[EM] Response to Kristofer Musterhjelm

[Kristofer Munsterhjelm]

Since many of your replies follow the same pattern, I'm going to address 
them where I first see each. Thus I may veer a bit from replying only to 
the issue quoted.

I've also been a bit slow in writing this. I'll blame work and play :-)

-

David L Wetzell wrote:

>>    KM:Your first points seem to suggest that if you had PR, the problem
>>    of IRV leading to two-party rule (or 2.5-party rule, as I call it,
>>    since the NatLibs aren't one party) would be ameliorated. But
>>    Australia has PR - STV, to be precise - for the elections for its
>>    other house, and still that doesn't break the 2.5-party rule of the
>>    Senate.

> No, I'm saying that we don't need to end 2 party rule in the USA to make 
> our political system work considerably better.   IRV makes it so that 
> the 2 major parties will have to center themselves around the true center.  
> 
> I advocate the use of 3-5 seat PR in "more local" elections so as to 
> handicap the rivalry between the two major parties, increase the number 
> of competitive elections and to give third parties a more constructive 
> role to play in the US political system: ensuring the protection of 
> economic/ethnic/ideological minority rights.

My whole point about pointing at Australia in this context is that 
Australia already has PR for one of its houses. Because IRV is used in 
so few places, there are very few examples we can use to see how IRV 
would behave, and in the absence of such examples, it would all come 
down to what we *think* would happen.

Therefore, I'll use Australia again. Its IRV chamber should give some 
idea of what you would expect to see in terms of IRV elections, because 
each member of parliament is elected on a single-winner basis from a 
single district.

If IRV produced competitive local elections, you would expect to see 
some of those local elections have an effect on the national IRV seats. 
Then the major parties would move to reclaim the seats, and so readjust 
themselves closer to the center. Say that Congressional seats were 
elected by IRV and because of the local elections advantage you mention, 
the Progressives grew strong in Vermont. Then you might see at least one 
Progressive member elected to the House from Vermont. This would then 
signal to the major parties (probably the Democrats) to consider whether 
to move to the left to recapture those seats.

However, in Australia, there are very few third-party seats in the IRV 
chamber at all. According to Wikipedia, minor parties only won seats in 
1946, 1996, and 1990-2010. The mean number of third party seats for 
these years were 2.9, or 2.1% of the house (because of changing number 
of seats).

Either the major parties were very good at anticipating the direction 
they were to move to the center (and very good also at avoiding any 
slipups), or the feedback is weak indeed.

State election results also show the state heads - the states' 
equivalents of prime ministers - to be either NatLib or Labour most of 
the time. See the Wikipedia article about Politics of Australia for a 
graph of that. I know it's not analogous to Governor elections because 
the legislature picks the state head, but in PR countries, small parties 
may get the prime minister anyway.

Finally, IRV may not actually improve ethnic divisions. See 
http://rangevoting.org/FraenkelG.html .

>>    KM:You might argue that Australia only has half PR (i.e. in one of
>>    its chambers) whereas you'd want the US to have PR for its
>>    legislature and IRV for executive elections such as Governor and
>>    President. Thus, it would make more sense to compare with a nation
>>    that elects a President by IRV and has PR everywhere else. Yet
>>    Ireland, which has an STV legislature and an IRV president, doesn't
>>    show a good record of that breaking two-party rule (fixed two-party
>>    rule, even) for the positions that *are* elected by IRV

> dlw:I take as given that the economies of scale in running for 
> single-winner offices will make it so that there are two major parties.   

I'll consider it likely that economies of scale will favor major 
parties. It does not, as such, need to favor them so severely that there 
are only two viable options for the candidacy, however.

 From a logical point of view, Plurality elections favor major parties 
because minor parties would split the vote. Therefore, it's important to 
know who the candidates most likely to win are, so that you don't 
actively hurt your case by voting for a minor party of your preference 
and splitting the vote. What is the problem here? It is that voting in 
the way you prefer would hurt you.

In IRV, voting in the way you prefer doesn't hurt you... until the third 
parties get large enough that they might affect the elimination order. 
When they do, the election can swing the wrong way, so voting the way 
you want can hurt you if your choice would have made a difference - i.e. 
if minor party candidates have any chance of winning.

So when we go at this from the logical point of view, the extent to 
which single-winner methods favor major parties seems to differ. I would 
point out that top-two runoff works, and that (as I showed in another 
post of mine) it does give a number of effective candidates 
significantly greater than two -- but you don't consider top-two runoff 
a single-winner method. Since I do, I don't think single-winner methods 
(or should we say, methods that are used to elect individual candidates 
for single positions) have such a steep barrier that only major parties 
can participate.

If you disqualify the runoff method, then pretty much all you're left 
with is Plurality and IRV. In both Plurality and IRV nations, you see 
major parties winning almost all the time, even when counterbalanced 
with PR (e.g. Australia). So then you could of course say that from our 
{Plurality, IRV} based sample, all single-winner methods elect 
overwhelmingly from major parties, and you could say that it's because 
of economics of scale.

To me, it seems these economics of scale are of the sort that are needed 
to compensate for the methods' failure modes. You need to unite to win 
in Plurality, and that kind of unification, almost by definition, favors 
the major parties greatly. You also need to unite to win in IRV because 
IRV has problems once the third parties get large enough.

