[EM] Sincere Condorcet Cycles

[Abd ul-Rahman Lomax]

At 06:34 AM 12/15/2009, Raph Frank wrote:
>On Tue, Dec 15, 2009 at 3:31 AM, Abd ul-Rahman Lomax
><abd at lomaxdesign.com> wrote:
> > Now, consider Deliberative Process as an election method. What would
> > deliberative process choose? Quite possibly, the negotiators would toss
> > another variable into the mix. Suppose that there is a Clarke tax. Suppose
> > that this tax is designed to be the value to each participant of
> > optimization of choice. If these are travel distances, it's the 
> cost of that
> > travel, paid to each voter, there is a travel subsidy given with the
> > proceeds of the tax to those who "lose" the vote, and there is a net tax
> > paid by those who "win." (I won't give the details.)
>
>A Clarke tax is only paid if the result is nearly tied (only voters
>who's vote changed the result are charged). If the voters have to pay
>no matter what, honest is no longer the best strategy.
>
>Your suggestion seems to be more a travel equalisation scheme.

That's correct. Technically, what I proposed was only schematic as to 
effect: the same tax is paid by all voters, but is refunded depending 
on the result, so that the result is neutral for all voters, ideally. 
The voters, really, are simply informing the system of what their 
cost is for each option. I did not specify details, how artificially 
high costs would be set aside, etc. What matters is the differential 
costs between the options, and with a deliberative decision-making 
system, it is no longer simply "voting," it is active negotiation and 
deliberation, an iterative process with far higher flexibility than 
voting systems.

>In Ireland, TDs are refunded for costs of travel and accommodation
>while attending the Dail.  TDs from "distant" constituencies are given
>a larger "mileage" expense.  Obviously, the system is massively abused
>(as TDs themselves decide the rules), long time TDs often end up
>receiving enough to pay the mortgage of a house in the capital with
>the funds.  But in principle, it is a reasonable idea.

Sure. That's often the case, a reasonable idea is abused; that's a 
sign of lack of independent monitoring. Asset Voting would actually 
create a penumbra of electors with real electoral power but without 
seats; rather, they choose the seats, the seats directly represent 
them, through visible votes, and those votes can be made reversible. 
Asset would be a true reform, with vast possibilities, it is far more 
than an election method, it is a method of *representation*, and even 
more, it is a technique for *communication* between the public and 
the representatives.

With Asset, the sense of people that government is "us" would 
increase, because the connections would be visible to all voters. 
They would know exactly what effect their votes had, and they could 
adjust their votes accordingly in the next regular Asset election. 
Candidate you voted for wasted your vote? Vote for someone else next 
time. You don't have to consider "electability" at all!

> >t is about time that the value of iterated voting is
> > recognized by voting systems experts. Because it hasn't been 
> valued, top two
> > runoff has been neglected and not supported against attack by 
> instant runoff
> > voting, which is inflexible and chaotic, as we well know. Top two runoff,
> > when the voting method is vote-for-one, is subject to an obvious flaw,
> > Center Squeeze, but that is easily fix by using a better method, such as
> > Bucklin or Range; if Range is to be used, the method should include an
> > explicit approval cutoff (I suggest mid-range).
>
>Well, the second round is a very obvious cost.  Even if IRV is
>actually more expensive, it is easier to convince people that 2 rounds
>must be more expensive.

That depends on how often they occur. Yes, a second round of voting 
is a cost, as is a third, etc. But standard deliberative process 
requires this, it's fundamental to democratic decision-making, and is 
never bypassed except through strong necessity. Robert's Rules 
recommends never bypassing it. But, of course, deliberative bodies, 
as part of their sovereign privileges, can make their own rules. The 
traditional rules based on long practice can be changed, and the 
tradition includes that.

My point is that communities willingly took on this cost when they 
began requiring a majority in elections. In some cases, candidacies 
blossomed, that's why some districts in San Francisco were always 
going into runoff. The solution, in my mind, the obvious one, is to 
use methods which are more efficient at finding a majority in the 
first round. Bucklin. But also educate the voters.

My general take, though, is that it should be the voters who decide 
if more than one round is necessary. That's clearly a situational 
decision, there is no general answer.

