But runoff voting can easily choose Left or
Right. This occurred in the 2007 French and 2006 Peruvian elections,
which use runoffs. (See RangeVoting.org/FunnyElections.html )
Returning to the previous Plurality Voting method is not a good
option. Remember, in Florida 2000, Nader took some votes from Gore,
causing GW Bush to win. Presumably, a majority of Florida voters
preferred Gore. Similarly, in 1992, Perot may have taken enough votes
from GHW Bush to cause Clinton to win.
Range Voting works like Olympic scoring. Voters give scores to all of
the candidates, and the candidate who gets the highest average score
wins. Range Voting tends to choose winners who give high overall
satisfaction to the voters. Range Voting, and an even simpler, related
method called Approval Voting, can be run on existing voting machines.
And these methods require only one round of elections - no runoffs!
See RangeVoting.org and ApprovalVoting.com for more info.
Jan Kok
Cofounder of Center for Range Voting
Fort Collins
Abd ul-Rahman Lomax
no-overvote rule and counting all the votes, with the candidate
receiving the most votes winning. In addition, I recommend leaving
the top-two runoff in place where indicated, it fixes a
commonly-perceived problem with Approval Voting. Counting all the
votes makes the need for a runoff less likely.
At 06:50 AM 6/7/2007, Jan Kok wrote:
>I recommend Range Voting or Approval Voting as better alternatives to
>the current runoff system used in Aspen elections.
Jan's letter brought to mind a possibility. Often communities that
want to pick a majority winner have runoffs if no candidate gets a
majority. Approval raises a new possibility, that more than one
candidate gets a majority. Sometimes the latter situation is used as
a criticism of Approval, because the first preference of a majority
can lose, where the majority also approved another candidate.
If there is good reason for a runoff if no candidate gets a majority
under Plurality, I don't see that the reason disappears with Approval
or Range (in the latter case, "majority" is not so simple to define).
My interest is in making initial changes be as simple as possible,
and this points to using Approval because of its use of the same
ballot and counting methods as Plurality, only we Count All the
Votes, instead of discarding over-votes on the assumption they are
unrecoverable errors.
(Over-votes were never discarded because of the spurious argument
that they are violations of one-person-one-vote, see Robert's Rules
on the subject. Approval votes are one voter per voter per candidate,
i.e., Yes/No, with No being assumed if Yes is not marked, and with
the winner being the candidate with the most Yes votes, exactly as
with Ballot Questions.)
However, there is no reason to discard the top-two runoff except for
cost. And discarding it for cost reasons is an entirely separate
issue. Leaving it in place, from an election quality issue, makes
complete sense to me. Without knowing how the runoff rules are
specifically worded in the law, I don't know if that part of the law
needs any changes at all. But I'd suggest that a runoff be considered
even if a candidate gets a majority, if there is more than one such candidate.
This top-two Approval process is theoretically superior to
single-stage Approval because it satisfies the Majority Criterion.
Some would argue that Approval chooses a better winner than the
majority criterion suggests, but this is more true for Range. The
black-and-white nature of Approval can cause anomalies where, in
fact, the majority might reject the election result. And that's a
possible violation of majority rule, which is a fundamental principle
of democracy.
(All election methods can violate majority rule unless some way is
included of gaining the explicit consent of a majority to the result,
the election of a particular candidate. While it may be possible to
incorporate that consent -- or withholding of consent -- on the
original ballot, it further complicates it. While I'm acknowledging
that Approval results can violate majority rule, this is also true of
Plurality and IRV, as examples, where the problem can be far more serious.)
(To incorporate consent on the ballot, a separate question would be
placed on the ballot. A description would be given of the rules for
determining the winner in this particular election, and the
conditions under which a runoff would be held, and then the Question
posed: Shall the election result as determined by the rules be
accepted without a runoff? Yes/No. Majority Yes waives the runoff.
The majority has a right to waive its sovereignty by delegating it to
a person or process. There is a problem with a general and permanent
waiver as occurs by "rule of law," that is, where the establishment
of a process and the assignment of the rights of the majority to it
has taken place in the past, with a different electorate. Making the
question explicit on the ballot resolves that and restores the right
of the majority to decide specifically.)
Dave Ketchum
Runoffs need discarding - ESPECIALLY those that, somehow, pick a subset of
the original candidates to choose among - because this picking too easily
omits the true majority preference.
Approval has its own problem as to ability to state preferences - easy
enough to vote here for beauty and against ugly - but how do Approval
promoters respond to "souse is tolerable if I cannot have beauty, but how
can I avoid voting for soso weakening my vote for beauty AND avoid voting
against soso weakening my vote against ugly?"
Range has a great argument - I can rank every candidate against a
scoreboard I establish. Trouble is, how do the vote counters match my
scoring to a single shared scoreboard.
IRV has problems that get it rejected, mostly for incomplete analysis of
votes.
Condorcet uses the same ballot as IRV, does complete analysis as to which
candidate is most preferred, and allows simple publication of its analysis
both as to complete election AND for any district.
C. concedes detail to Range, but notes that all its voters respond
to the same scoreboard.
NOTE: Various happenings encourage existence of more serious candidates -
and thus make quality of voting systems more important.
NOTE: NOT useful to brag that multiple voting systems often agree as to
winner - MOST do - but matters more as to their failures in recognizing
voter desires when they disagree.
Statements about existing voting machines are both:
Unimportant, for problem systems likely need replacement anyway.
Misleading, for they likely require that a machine capable of a
dozen Plurality races be replaced by up to a dozen machines - even if of
the same type.
DWK
--
da...@clarityconnect.com people.clarityconnect.com/webpages3/davek
Dave Ketchum 108 Halstead Ave, Owego, NY 13827-1708 607-687-5026
Do to no one what you would not want done to you.
If you want peace, work for justice.