If I'm right, then the extent of major-party dominance isn't given by 
the single-winner approach, but by the tool used. What we'd really need 
would be examples of non-{IRV, Plurality} methods being used. How did 
the Bucklin method fare in Minnesota?

(And I know about Borda and Nauru Borda. The latter is pretty much 
Plurality, and the former has severe cloning problems that rewards those 
who can field the most candidates.)

>>    KM:Ireland doesn't have compulsory voting, yet the President has
>>    been elected from the same party almost every time. There have been
>>    two exceptions: the first, before the setup of parties, and then a
>>    Labour victory recently.
>>
>>    Unfortunately, it's hard to find further examples of IRV going in
>>    either direction - simply because IRV is not widely used. What I
>>    *can* find on a national level is Malta (which is somewhat of a
>>    special case because it's two-party even where STV is used) and Fiji
>>    (which is no longer a democracy, to my knowledge).

> dlw: My ideal type isn't in use anywhere.   
> <http://anewkindofparty.blogspot.com/2011/03/strategic-election-reform-explained.html>
>
> It was sort of used in the US, since the economically important state of 
> IL used a 3-seat quasi-proportional election rule for its state 
> representatives from 1870-1980.  This handicapped the rivalry between 
> the two major parties so that states who were economically dependent on 
> IL were not politically dependent on IL and had freedom to experiment.   
> (It'd be hard to know how things would have turned out in the US without 
> the innovations done in WI under LaFollette's Progressive Party in the 
> early 20th ctry.  )  Our system, without the use of PR in "more local" 
> elections tends to have had one party dominate a state's politics.  And 
> yet, because there are so many states, there's been scope for a third 
> party to become dominant in a state's politics, which has forced the two 
> major parties to accommodate them(like with the Democrat-Farm-Labor 
> party in MN).  
> 
> If there tends to be two major parties under my system then it will be 
> two parties per state and it won't necessarily be the same two parties 
> per state.  In this manner, the duopoly will be contested by minor 
> parties.

That sounds like a runoff method. You have an Approval first stage and 
an IRV second stage. (This initially threw me off because "AV" also is 
an abbreviation for "Alternative Vote", another term for IRV.)

Perhaps the runoff aspect of the combined method can outweigh the 
two-party dominance of IRV. On its own, IRV seems to give two parties, 
but also to give them such power that they aren't really contested by 
minor parties. IRV has at least kept the same major parties major in 
both countries where it's used, despite the PR that pulls in the other 
direction. So again, I think IRV's hierarchy force is stronger than PR's 
pluralizing force. It might be weaker than Plurality's, but if you get a 
two-party situation in both cases, then what's the difference? As for 
trying to fix IRV by applying a runoff round, it seems like a hack. If 
you're going to depart from the one-stage-IRV approach, let's get it 
right. Let's use a good method, not a fix that may or may not work on a 
method that, in my opinion, doesn't get far enough away from the 
uncontested duopoly tendencies of Plurality.

Even if you disregard that, IRV still tends to amplify small changes 
into great ones. If you have a narrow victory in any round of IRV's
elimination, you might have to do a recount in that one  (assuming it
has an impact on the subsequent eliminations), not just in the final round.

I also think that if you have to have an IRV3 second round, the first
round should be proportional. Otherwise, the cloning problems I have 
mentioned earlier may happen and thus make IRV3 itself redundant.

Consider this situation, for instance. You have Republicans and 
Democrats, and either R or D will win the approval vote. Then both 
parties field, say, ten candidates, and all who like the Republicans 
approve of only the Republicans - and all who like the Democrats approve 
of only the Democrats. Say without loss of generality that the Democrats 
win. Then all ten Democratic candidates will have more Approval votes 
than any Republican, and so the next stage will be filled with 
Democrats. Therefore, a Democrat will win, no matter how the IRV3 stage 
behaves. The same thing holds for the Republicans: if there are ten R 
candidates and all R-votes strategically approve of all of them, the 
IRV3 stage will be populated entirely by Republicans and so have little 
meaning (except to apply the IRV logic to determine *which* Republican 
wins).

>>>       dlw: A two stage election has a winner-doesn't-take-all first
>>>       stage and a winner-take-all second stage.  So it's still
>>>       consistent with single-winner leading to hierarchy and
>>>       multi-winner leading to plurality.

>>    KM:I'd still consider it single-winner, since at the end of the day,
>>    a *single* winner is elected to each district. I can be flexible
>>    here, though. The top-two article on Wikipedia mentions that top-two
>>    runoff is a game (in the game-theory sense) because the voters can
>>    adjust their votes from one round to another.

> dlw: All election rules can be viewed as like games.  My point is that 
> one can account for the greater pluralism with a top-two runoff by the 
> fact it's first stage is not single-winner.  So it's consistent with my 
> theory that it's single-winner elections that lead to relatively few 
> parties.  And since there inevitably have to be important single-winner 
> elections, it's also inevitable that there will tend to be two dominant 
> parties (or two dominant party coalitions, which will act not unlike the 
> two dominant parties)...

(Regarding games, I meant more properly that it's a dynamic game, i.e.
one where the games-players can adapt to earlier moves. An ordinary
election would be a simultaneous game, but it could perhaps have some
aspects of a sequential/dynamic game in the presence of polls.)

