[EM] Preferential Party-List Proportional Representation (PPLPR)

[Vidar Wahlberg]

On Thu, Nov 06, 2014 at 03:34:52PM +0200, Juho Laatu wrote:
> On 06 Nov 2014, at 11:26, Vidar Wahlberg <canidae at exent.net> wrote:
>> The "wasted vote" mentality is still around in party-list democracies,
>> and parties that are liked by most, but not as their first preference,
>> will not gain much influence.
> 
> Yes. One approach to fixing this problem is to use systems that
> guarantee a seat to all parties that reach e.g. the v/r votes limit
> (where v = number of votes, r = number of representatives). One could
> go also further and for example allow "cumulative votes" and
> "proportionality in time" in the sense that "lost" votes (and maybe
> negative "undeserved" votes) in this election will be added to the
> results of each party in the next election. That would be one approach
> to eliminating the problem of lost votes as well as we can.

Interesting thoughts. I'm generally not a fan of election thresholds as
they do appeal to strategic voting (in Norway with 4% election threshold
there's indications that people voted strategically on the party
"Venstre" in 2005 to increase chance of victory one coalition by keeping
them above the election threshold, the same thing likely happened last
year to another party, "Sosialistisk Venstreparti", to increase chance
of another coalition winning), as well as it makes it very difficult for
new parties to gain influence.

Even my distaste for election thresholds I do see a potential problem of
fragmentation, and too many parties. Your example of accumulating votes
from the previous election to reduce "lost" votes is intriguing. My own
thoughts on the subject has been preference IRV-style, meaning that
votes to a party which doesn't meet the election threshold goes to the
voter's next preference (repeat process until all remaining parties are
above the election threshold).
And obvious advantage of your approach is that it doesn't require a
change on how votes are cast, another advantage is that a party
remaining just below the election threshold would sometimes still win a
seat. Some other features, which may be considered advantages by some
and disadvantages by others is that it could lead to more parties with
few seats, and such parties would likely repeatedly drop in and out
between elections, not keeping a stable presense.


> For sure the incumbents try to stop any changes as a general rule
> since the existing rules (whatever they are) are the ones that gave
> them the power. Any changes to the system would probably change some
> of the incumbent leaders. That is already a good enough reason for
> many to oppose (intentionally or instinctively) any changes. Any
> method that would not elect them would surely be a faulty method :-) .

Indeed. Sadly, this is why I believe certain nations, such as USA, is
doomed to their two-party system. Both democrats and republicans would
lose influence with a new system that opens up for newcomers, so
obviously they have no desire to change the system. And the people won't
dare voting for a 3rd party, because then, to phrase Douglas Adams, «the
wrong lizard might get in».


> Although methods that are based on candidates only, without any party
> structure, do have some benefits, so do the party-list methods. One
> key benefit of parties is that they give voters a simplified structure
> of the possible ways that the society could be developed further.
> Democracies rely on regular people making decisions on how the society
> should be run. Since they are not experts, it is important to make the
> alternatives clear enough to them, so that they can truly make
> decisons on the key questions. They should have some understanding on
> what direction the system should take, and which groupings are
> committed to which changes. This makes them less vulnerable to cheap
> advertising tricks, and makes it less likely that they will simply
> vote for those candidates whose name is best known, or that have the
> most expensive advertising campaigns, or that just have a nice smile.
> 
> Parties, ideologies and similar simplifications and groupings thus
> serve the basic need of a democracy to make it possible for the final
> decision makers, the regular voters, to make rational decisions on how
> the society should be run. A typical voter is not interested in
> following and studying the opinions and achieved results of maybe tens
> or hundreds of candidates. It is better to try to provide some simple
> structure that helps them in making these decisions in some rational
> manner.

This sums up pretty well my reasons for prefering party-list systems
over candidate based systems. You have one election for an ideology
(party), and then you have another election for which person within the
ideologies that should be elected. Granted, the second election doesn't
really matter in Norway, as currently the only way to modify the order
of the candidates is when 50% or more explicitly cross out a candidate
as unwanted. Something that never has happened before. This is however
something that likely will change within a few years, it's currently
being looked into.


>> So why doesn't this gain more attention?
> 
> It will. Actually the EM list has improved quite a bit on this respect
> in the recent years.

I do hope so. I think there are many borders we can expand upon in PR
voting systems.

Until then, I thought I'd show a fundamental difference between my
"PPLPR" and an Approval Voting style PR system:

For simplicity, there are three parties (L, C, R) and two voters (one
L-supporter, one R-supporter).
With Approval, where the two voters only approve of their first
preference, you would have one vote for L and one vote for R. 50%
representation to both parties. With PPLPR you would naturally end up
with the same result.
However, if the L-supporter decides to throw in a vote for C as well,
because it's better than R (while the R voter sticks to only voting for
R), then Approval would end up with one vote to each party, or 33.3%
representation to each party. In other words, by adding another party to
your ballot, you decreased your own party's representation (as well as
R's representation). This is where PPLPR differ.
With PPLPR, L would still get 50% of the representation, C would get
12.5% representation and R would get 37.5%. You increase the
representation of one preference by reducing representation from the
lesser preferred alternatives, while the more preferred alternatives
remain unaffected.
Why it becomes 50-12.5-37.5 and not for example 50-25-25 is because C
and R voters have the same opportunity to affect later preferences
(which wasn't used in this example), I went a bit more in depth on this
earlier.

I hope this is a more simple explanation of my goal. The way I see it,
systems where your most prefered alternative is directly affected if you
vote for other alternatives, are more likely to cause tactical voting
than systems that won't directly harm your higher preferences.
There still is an indirect harm with PPLPR, by ranking an opposing
alternative to your first alternative high up on your preference, you
may cause that opposing alternative to form coalition without your most
preferred alternative.

And then there's a problem with strategic nomination, namely splitting
up parties to gain influence and introducing mock parties to decrease
later preference strength (I wrote about this earlier in the thread).


-- 
Regards,
Vidar Wahlberg