[EM] SPPA - was STV district magnitude

[Stephane Rouillon]

I am back from vacation and I kept you for dessert...
I think this issue is most interesting to me.

James Gilmour wrote :

> Stephane wrote (> )
> > I used pure PR in the sense of any method that can produce
> > a perfect match between popular support toward parties
> > and seat won by those parties, to the integral limit of not
> > splitting representatives apart.
>
> You have again defined your objective as "PR of political
> parties".  This is the narrow objective that I reject.  There can
> be PR of many other aspects of "political" life, as expressed by
> the voters in their responses to the candidates.  What about PR
> by gender, by religious affiliation, by ethnic origin, by
> professional or work activity, by significant issues that cut
> right across all party boundaries?

Political parties are used:
1) to identify the government after an election;
2) to assemble activists around a unified and coherent vision
of what are the priority debates and what should be the society laws
in those regards.

I understand that not using the political party as an entity at the election
level
does not stop from using it to identify the government after.
But if it is not necessary, it does not make it useless. On the contrary, I
believe
most voting losers would apreciate some sort of representation using
some other candidate from the same party instead of nothing to say to
their representation. There can be and I would like to see after an election
"PR of many other aspects of "political" life". Nothing stops these to be
reached through political parties... Is there a law that forbids creating
parties if
debates "by gender, by religious affiliation, by ethnic origin, by
professional or work activity" occured? PR by political party is the best way
that PR
per issues of political activist of all kinds can be obtained!

> In these circumstances, it seems pointless to me to pursue some
> concept of "pure PR" to the ultimate arithmetic limit.  Even if
> you could achieve this goal, you would have PR only in one
> dimension.  Voters' responses are multi-dimensional - or at least
> they are when the voters are given the chance to show that.

This is why several (sub)districts that allow different debates are useful.
SPPA is designed to obtain PR representation in multiple-dimensions
as STV does too.

> > > Stephane wrote (> > > ):
> > > > I do agree, but you need to prove that the adverse
> > effects are not ghosts...
> > > >
> > > > I do not think that the advantages of a "reasonable
> district size"
> > > > outcomes obviously its disadvantages when compared to pure
> PR
> > > > systems.
> > >
> > > You need to define "pure PR".  What is "pure"?  On what
> > criteria?  In
> > > whose view?  You appear to ignore completely all the previous
>
> > > discussion in this thread and in related threads where it has
> been
> > > suggested there is more to "PR" than "PR of registered
> political
> > > parties".
> >
> > I do agree with you to the fact that there is more to only
> > political parties. I would not want people voting for ideas
> > without knowing who is going to implement those ideas and
> > how. I think independents are an asset as a guarantee that a
> > person can run for office without being linked to some
> > oligarchy. Both this element are present in SPPA as voters do
> > not vote for parties but for persons (as with STV) and
> > independents can run (as with STV). So please precise your
> > thoughts... What obstacle is so unavoidable that we have to
> > abandon PR systems fully proportional, pure PR systems using
> > my vocabulary, and not only PR list systems?
>
> But unless I have misunderstood your descriptions of SPPA, votes
> are aggregated within parties and seats are allocated nationally
> to parties. That makes party PR the objective and, de facto,
> everything else will be secondary to that.

Yes PR of parties is the first objective (chronologically during the
resolution)
but it does not make individual support unimportant at all!  Secondary
is not negligeable and can be nearly important at primarly. SPPA is based
on individual votes exactly like STV. It is only the mechanics that are
changed
(using polling and sampling-like strategies) to obtain the result equivalent
of
one huge STV Greater district without voting on a ballot with a thousand of
names.

> > > > Alex says that "a reasonable district size will also keep
> the
> > > > representatives closer to those they represent, and will
> keep the
> > > > focus in elections more on the individual candidates
> because the
> > > > field of candidates won't be as crowded."
> > > >
> > > > Geographical "closerness" is a bad thing.
> > >
> > > This may be your opinion.  It may even be an acceptable
> > opinion to a
> > > majority of the citizens of your country.  But all I can
> > tell you from
> > > more than 40 years of practical campaigning for electoral
> reform in
> > > the UK is that "closeness" matters to electors here.
> >
> > I do not have your experience.
>
> I do not know what you mean by this response.  If you are
> referring to my 40 years of campaigning, my personal involvement
> is irrelevant.  What matters is what electors say they want.  UK
> electors value "closeness".  My opinion may be reinforced by the
> fact that I have never heard electors say anything different (and
> I have been listening for quite some time).  I suspect it is the
> same in many other countries, especially those with a legacy from
> the British parliamentary system, but I would not presume to make
> such assertions without direct local evidence.

I would never want to impose a non-geographical electoral system
to any electorate who does not want it. But I think even you James
who take the time to read me (thank you by the way) do not see
what you would keep and what would be changed by such a modification.
I am sure that what you value as "closeness" would still be available.