You're equivocating here. First you say that "single-winner elections
lead to relatively few parties" (where single-winner means those that
have only one stage). Then you say that "there inevitably have to be
important single-winner elections", where single-winner means elections
to single seats. There inevitably has to be no such thing by your first
definition of single-winner. France uses single-winner in the second
sense by having legislative elections, as well as presidential ones, use
top-two runoff.

To avoid further confusion, let's call the one-round methods
"single-winner" (as by your terminology) and the methods used to pick a
winner for an individual position "single-office". You haven't shown
that there must inevitably be single-winner elections, only that there
must be single-office elections.

>>    KM:In any case, if the runoff method leads to plurality, why not use
>>    it? It clearly doesn't produce paralyzing Plurality - France tried
>>    PR and then returned to top-two as they considered PR to fragment
>>    too much, which implies that top-two didn't. Furthermore, runoffs
>>    have seen wider use than IRV, including in the US, and so practical
>>    cases can be used to argue for it; and it seems to overturn
>>    Plurality more often than IRV does.

> dlw: I'd rather use 3-5 seat forms of PR in "more local" elections to 
> ensure plurality.  I'm okay with using IRV for "less local" elections, 
> because I value both plurality and hierarchy at the same time.  

I think IRV gives too much hierarchy - so much that its dynamics isn't
noticably different from Plurality. I try to back that up with 
references to Ireland and Australia, as well as pointing out where the 
problems start to show up with IRV (when third candidates start to 
matter and are no longer just for show).

If you want a Plurality level of hierarchy, just use Plurality plus PR.
(I wonder if any country actually does that.)

>>    KM:FairVote claims that IRV is cheaper than top-two, but others
>>    disagree. Warren argues that IRV needs more complex voting machines,
>>    and that this and the fact that IRV can't be precinct-summable could
>>    erase most of the gains from not having a runoff.

> dlw: I'm not an expert on the cost of the machines, but extra rounds of 
> elections are quite costly... and IRV can be precinct-summable if 
> IRV3/AV3 is used.  If one treats the ranked votes as approval votes then 
> one can get three finalists almost immediately.  Then, one can summarize 
> the votes by sorting them at the precinct level into the ten ways one 
> can rank the three finalists: 6 rankings of 2 of them, 3 rankings of one 
> of them, and 1 ranking of none of them.  
> 
> So the cost arg for IRV is a valid one.  And I'm cool with advocating 
> for IRV plus (more local) PR, as opposed to IRV alone...

Is IRV3/AV3 single-winner or just single-office? If it has two rounds,
then you are dealing with a runoff. If it has only one round, and you
infer the Approval data from the ranking, then you're going to have to
have some sort of communication (or central counting) to know which
three candidates you're going to pass to the IRV stage, and thus it
isn't strictly summable.

>>>       dlw wrote:And could we not argue that the difference between two
>>>       major parties and two coalitions of parties isn't as great as we
>>>       may think it is?  

>>    KM:The greatest difference, to my mind, is that having coalitions
>>    brings politics out in the open. Anybody may read the record of the
>>    parliament or legislature, and alliances between what would be the
>>    wings of the two parties are very clear, particularly in
>>    parliamentary systems where the continued existence of the executive
>>    rests on the coalitions remaining as they are.

> dlw: It's also possible that one can bring too much info into the open 
> and thereby gain an degree of obscurity.  Lots of inevitable compromises 
> in politics are easier to work out when there's less transparency.   

I can only speak for my own country, but that has never been a problem
here. I haven't heard of such being a problem except for places where
they have other, more serious, problems too (like the participants of
the political system trying to game said system by any means available 
and so actively muddy the situation as much as they can).
Perhaps Juho can tell us whether it's a problem in Finland, but I don't 
think it is there, either.

The transparency means that parties can't promise an unwavering position 
at the center of their area of political space. They will have to 
compromise in order to get what they want, and so pull the mean position 
of political decisions closer to their point. In a world where parties 
are expected to be hard and unflinching, this may seem hypocritical, 
since they can no longer be the defenders of their positions through and 
through. In reality, the people get used to the more pragmatic nature of 
give-and-take, and can use the record to judge how well parties are 
negotiating for their position when they can't get it all.

I also have a more abstract argument in favor of transparency. If you 
have too little transparency, you can't really synthesize more 
transparency out of it. However, if you have "too much" transparency, 
the media or independent efforts like OpenSecrets can aggregate the 
wealth of data for the people's benefit.
So if there are disadvantages to transparency, it's either that the 
feedback by the people upsets the process too much -- or it is that said 
media and independent efforts not being up to the task, the torrent of 
data overwhelming the people.
When small towns feel too small because everybody knows everything about 
everybody else and thus nobody has any privacy, it's the former. When 
one speaks of information overload, it's the latter.

>>    KM: Beyond this, coalitions also give a greater flexibility. To give
>>    an example of this: in Norway, lately the christian democratic party
>>    has made statements to the fact that they would ally with the
>>    right-leaning parties rather than the left-leaning ones, but only if
>>    the populist rightmost party isn't too strong in the remaining
>>    coalition. The Christian party has conservative and liberal shades
>>    of Christians in it, and this is probably an internal compromise. As
>>    such, it keeps the right bloc from veering too far to the right.
>>    If we'd had a two-party system, all of this would have happened
>>    under the covers, out of sight. You could argue that the
>>    right-populist party wouldn't have had a chance at all and so the
>>    outcome would be the same, but the right-populists' voters know why
>>    they have less of a say now than if everything was done internally.
>>    These voters can also consider whether they need the Christian
>>    democratic party at all, and the conservative voters of the CDP can
>>    consider if they should vote further to the right.

