On 09/04/2015 01:11 PM, Vidar Wahlberg wrote:

I think an intuitive first approach is to consider party list like a
single-candidate system where each party runs a bunch of clones and
where voting for the party is translated into voting for all the party
candidates in that order.

In open list PR, a voter may rearrange the candidates on the party's
list. This would be similar to the voter voting the party candidates in
a different order, though in practice, it usually is more like each
voter rearranging the order of the clones according to another voting
procedure run only on the ballots for the party in question.

If there are three parties (X, Y, Z) and three candidates for each, and
each party lists the candidates in order 1, then 2, then 3; then voting
for X in a party list system is similar to voting for X1 > X2 > X3 in a
candidate-based system. The analogy is exact (I think) for largest
remainder methods: if you do that kind of expansion with party list and
run the result through STV, you get the same result (per party) as if
you did party list PR allocation by largest remainder with a Droop
quota. I'm not completely certain which candidate methods correspond to
Sainte-Lague; I've heard that Monroe's method does this, but I'm not
sure if it does or if it corresponds to Hare instead.

Part of the reason that you'll find more discussion about candidate
methods than party list, I think, is that there's an idea that candidate
methods give greater freedom. In candidate methods, the voter can decide
which candidates in a party he prefers the most, and thus not only rank
between parties but also within them; and he can also vote for an
independent if so inclined, or mix candidates from different parties
(e.g. rank the candidates he thinks are particularly honest ahead of
those he thinks are less so, irrespective of party affiliation). Thus,
for candidate methods, there's no need to favor parties as such: a
candidate may run with or without party support.

Party list methods are usually a lot simpler than similarly fair
candidate methods, though, and you can do biproportional representation
with them. It is possible to do something similar to biproportional
representation with candidate methods (see e.g. Schulze's STV-MMP[1]),
but again, that's more complex and then parties are part of the system
again.

Going by the reduction or analogy above, that seems right. The method
passes LNHarm if appending some parties to your ranking doesn't reduce
the number of seats given to the party/parties you ranked before
appending. The method passes monotonicity if raising a party never makes
it lose seats, and lowering a party never makes it gain seats.

In http://www.votingmatters.org.uk/ISSUE3/P5.HTM, Douglas Woodall
defines some clone criteria. In particular, he defines "clone-no-harm"
as "replacing a candidate x by a set of clones should not harm another
candidate y". He then says that this is incompatible with the Droop
proportionality criterion. Imagine, for a candidate method, this election:

1000: A
1: X>Y>Z

for three seats. The outcome would be {A, X, Y}. But then clone A:

1000: A1>A2>A3
1: X>Y>Z

and it's evident that the outcome will (and IMHO, should) be {A1, A2, A3}.

But for party list, this is not a problem as long as the rules require
that every party needs to field at least s candidates to run. In the
situation above, nothing would happen between the first and second
instance: A would get every seat in the first election and A1 would get
every seat in the second (assuming a 3-seat election).

So in short, yes: party list can pass independence of clones whereas
candidate methods with fixed seat sizes have to give clones some seats
in certain cases. One way of looking at my question to Steve Bosworth
about cloning is as a question of whether his weighted vote system
suggestion is closer to party list or a candidate method.

[1] In particular, this:
http://permalink.gmane.org/gmane.politics.election-methods/13620 seems
quite akin to biproportional representation.

On Fri, Sep 04, 2015 at 04:51:16PM +0200, Kristofer Munsterhjelm wrote:

I'm not entirely convinced that candidate methods are superior to
party-list methods. I do agree that they give greater freedom, but
they're arguably more likely to turn the election into a popularity
contest where who the candidates are become more important than their
policies. Party-list systems won't fully remove the popularity contest
either, but they help shift the focus away from the candidates to the
policies of the party the candidates are representing.
Shifting focus away from charisma to ideology/policies I would argue is
an important goal in political elections, albeit it's not an easily
measurable goal as opposed to the voting system criterions.