> > > > Yes in one hand it gives
> > > > an elected representative that knows the district better
> than
> > > > others. But on the other hand, it attracts several
> dysfunctional
> > > > behaviors. It attracts lobbyist not legislators.
> > >
> > > Where is the evidence for this?  The worst lobbyists (or the
> best,
> > > depending on your point of view) are not those who represent
> the
> > > people who live in their locality, but those whose promote
> single
> > > issues.
> >
> > You have a local council to defend your towns problem. You do
> > not need a local
> > representative when it comes to make national laws or budget.
>
> I find this statement amazing, especially from a citizen of
> Canada.  I know what you mean when you say that you do not want
> members of the federal parliament to make decisions purely on the
> basis of local (geographical) interest.  But surely there is a
> legitimate state or regional perspective when deciding the
> federal budget or making national laws?  There certainly is here
> in the UK.

The problem is that you have to rely on MP's ethics. Even scandals journalists

cannot put public some "localization" scandals because it is always possible
to find some arguments defending any choice. Why accept such a situation when
the
system could provide a garantee by removing any incentive?

<snip>

> > > > It attracts people who want to get
> > > > the best to their local community,
> > >
> > > That surely is a laudable aim?
> >
> > by itself yes, but it depends how...
> >
> > > > at the detriment of the country or
> > > > other districts nearby if necessary.
> > >
> > > This is not a necessary consequence.  It suggests to me that
> the
> > > structures and functions of government at federal, state,
> > region and
> > > town/city level are not adequately defined or separated.
> >
> > No. It is because representative at each of these levels
> > defend first the interest of the part that elected them and
> > next the whole entity. And it is normal human behavior as
> > long as the system retributes such behavior by reelecting
> > those who do so.
>
> But this is a representative democracy.  If that is what the
> voters want, is it really for you or me to say they should not
> get it?  Whether you or I agree with such an outcome is
> irrelevant - that is democracy at work.  If we don't like what
> our fellow voters want, we can, of course, seek to educate them
> to "better ways" or campaign for those who take a broader view -
> but that is quite a different matter.

Voters want the most representative, least corruptible, most efficient
government.
I do not want to sacrifice anything to propose a less corruptible governement.

SPPA just proposes a pareto-optimal modification of the electoral system to
obtain
a government more representative, less corruptible and as efficient.
At least I think so...
More representative using PR definition, less corruptible in regards to
localization
decisions, as efficient in terms of stable majority or coalition.

> > > > It institutionalizes ghettos,
> > >
> > > I cannot see that this follows at all.
> >
> > When laws affecting in a specific way all rural persons in
> > brittain are taken,
> > who protects the interests of that minority. The rural
> > districts representatives. I do not know from which political
> > party they are, but I can tell you they are minoritarian in
> > the english parliament because the system of districts
> > assigns a finished number of seat to those localities.
> > Minoritarian they are and minoritarian they will be after any
> > election, even if they were the target of a major unfair law
> > that would become the main subject of debates.
>
> I'm afraid I cannot follow your argument here.  The rural
> electors in the UK (or in any part of the UK, like England) are
> in the minority, but not because of any scheme of districts.
> They are in the minority compared to the urban electors because
> there are fewer of them.  No scheme of districts is going to
> change that.  Of course, you could "balance" the representation
> by having very different numbers of electors per elected member
> in the urban and rural districts, but such extreme manipulation
> would not be acceptable (not since 1832).
>
> That said, we do have a "rural bias" in the districting in most
> parts of the UK, even when we use a system of PR, as for the
> Scottish Parliament.  It is both accepted and expected.  In terms
> of practical politics, there would not have been a majority vote
> in favour of (re-)establishing the Scottish Parliament unless the
> proposed electoral system would ensure over-representation of the
> rural and remote areas of Scotland.  Of course, that distorts the
> arithmetical equalness that some say is implicit in "one person,
> one vote", but in the real world there is more to equality of
> representation than arithmetical equalness.

Assume some percentage of the population can figure out that resolving
unfairnesses makes them lose something to their neighbour but improves the
global result. So resolving several unfairnesses profits to most of the
population
and at worst does not change the quality level of others life.

A cake analogy would be two groups wanting more than their fair share of the
cake.
FPTP uses an advocate that gets paid using some part of the cake while
the unfairnesses the advocates puts in place justifies giving each group its
fair share
of the remaining piece of cake. SPPA makes it easier to identify the
unfairnesses
and eliminate them thus everybody would get their fair share of the initial
cake
getting rid of the advocate. One way of doing it, is not to put the power of
distributing
the extras in the hand of someone that knows which side is piece will be taken
from.