> dlw: If the rivalry of the two major parties is handicapped so that 
> there's more circulation among the elites and third parties can win some 
> seats and threaten the duopoly, there'd also be greater transparency.  

By indirect means, perhaps. Yet we don't know that, as we have so few 
examples. What I've seen of IRV countries seem to indicate that the 
rivalry of the two major parties isn't handicapped enough. The major 
parties stay major parties, and not even the PR counterweight dislodges 
them to make other parties the new major parties.

And if you want to optimize a certain metric, it's best to directly 
optimize it, rather than something that is corelated with it. If you 
want transparency, I think it's better to make the political system 
transparent than to try to find some change (IRV) that leads to a change 
(handicapping) that leads to transparency. If any link of that chain 
fails, the whole thing fails, so the fewer links, the better.

>>    KM:So, in simple words about the above: having multiple parties lets
>>    the voters decide more about how politics is made. Since coalitions
>>    are fluid, they are ultimately responsible to the voters.

> dlw:  Do voters decide how ad hoc coalitions among multiple parties are 
> worked out after a PR election?  Hardly.  If coalitions are fluid then 
> it's also hard to get things done and it's possible to blame the other 
> coalition members for problems.  One can even bet on becoming a part of 
> the next coalition if this coalition falls apart so long as you keep 
> your base happy. 

The voters don't directly determine what the coalitions are like. The 
voters don't veer off from one extreme to another, either, so they have 
some idea of which coalitions are realistic.

In the Christian party example, a voter might consider whether to vote 
for the party, or, if he thinks their compromise is too soft, vote for 
another more right-leaning party instead. The advantage to voting for a 
more right-leaning party is that he knows he'll strengthen the 
particular coalition that he does want; but the disadvantage is that he 
has to accept that party's position on issues where he agrees more with 
the more moderate party. He can think about which is better, then 
express that opinion in terms of his vote. If the party sees that its 
voters are fleeing, they then know the voters didn't want that 
compromise after all, and they can reposition themselves.

Or in US terms: with a coalition, a tea-partyist or an Occupier might 
directly vote for a further right-wing or left-wing party knowing that 
this party will either make it more likely that the more moderate 
parties will ally in that direction, or that the further right/left 
party will serve as a stronger opposition if the other coalition wins. 
Tea-partyists or Occupiers have little chance to alter the power dynamic 
directly by voting in the current system, since they would vote 
Republican or Democratic respectively anyway - the only way they *can* 
change things is either internally (where the tea party has had quite a 
bit of success) or by threatening not to vote at all.

Thus the coalition system is more responsive because the voters can say 
what part of what would otherwise be one party that they like, instead 
of just having to take it or leave it. It's not perfect, of course: 
voting power is not linear and there's a lot of negotiation going on 
that may turn the power distribution less representative of how the 
voters voted - but it's better.

In a way, the coalition systems behave like the proxy methods that have 
been discussed here, though of lesser fidelity. In a proxy method, you 
vote for someone who votes the way you want, and you change your vote if 
he stops voting the way you want. The alliances then shift because the 
voters' support of the proxies (or the policies of the different 
proxies) shift. A coalition does the same, but more coarse-grained, and 
a major-party system more coarse-grained still.

> If you have two major parties then the replacement of one with the other 
> being in power is a meaningful incentive to rule well.  It can be 
> improved upon, but having two major parties is not inherently bad.

Not necessarily. Another pattern that might emerge is that one party 
appeals to one group, the other to the other. Then the first party says 
"these guys are all corrupt, as you've seen, so vote for us instead". 
They win and resume their corrupt practices with the second party in 
opposition. Then the first and second party switches places and the same 
thing happens with signs reversed. Each party has a solid base to which 
go the spoils, and an undecided middle who vote for "the other party" 
(whoever was in opposition the last time around) because they want 
things to change. No party wants to permanently claim the middle group 
by becoming more honest, because they gain more by perpetuating the 
system than by cleaning themselves up.

AFAIK, that happened in certain South American states, but I am not 
completely certain of this. You could also consider the lobbying dynamic 
in the US weakly related, with each party's base being made of different 
types of wealthy donors.

This might be fixed by making the field competitive enough that the 
model doesn't pay, but I can't find evidence that IRV gives honest 
parties a chance to unseat the old major parties and become new major 
parties. At least I know that a multiparty system will give competition, 
if for no other reason than that it's much harder to buy off ten parties 
than two.

>>>       dlw wrote: thus, I don't think it requires the end of a
>>>       two-party system to restore the US's democracy.  If we have a
>>>       contested duopoly and a host of LTPs(local thid parties who
>>>       specialize in contesting "More local" elections and vote
>>>       strategically together in "less local" elections as part of
>>>       their more general issue-advocacy) checking the influence of
>>>       $peech, it'd suffice.