This (clone-no-harm) is a good example. If you were to translate a
party-list method to a candidate method where candidates were simply
ordered X1>X2>X3>Y1>Y2>Y3>Z1>Z2>Z3 (and disallow voters to rearrange the
candidates, i.e. closed list), then I believe STV would pass this
criterion. It would not if you allow voters to rearrange the candidates
(i.e. open list), as the votes may end up being evenly spread for one
party, causing that party's candidate to be excluded earlier in the
election process.

--
Regards,
Vidar Wahlberg

Some discussion on properties of list and candidate based methods follow.

Some people don't like party lists because they don't like parties in general, and they think using a method that does not use the party structure at all could help in reducing the power of the parties (or their main offices) somehow.

Party lists at least seem to force candidates to represent a fixed policy. This is however more related to the ability of parties to nominate the candidates.

Party affiliations make it more difficult to the candidates to collect votes by telling different stories to different audiences. This is one key reason why party affiliations can be seen as a central feature supporting voters in making sensible decisions. I mean that one key problem of the concept of democracy is that most voters do not follow politics vary much and do not have time to evaluate all the candidates and their opinions. Therefore they will benefit of having some clear groupings that represent some typical policies that voters might want to support. The true opinion of the voters can be measured better this way.

Party lists (and other methods that support groupings) make it possible to have a large number of candidates (more than the voter wants to handle when filling the ballot). It is easy to indicate support to a named grouping, or allow a vote to an individual candidate to benefit also the related party / grouping.

One problem of pure party list methods is that they do not measure second preferences. This makes it difficult for small parties or other groupings or individual candidates to run. The voters feel that their vote is wasted if they vote for a small grouping that will not get any seats. This problem could be alleviated by some tricks (e.g. by allowing second preferences to be indicated by voters (or by parties)).

One problem of pure candidate based methods is the above mentioned problem of handling numerous candidates. This problem could be alleviated by allowing voters to name parties / groupings.

As you say, candidate based methods allow some populist candidates, or candidates that are well known for other reasons, or just pretty to look, to win seats easier that candidates that are good but without such vote collecting features. Also open lists have been criticised because of this possibility. Closed lists are not immune to this problem either since parties tend to nominate popular candidates also on the closed list. I don't think the number of candidates that are very popular for other reasons but poor politicians is very high anywhere. And on the other hand pretty faces and deep voices help everywhere. Note also that allowing voters to decide freely this way fights another possible problem, namely the ability of the party offices to favour the inner circle and candidates that are most loyal to it. What is the best balance between allowing the parties to decide which candidates will be elected vs. allowing the voters to freely decide (including the ability to elect some pretty and popular faces)? Note that the dividing line may be more already between open and closed party lists, and not between party list and candidate based methods.

(I'll skip the influence of primaries. They would just make the discussion more complex without adding much to it.)

I have sometimes promoted also methods that allow party internal groupings to be named. This would help voters in making decisions, and this could help implementing party internal proportionality in list based systems. Some parties may not like this approach since they want to promote uniformity within the party (maybe to be stronger and to promote their key values in an uniform way, or maybe to emphasize the role of the central office of the party and to eliminate any risk of opposition or even party fragmentation).

If one wants to avoid party internal groupings (fro the aforementioned reasons), one could still allow candidate ranking within the party. This is one way of trying to capture the best parts of both list and candidate based methods. (This would work specially in countries that already are list oriented. In countries that are candidate based method oriented, one can approach the same problem by introducing groupings to the candidate based methods.)

There are thus benefits and problems on both sides. Some mixtures of the properties of both sides could make some good methods.

Juho

On 09/06/2015 11:10 AM, Vidar Wahlberg wrote:

That's a valid concern, and you could argue that it even happens in very
small parties in party list (e.g. The Christians fielding an Elvis
impersonator). On the other hand, some countries already use STV, and I
imagine that if those elections were to become too much like a
popularity contest, I would have heard of it. I haven't heard of say,
concerns that the the local elections in Scotland are turning to
popularity contests - or for that matter, that lower house elections in
Australia have been thus transformed.