>   Also, it is reasonably likely that top-2 is
>more expensive for the winning candidates, in terms of campaign time.

Not necessarily. They may spend less in the first round, because it 
is no longer their only chance. In theory, 2-round should be more 
efficient overall, because campaigning can be more targeted.

>Also, people see value in a decision being made, even if it isn't the
>optimal decision.  In principle, for something like the House of
>Representatives, this might even be true, as if the seat is not
>filled, the State gets less representation.

Failing to find a result in two rounds would be very rare, even if a 
majority is required in the second round as well (usually it is not 
in public elections); Robert's Rules, however, does not limit the 
number of rounds, and for good reason. The majority can at any time 
decide that this is enough and can develop a method, agreed upon in 
advance, for completing an election. All it would take is a majority. 
But if they are not ready to decide, they won't do that. The majority 
is sovereign; if the king's delay in making a decision is causing 
harm, what do you do? You inform the king, but you do not take away 
the king's sovereignty, it's still the king's decision. If the king 
needs more time to consider the evidence, the king takes more time, 
if the king has come to two possible choices, and delay is causing 
harm, and the king can't choose in the available time, the king may 
toss a coin, or defer the decision to a council, or do any of many 
possible things that are within the power of the king.


>A compromise would be to add a question to the ballot asking if the
>voters agree that the plurality winner should win, and there is no 2nd
>round if the majority vote Yes on that option.

Yes, that's an option. However, an informed electorate would know 
that this would be the consequence of not adding lower ranks. It has 
the proper balance with Bucklin, which can be very flexible.

>Alternatively, voters could indicate which candidates are acceptable
>and the election finishes if the plurality winner is acceptable to the
>majority.

Bucklin! That is the function of the lowest rank, even in traditional 
Bucklin. Voting for more than one in that rank was allowed.

>   The 2nd round could be between the plurality winner and the
>most approved/accepted candidate, or maybe the election is just
>repeated in full.

Probably not pure repetition in public elections, but, yes, that's 
one decent way of determining the named candidates in the runoff. But 
the runoff should still be open for write-ins; if it's a two-rank 
Bucklin election, they are harmless. And they just might uncover a 
real Condorcet winner.

>Also, this breaks the prisoner's dilemma effect of plurality.
>http://leastevil.blogspot.com/2009/12/prisoners-moderate-voters-dilemma.html
>
>If supporters of one of the 2 large parties vote for a moderate, then
>the other side can betray them and then the other side's favourite
>wins.  Thus there is an incentive to defect.
>
>Thus both sides defect; each side only votes for their own candidate.

Complex strategies are unlikely in Bucklin.

>With the above rule, they could vote for the moderate, and designate
>the other party's candidate as non-acceptable.  Thus either the
>moderate wins (if the other side cooperates) or the election fails.
>This eliminates the prisoner's dilemma effect.

Majority requirements defuse the effects of complex strategy.

>Ofc, using approval directly would be better, but the above method
>only requires counting the first choices and then count that
>candidate's approval total..

Bucklin is simply ranked approval. (Thus it is quite compatible with 
Score Voting, later.)

> > Note that in some top two runoff implementations, write-in votes 
> continue to
> > be allowed in the runoff, which under some conditions fixes, to 
> some extent,
> > Center Squeeze, if there is sufficient preference strength in the
> > electorate, and with a good method, the possible spoiler effect in the
> > runoff can be avoided. TTR with good voting systems is really 
> quite close to
> > ideal.
>
>In principle, assuming that write-ins are allowed, the 2nd round could
>fail to elect a majority winner too.

Very unlikely with Bucklin, first of all, but, second of all, the 
plurality winner in such an election is quite likely to be the best 
choice. Only if the electorate actually prefers a third repetition 
would choosing the plurality winner be a problem. Rare.

>A similar effect could be achieved by adding a NOTA option.

In theory, that should always be allowed, and when there is a 
majority requirement, it's easy. Just write in NOTA. Under Robert's 
Rules, that's a vote, but it would not count for any candidate; it 
would be, nevertheless, part of the basis for a majority.

Requiring a majority, or at least attempting to require it, is 
clearly superior to making premature decisions!