>>    KM:.. but it seems like it's just the effects of a steady erosion.
>>    You had something a bit like social democracy with FDR. If you'd had
>>    a viable multiparty system, I think that system itself would have
>>    given support to that position, as well as countered the later
>>    departure towards one guided increasingly by money.

> dlw: Well, we had an infusion of greater concentration of wealth among 
> the financiers who funded the rise of Nazi Germany in t he US...  The 
> military-industrial-congressional complex took advantage of the 
> chronically non-competitive elections in most "more local" elections.  
> 
> I do not contest that the US's political system has been in need of 
> reforms for some time.   I simply do not believe that an EU multi-party 
> system is the goal we should be aiming for.

I don't believe IRV can give you the competition you need for a truly 
*contested* duopoly, either. IRV is a patch, but the patch doesn't go 
far enough.

> dlw: Well, I am in favor of the use of 3-seat Largest Remainder Hare for 
> state representative elections so as to enable the proliferation of LTPs 
> who would collectively check the influence of $peech on both major 
> parties.  They won't try to rival the two major parties, but they could 
> help a minor party gain ground on them.  This could lead to a merger so 
> we might end up with somehting like a Green Democratic party and a 
> Libertarian Republican Party, with both parties being more green and 
> libertarian(or less corporatist).

The only Australian mergers I know of are ones between minor parties 
(e.g. Australian Party and Queensland Party) and between major parties 
(the proposed Liberal-National merger which, if carried through, would 
make Australia a true two-party system, not just two-and-a-half). I 
could have missed something, though: do you know of any such mergers 
there, or in any of the other IRV nations?

> I'd rather push for a contested duopoly than a contested monopoly.  I 
> think it's a more feasible "evolutionary step" that will engender less 
> opposition from those in the two major parties in the US.  As for what's 
> next, we will see...

A contested duopoly is of course better than a contested monopoly. I 
think a free "market" would be better still, and even if you want a 
contested duopoly, IRV won't give it to you.

There's an unfortunate side-effect to the way FairVote has been linking 
the ranked ballot itself to IRV, as if to imply: "there's one ranked 
ballot method, and that is IRV: if you want a ranked ballot, you want 
IRV". To some extent, I think the IRV-supporters among liberal wings of 
the parties also have this idea. They support the ranked ballot (the 
idea is obvious! Have a contingency if your main vote doesn't count), 
and then, because IRV has been marketed so heavily, they say "Ah! I want 
a ranked ballot, that means IRV!". So those who don't like fairness 
won't support a ranked ballot anyway, and of those who do, they're 
supporting the ranked ballot more than they're supporting IRV.

(A Washington Senator tried to get the legislature to vote on using the 
Schulze method, so non-IRV methods aren't entirely unknown. However, it 
was blocked by the committee leadership. See 
http://groups.yahoo.com/group/Condorcet/ for more about that.)

>>    Perhaps this is akin to the extent to which one would expect
>>    strategic behavior to appear in single-winner elections.
>> 
>>    Instability can also occur in places that are transitioning to a
>>    democratic form of government. Early post-Soviet Poland and Hungary
>>    are good examples. I think they settled down, though.

>     dlw: Habits matter more than rules for democracy.

Maybe one could say that the rules make some paths easier to reach than 
others, but the habits determine where one actually *does* go.

>>>       I think we can trust in the politics of Gandhi/MLKjr (and
>>>       hopefully #OWS) more so for the crucial sorts of changes needed.

>>    Should we? Movements like Gandhi's, MLK Jr.'s, and Occupy Wall Street
>>    are "release valves". 

> dlw: they are more than release valves. Self-sacrificial acts on behalf 
> of others can move people to be more selfless or to rebuild a political 
> culture that enables other much needed reforms..  

>>    When conditions get too bad for the groups in
>>    question, change forces its way through. Nations that are less
>>    democratic get their own release valves, too, which fire less often but
>>    more fiercely. If they are blocked altogether, say by a secret police,
>>    the "vessel" eventually bursts in a revolution.

> dlw: IMO, We have just begun to see the effects of such movements and 
> their potential to move the political center that the parties center 
> themselves around.

Yes, after a long while, you're getting some change. My point is that 
this change takes a long while and it doesn't happen at once; in the 
meantime, you have a period of discontent that builds up. You could say 
that therefore we can go with the most marketable slight improvement 
because if it's not good enough, groups outside the political system 
will reform it later -- but that ignores the long period of bad rule 
before the group grows strong enough to "not take it anymore" and large 
enough to actually have an effect.

Let's have proportional control, not bang-bang control. It's gentler on 
the system and reaches the (changing) target more quickly.

>>    KM:So it seems here that you have a continuum. On the one extreme,
>>    desires for change are picked up and made part of the political
>>    course. On the other, the dissatisfaction keeps building up until
>>    there's an explosion.
>>    Now, the US is a democracy, so I don't think there will be a
>>    revolutionary explosion, but the movements suggest that those that
>>    are part of it don't get their say in the ordinary political
>>    process. This means that there's been a significant period of time
>>    in which they've considered the situation quite bad - enough for the
>>    metaphorical pressure to build up. Wouldn't it be better for the
>>    system to adjust itself earlier? That is better for the people -
>>    because they get a world closer to the one they want sooner - and
>>    it's better for the system, which doesn't exhibit as sudden and
>>    drastic changes.