Granted, I have already argued that the upper house elections are
anchoring the lower ones in Australia -- that IRV has a Duvergerian
effect and that this effect is stronger than STV's multiparty effect. So
in Australia, one could claim that the same holds for STV's personality
focus as a candidate method: that IRV pulls more strongly towards party
focus than STV can push towards candidate focus.

It is, of course, also possible that I simply haven't heard of the
concerns, rather than that they haven't been made. But as you say, it's
not as easily a measurable goal as voting criteria, and so I have to go
with what I know :-)

Perhaps hybrids with party list (or MMP) on a national scale and
candidate methods on a local scale would help retain policy focus while
still giving popular local candidates some pull in their parties. On the
other hand, I'm not aware of any personality contest complaints in
Ireland either, and Ireland uses only STV, not an AMS or IRV hybrid.

I think the Droop proportionality criterion ensures what you say for
closed list to some degree, but I'm not sure if that's enough. The paper
that I linked to shows that STV also fails Woodall's other two clone
criteria.

But the open list scenario you mention shouldn't be a problem in itself.
Say we have something like

1000: X1 > X2 > X3
3: A1 > A2 > A3
2: A2 > A3 > A1
1: A3 > A1 > A2

(three seats)

If we were to use IRV directly, then what you say happened would happen:
X2 and X3 would be eliminated and the outcome would be {X1, A1, A2}. But
STV guarantees some representation by the Droop proportionality
criterion, which it achieves by handling surpluses. In the setting
above, the Droop quota is 251.5. So X1 would be elected and the quota
deducted, leaving 748.5 votes. Then X2 would be elected and the quota
deducted, leaving 497 votes. Finally, X3 would be elected.

If you have one party that's evenly split and another that isn't, say:

1500: A1 > A2 > B1 > B2
750: B1 > B2 > A1 > A2
750: B2 > B1 > A1 > A2

(two seats, Droop quota is 1000)

first A1 would be elected and the quota is deducted to give 500 votes
for A2. Next, A2 is eliminated and B1 is elected. So it doesn't seem
that evenly splitting made much of a difference here. The Droop quota
still says that {B1, B2} has more than a Droop quota and so should get
more than one seat, while {A1, A2} also has more than a Droop quota and
should get a seat.

The tricky part is finding out if the wiggle room provided inside the
DPC gives open party list STV enough rope by which to hang itself, so to
speak. We know it can trip up as a candidate method (as Woodall shows
in that it can fail both clone-in and clone-no-help); the question is
whether one can reach any of those failure cases from a party list
expansion.

On Sun, Sep 06, 2015 at 04:31:13PM +0200, Kristofer Munsterhjelm wrote:

I must admit that I'm not very familiar with politics in Scotland nor
Ireland. Like you, I too rely on my own experience as I don't know of
any studies or papers on how party-list methods vs. candidate methods
affect the political discussion. Looking at the US and UK elections,
which arguably are examples of quite bad voting systems, it's my
impression that there's far more focus on the candidates than you would
typically see in Scandinavian countries. Obviously, correlation doesn't
imply causation, so there could be entirely different reasons for this
(and of course, I may have a bias that cause me to notice what
strengthens my beliefs, while [involuntarily] overlooking facts that
doesn't support by beliefs).
If you expand a party-list system to i.e. STV, I suspect that people are
more likely to say "I vote for <candidate>", rather than "I vote for
<party>". This may be a driving force to focus on the candidates rather
than the policies of the party the candidates are subscribed to.

I see that what I wrote was both misleading and incorrect. STV fails
clone-no-harm regardless of whether voters are allowed to rearrange the
candidates or not. I'm not sure where my mind wandered when thinking
otherwise.

Did you mean that it can fail both clone-no-harm and clone-no-help?
Woodall claims multi-seat STV should meet the clone-in criteria (but
presents no formal proof of this claim).
What I fear with an expansion from party-list to STV is vote management.
As shown in the paper from Woodall you linked to, adding clones can
help/harm other candidates. Arguably Schulze STV solves some of this
concern, but that system adds a fair amount of complexity.