> dlw : I agree.  This is why I'm pushing for the electoral reform that is 
> most likely to be adopted the soonest.

If it isn't good enough, you'll simply have traded off gradualism in the 
political process in general for gradualism in adopting the election 
method. Since there are many more decisions not involving elections than 
there are decisions involving them, the benefits to the smooth 
transition of the election method is outweighed by the disadvantages in 
having to wait for another extra-political movement to correct it 
properly the next time (or for extra-political movements to fix what 
would otherwise be fixed through the political process).

Further, if the method isn't good enough, the political movements may 
end up moving away from it. "We tried that, it didn't work", as I've 
said before. FairVote's attempts to link the ranked ballot and IRV (such 
as by using the name "Ranked Choice Voting") would make such a reversal 
more likely. "We tried the ranked ballot, and IRV didn't work/just gave 
us weird results, so forget about it" - something like that.

>>>       This would elect a president with broad appeal who will then be
>>>       sheltered from the partisan rivalry for control of the Senate or
>>>       House of Representatives.

>>    If that's what you want, I don't think you need a single-winner
>>    method at all. Either have party primaries or use a method like STV
>>    in a grand jungle primary, in either case narrowing the field to 6
>>    or 7 candidates. Then use a PR method (probably the same as in the
>>    first stage, if that's what you'll use) to narrow it down further to
>>    three. Finally, have the electoral college decide. Since the college
>>    itself is small, it doesn't need a voting method - it can just reach
>>    a decision in the same way that the House or Senate does.

> I agree that there are plenty of alternatives for how to do the first 
> two stages.  I'm sure some sort of STV will be involved in the state 
> primaries...   

Well, that would get around most of our disagreement. If you use a 
proportional method like STV, then IRV never enters the picture, while 
you also get the election form you want.

> dlw: There'd be no advantage to Republicans creating shadow parties. 
> There'd be no official restriction (other than seven) on the number of 
> candidates a party can have among the final seven.

I misunderstood the method. See an earlier paragraph about how cloning 
could affect the method, but very simply put: it can make the first 
(non-PR, Approval in your case?) method so decisive the second round is 
little more than an intra-party formality.

> I think it'd be hard for a party to get 3 virtual clones among the 7 
> finalists if a form of PR is used in the first stage.  The strategy 
> would be to get two of your candidates who get along okay among the 
> three finalists so as to max the chances of one of them becoming 
> president.  

So use PR for the first stage, yes. (Or cut the second stage out 
altogether.)

>>    KM:This strategy would be a kind of decoy list strategy, in that the
>>    party that wishes to game the system makes "shadow parties" (instead
>>    of shadow or decoy lists) to get around limits that are enforced per
>>    party.

> dlw: You're forgetting that the US will likely still have a two-party 
> dominated system with a relatively even split between the big two due to 
> the exigencies of having so many single-winner elections and if PR is 
> used, there'd be no need to have quotas.   It'd be quite hard for a 
> party to get more than 4 candidates among the seven.  

What do you mean by "no need to have quotas"? All forms of PR has some 
type of quota, whether fuzzy (in Sainte-Lague or Webster) or strict (in 
the Hare or Droop quota).

If you're referring to my reply regarding cloning, you don't have an 
explicit quota, but you have a limit on the number of candidates. 
Strategic nomination combined with a non-PR method can fill the 
available slots with members from one party alone under certain 
circumstances.

>>>       dlw wrote: I think more practically that IRV3/AV3(uses a limited
>>>       form of AV to get three finalists) enforces the maintenance of a
>>>       two-party dominated system.   It also tends to be somewhat
>>>       incumbent friendly.  This makes it easier for it to get adopted
>>>       by legislators, who mostly are going to be incumbents.  

>>    KM:It is possible to do a reform if the people wants it, even if the
>>    legislators do not. An example I like to point at for this is the
>>    adoption of STV in New York. It might even fit better with your idea
>>    of "politics by movement", since the proportional representation
>>    league (I don't remember its exact name at this time) was involved
>>    in promoting it.
>> 
>>    (Later, said league suffered setbacks as those in power linked PR to
>>    "Stalin Frankenstein" and called STV a Soviet invention - and STV
>>    was eventually removed. Such tricks, needless to say, would not be
>>    feasible now: the USSR is dead.)

> dlw: It's possible, but it's easier to get it sooner if you play 
> political jujitsu.  But I'm fine with local strategies.  If another STV 
> movement gained momentum in NY, I'd support it. 

It's easier to get something done if you play political jujitsu, but 
this something is limited by what's well known. Currently, that's 
FairVote's IRV as well as variants on the same format (contingent vote, 
three-choices-of-many IRV, etc). If I'm right and IRV won't lead to 
multipartyism nor weaken the big two enough, then selling IRV is a huge 
wasted chance. Instead, one should try to show better systems to the 
friendly politicians. Being politicians, they don't have time to go into 
detail and find out which of the many ranked methods are good, but 
that's where the declaration comes into play.

>>>       dlw: I've seen that before.  Like I said, if the center is
>>>       dynamic and so are the two biggest party, it's not that big of a
>>>       deal...

>>    At the moment of the election, the center is static, though. I'll
>>    reply to this in greater detail in the other post, but to be simple
>>    about it: all dynamic centers give you is a capacity for the parties
>>    to compensate. But that will discourage parties from forming where
>>    they could cover most voters, because it's at just this point where
>>    the weird effects occur.