And finally a slight digression, what I dislike with STV is that popular
2nd preferences can be excluded early on:
20 A>B>C>D
1 B>C>D>A
20 C>B>D>A
20 D>B>A>C

To reduce the amount of e-mails and since I don't have much to add, I
won't reply to Juho's response directly, but I found his input
interesting and valuable.

--
Regards,
Vidar Wahlberg

On 09/07/2015 11:44 AM, Vidar Wahlberg wrote:

I agree about the US; I'm not as familiar with the UK. But the US is
untypical in other respects as well. What we'd really need is, as you
say, studies about which countries focus more on candidates' charisma
and personal attributes and which focus on more on issues and policies.
I suspect there are also confounding factors: I'd imagine less
democratic countries focus more on candidates as people. The ultimate
example of that would be "strong man"-type presidential systems.

The size of the constituency might also be of importance. I was reading
a weblog post about the direct mayoral election trials in Norway, and
the post said mentioned that, roughly translating, "Electoral research
indicates that the less populated the county is, the more the people
themselves matter" (http://bit.ly/1hQRm3S). If that is true and can be
generalized, then that would imply that Plurality would be more focused
on personal aspects than say, STV would, because the former only elects
a single winner per constituency and hence these have to be smaller.

I misread. Yes, you're right; Woodall doesn't show that STV fails
clone-in, and he does show that it fails clone-no-help.

Right. I don't mean that we should use STV as a party list method, but
that imagining doing so will go a long way in showing what kind of
criteria can be applied to ranked party list methods: if a candidate
multiwinner criterion makes sense after you force ballots to fit the
expansion, then the criterion makes sense for a party list method as
well. And if a criterion doesn't, it might do so after the ballots are
restricted to party list type (as was the case for clone-no-harm).

That's probably because it has an IRV-like core. The surplus rules act
as a kind of "hard stop" that prevents IRV's dynamics from violating the
Droop proportionality criterion, but STV is still IRV within those
constraints.

A similar effect could occur in your party list plus elimination method
(where you run party list and eliminate the weakest parties by first
preference until every party gets at least one seat) if there are
multiple very small parties. "Everybody's second choice" among those
small parties might get eliminated prematurely.

More generally, within the constraints, elimination methods will follow
their unconstrained logic. An IRV-based multiwinner method will act
mostly like IRV (picking the strongest wing), and something like "STV
but eliminate the Condorcet losers" will act mostly like single-winner
Condorcet (electing candidates closer to the general center; i.e. moving
the sweet spot that will get a candidate elected as close to the
over-all center as it can until it bumps up against the Droop constraint).

Neither of these is ideal, but to deal more properly with it, I think
you have to use more complex rules. Methods that don't directly
eliminate (e.g. my Bucklin method) will probably be somewhat better, and
methods that optimize for some condition or take all combinations into
account (Monroe or Schulze STV) will be better yet. But they are complex.

Well, non-fractional Monroe can be stated quite simply in this form:

1. Each voter has a satisfaction score for each candidate (his rating of
   that candidate).
2. Each candidate can be assigned (linked to) either zero or v/s voters.
   If v is not divisible by s, duplicate each ballot until it is. (For
   party list: each party can be assigned kv/s voters, where k is an integer).
3. Assign voters to candidates so that the total satisfaction is maximized.
4. The candidates who are assigned more than zero voters win.

... but the problem is that it's both very hard find the true optimum
and to verify that it is indeed the optimum. It's also Range-like; it's
possible to make an MJ version, but it would be more complex. Finally,
it is (I think) quota-based and so would fail population-pair
monotonicity. Making it divisor-based would again (probably) require
greater complexity.

Making a simple well-behaved proportional method is hard.

I suspect that Webster/Sainte-Laguë is already maximally well-behaved
for its simplicity: if you want greater fairness, it's going to come at
the cost of it being more complex. But how much more, I don't know.

(I also suspect that perfect proportionality is going to be NP-hard for
most ranked and rated ballot systems. So if you need to go beyond a
certain level of quality, it'll always be worst-case very hard to find
out who won. But I don't have a formal proof of this. My suspicion is
based on that similar problems, like clustering, are NP-hard, and some
voting problems can be formulated as instances of these.)