> dlw:  sure.  But that compensation is what makes the diffs among the 
> rules not so great, and even if the center at the end of the election is 
> not covered that is less bad when one considers that the center is 
> inherently dynamic...

If the people change their opinions quickly, then they would tend to 
land in the middle of one of the parties' areas and the other parties 
would then ideally adjust their positions to move to the new center 
(because that's where the voters are); but by the next time, the center 
would have changed so much that it would land in the middle of some 
other party's coverage, and so you wouldn't expect to see the artifacts. 
That is true.

On the other hand, if the people change their opinions less quickly, or 
the way they move can be predicted by the parties, then the clustering 
happens before the election. The center according to the voters' 
preferences could be in the middle of a strange region, and as the Yee 
diagrams show ("Shattered" and "Disjoint"), these regions can be quite 
large.

Furthermore, if people change their opinions quickly, their standard 
deviation might be greater - and a greater sigma amplifies the oddities 
of IRV. To see this, consider the opposite situation, where the standard 
deviation is so small everybody agrees on the complete ordering. Then 
every nondeterministic method that passes the unanimity criterion would 
elect the same winner.

>>    KM:The question boils down to whether IRV (or IRV plus PR) is good
>>    enough to give meaningful political diversity, more precisely in the
>>    form of the major parties changing. It isn't in Australia. You could
>>    say that's because there's compulsory voting and how-to-vote cards
>>    in Australia, and then I could point at Ireland, and then you could
>>    say that the President there is ceremonial so the parties' hearts
>>    aren't in it. 
> 
> dlw:You raise good points.  
> 
> I doubt Aussies would want to switch to FPP plus PR or FPP alone.  
> I believe the Aussie Prime Minister is derivative of its lower house, 
> which uses IRV.  I would not recommend this.  The lower house should use 
> a 3-5 seat form of PR and the upper house should have IRV. 
> Single-winner elections are more meaningful in bigger elections where 
> the balance between the biggest parties tend to be stronger.  
> The strength of the US presidency is significant.  It makes our 
> prez(single-winner) elections important elections, which is another 
> reason why it's damn hard to end effective two party rule and very 
> important to prevent effective single-party rule.

I doubt they wuld want to do that, too. Australia's minor parties would 
prefer instead to have a change to PR in both houses (e.g. 
http://www.greens.org.nz/press-releases/trans-tasman-action-ge-labelling-proportional-representation 
).

I also agree that presidential elections are important. I might even say 
they are so important that they should be left to people who can elect 
and recall them at any time and keep them under scrutiny (i.e. 
parliamentarism), but I know that has absolutely no chance of succeeding.

Since that's not an option, I think it's very important to get the 
presidential election right. If the playing field under the new system 
is still tilted too far to give other presidents a chance, then the lack 
of alternatives for the presidential position can propagate throughout 
the system and harden the other branches. Also, the more imbalanced the 
field, the more important it is to be seen as one of the two "who have a 
chance", and that takes marketing and the heavy weight of the party 
structure.

Here again we come to the disagreement. I think IRV's field is too 
tilted to give the required amount of responsiveness to people's votes 
and that it thus won't lead to enough competition.

Finally, I also agree that in a parliamentary system with multiple 
chambers with different methods, the PM should be elected from the 
proportional chamber.

>>    KM:At some point, though, I run out of nations that use IRV. It
>>    could be the case that the few countries who did try IRV got
>>    two-party (or two-and-a-half-party) domination for different
>>    reasons, but it could also be the case that they did it for the same
>>    reason -- IRV.: 
>> 
>>    Then one has to ask, do we risk it? That depends on the expected
>>    benefits of getting IRV, weighted by the chance of actually getting
>>    it (by whatever means most likely), versus the expected benefits of
>>    a given other method like Condorcet weighted by the chance of
>>    actually getting that.

> dlw:That's a way to put it.   I think so long as we don't have too many 
> serious candidates in an election and a good number of considerably 
> boundedly rational voters then the purported diffs in values would not 
> make it worth it to push for a seemingly more ambitious election rule.  

You need at least some serious candidates besides the main two under at 
least some conditions, because otherwise the main two can just follow 
the strong pull of money instead of the weak pull of the voters. Beyond 
that, I can only rephrase that I don't think IRV can handle the domain 
between enough candidates to give the main two a challenge and too many 
candidates well. (For that matter, I don't think it can handle the case 
of many candidates, true multiparty style, either, but you don't 
consider that a problem.)

>>    KM:I look at Australia and at Burlington, and to me, the former
>>    doesn't speak very well of the benefits of getting IRV vs Plurality
>>    (you might compare parliamentary IRV-Australia to parliamentary
>>    FPTP-Britain here); and the latter doesn't speak very well of the
>>    chances of getting it *and* retaining it. In Burlington, IRV was
>>    left alone when it elected the CW, as it did in the first election,
>>    but then it failed to elect the candidate that was the CW and would
>>    have won under almost every other method considered, and the repeal
>>    followed.

> dlw:  You mean after a strong campaign was run against it???  If they'd 
> let well enough alone, the problem would have fixed itself and they'd 
> still have IRV, as opposed to FPP.  How many well-heeled folks wage 
> campaigns against FPP after it "spoils" an elections?

Note that Burlington IRV survived when it elected the Condorcet winner. 
I have a little idea of my own regarding this: when the Condorcet winner 
is elected, for any single group you may rally against the vote because 
their candidate didn't win, there's another group that preferred the 
Condorcet winner to that candidate, and the latter group always has a 
majority.

It is then considerably easier to organize a push against a method if it 
fails to elect the Condorcet winner than if it does elect the CW. 
Marketing and money probably had its part, but Condorcet failure could 
well have made it easier.

IRV's artifacts didn't stop there, and so gave the campaigners more 
ammunition. I'd say IRV's failure to elect the Democrat was a good 
example of its center squeeze, as well - its tendency to pick the wing 
with the greatest support rather than the center with even greater 
support because the center's votes are obscured by first preferences for 
the wings. The Republican candidate was, in a sense, a spoiler because 
his presence hid the Democrat that would otherwise have won.

I agree that the campaigners wouldn't have waged their campaign against 
FPTP. I do think they would have done so if the system had been changed 
from status quo to FPTP.

(And as a correction: Burlington doesn't have FPTP, it has top-two 
runoff with an odd 40% threshold instead of the more sensible 50%.)

>>>       dlw:What we need most is electoral pluralism.  We needed that
>>>       decades/centuries ago.  We had more electoral pluralism in ther
>>>       US in the past (like with the 3-seat cumulative voting in IL
>>>       from 1870-1980).  I believe FairVote can market critical reforms
>>>       to the US population.  I can't say the same thing for folks
>>>       pitching other alternative election rules and given that our
>>>       system uses primarily FPTP right now in the US, we can't afford
>>>       lots of alternatives being on the market.  It's too easy for
>>>       those who benefit from the status quo to divide and conquer us.

>>    KM:I agree with the need for pluralism, and so I would support
>>    proportional representation without a thought. A 90% incumbency rate
>>    is pretty appalling for a legislature.
>>
>>    Linking the proportional representation method to IRV, however...
>>    that's another matter. FairVote's strategy seems to be that the
>>    people would see the better results of IRV and so also support STV -
>>    but if the "better" results of IRV end up causing backslides over
>>    costs (from having to remake voting machines), lack of transparency
>>    (due to no summability), and weird results - then that could hurt
>>    STV too.

> dlw: costs aren't that great.  summability is possible under IRV3/AV3 as 
> I have described it, and the weird results are not common.

The University of Vermont seems to think it would be significantly more 
expensive to run IRV than top-two elections in Vermont. See 
http://rangevoting.org/VermontIrvCost.html . Besides that, I don't know 
enough about the use of voting machines, versatility, counting costs and 
so on for the US to comment, so I'll grant you the part about costs.

Similarly, I can see IRV/AV3 is summable, but with the caveat I 
mentioned earlier. If IRV/AV3 is in one round, where the approvals are 
inferred from the full rankings, then it isn't summable since you'd need 
to do an intersection operator on the rankings to get the 3-cand-only 
rankings; and if it is two-round, then it is a single-office election 
method, not a single-winner one. Then it *would* be summable, but you 
would be arguing against yourself when you say single-office methods 
can't be used.

As for the weird results, those can happen when the election is hard to 
call (close to the border of the cell in the Yee diagram given by 
candidate positions at the time of election) or voter opinions move 
slowly (for their sigma) so that the center-squeeze and nonmonotonicity 
phenomena show up. While not ubiquitous, the latter puts strain on IRV 
precisely when it has to meet the challenge - when it exits the 
Plurality domain and minor candidates start to matter.

> The big thing with FairVote is that they get people out of the FPTP 
> stupor, which is a major step in the US!!! 

FairVote gives by showing people that an alternative to FPTP is 
possible. They then take away by telling people that IRV is both a good 
alternative and the only alternative. I don't agree about the former, 
and the latter is somewhat of a self-fulfilling prophecy. FairVote 
markets IRV so heavily that their actions serve to turn IRV into "the 
only alternative".

>>    KM:You are right that we have been divided. I hope the declaration
>>    helps. Whether it does, time will tell, but it might, particularly
>>    if authorities within the field sign it.

> dlw: Endorsing 4 election rules and waving your hands over IRV hardly 
> seems very helpful.  
> Why not endorse pushing hard for American forms of PR in nat'l/state 
> representative elections (and city council elections) and trusting that 
> there'll be more demand for alternatives to FPTP once the rivalry 
> between the two major parties is handicapped and third parties have more 
> potential to spoil more elections...

I have no problem pushing for American PR. I get the impression that the 
single-winner first strategy was chosen because single-winner 
(single-office) elections are more familiar to voters than are PR 
elections. Is that wrong? If so, that's another way of handling the 
disagreement: unify around PR first and leave Condorcet vs IRV for later 
- but then so should FairVote.

However, even if we wanted to choose that strategy, those who organize 
voting might at any point ask "well, what of single-winner elections?". 
Then we can say "pick Approval, Schulze (e.g.), MJ or Range; authorities 
X, Y, Z, think they're all pretty good". We just have to get X, Y, and Z 
to sign.
If some local governments try any of them and find out that, say, MJ is 
good enough, then we can later say "X, Y, Z think they're all pretty 
good, and [county W] says they've had good experience with MJ".