> > To be elected
> > somewhere with geographical districts, you need to match the
> > positions of this
> > district interest groups.  So if any issue can advantage a
> > majority of these districts, forget defending the minority
> > point of view about it, how unfair it is. National
> > representative can fight such unfairness by using the
> > minorities to make the difference, because they only need a
> > small margin to beat other candidates of the same party and
> > get elected.  FPTP or STV can't (except for huge STV
> > constituencies in number of seats)
> > because minority defenders need to beat the majority
> > defenders: in the best case they can be elected in some
> > districts and shout, in the worse case they lose everywhere.
> > With SPPA if a particular unfairness rises in all the
> > election, a majority of candidates wanting to repair it can
> > be elected because it is the electorate that chooses the
> > debates that matter.
> >
> > > >purshasable votes for the next metro station or the next
> plant
> > > >subvention.
> > >
> > > At worst, that COULD apply in any political system, but it is
> a
> > > caricature because rarely will any elected member have such
> > > power or influence.
> >
> > No!! It is not the case with SPPA. Again I ask you as I did
> > Alex: where in England should a politician put a metro
> > station to favour the voters from the virtual district made
> > of the persons born between January 6th to January 10th of
> > any year? Ministers do have to take such decisions, you know
> > that very well.
>
> There is no answer to your question.  Of course ministers do have
> take such decisions and they are expected to consider competing
> priorities.  If they are blatantly partisan in their decisions,
> they will be called to account by parliament, by the media, and
> ultimately by the voters.

And because of the system, the voters that make the call are the ones who
got the metro station and elect back the politician in place...

> > > > I agree, it
> > > > worked like that for decades everywhere.
> > >
> > > I don't agree at all.
> > >
> > > > But if a system can offer
> > > > elections using principles and not interests, ideas not
> fashions,
> > > > shouldn't you consider this more closely.
> > >
> > > Yes, of course.  But in practice, most electors look to
> > their elected
> > > representatives to fulfil a number of different functions
> > > simultaneously.  Among these are (very) local interests and
> > national
> > > interests and international interests, as well as interests
> of many
> > > different kinds on political issues, social issues and
> > ethical issues.
> >
> > So use a representative from the institution that fits, using
> > districts that fits. For international issues, use
> > international committees with internationally elected people
> > who care as much about the health of english people than
> > tanzanians.
>
> This is pie in the sky.  For the foreseeable future, all
> international decisions of any real importance will be taken by
> national governments.  International bodies will be allowed to
> take only the decisions that don't really matter.  Of course, the
> realist might go further and say that what the USA wants, the USA
> will get!

The darn planet is physically one country now. What we say, think, show, write

buy, pay, sell or whatever can go around the world seven times in a second
so what about some realism and admit we need one world parliament
before losers of the actual systems decide to show how unfair it is in their
eye to the others.Your logic leads us to far more dangerous things than pies
in the sky.
Such a government can't exist using geographical basis but it is the necessary
step
so people feel represented as part of the world, not as Canadian, English,
Chinese
or French people.

<snip>

> > However, your paralysis argument is
> > unfounded when compared>
> > to the "crutch" option described to fix it in SPPA. I am a
> > serious electoral-reformer, so I would solve a problem
> > (misrepresentation) by creating another (unstability). Please
> > read before saying it is impossible. SPPA guarantees a stable
> > government that could produce bipartite coalitions for a
> > reduced mandate at worst. So please do not remove small
> > parties without necessity.
>
> This makes a complete nonsense of your striving for arithmetical
> exactitude in the PR of parties.  Your "fix" appears to say we
> shall accept the results of "pure PR", but only if one party wins
> a majority of the seats so that it can form a stable,
> single-party government.  Otherwise we shall distort the "pure
> PR" to manufacture a single-party majority.  If you believe in
> this definition of "strong government", why not build that into
> the voting system?  This has been done before: according one
> professor of politics, in Mexico the party that won the most
> votes was given 55% of the seats; in the 1920s Mussolini
> introduced a new electoral law that gave two-thirds of the seats
> to the party that won the most votes.  (In the event, the
> Fascists did not need this distortion to secure power  -  they
> won two-thirds of the votes!)

No, it is perfectly logical. All seats distribution other than Hare
(D'Hondt, St-Lague, Droop, if I am correct) are motivated by
a necessity for stable governments (actually Droop is more for
a strategical voting issue) thus a prime of seats to the winning party.
The divisor technic with odd quotients is a classical example.
I acknowledge and recognize the need for a stable governement.
However there is no need to give bonus seats to the
most popular parties if the resulting government is already stable.
My own analysis showed that in general bipartite coalition were
stable and three parties or more unstable. Typically, two parties almost
never disagree on every issue, while the intersection of three different
programs can be null or at least very thin. So why not correct the problem
only when it rises. If after the election, we see a stability problem, then
and only then exist a motivation to boost the most popular party. If not,
maximized PR seems the most representative stable government.
Reducing the mandate in exchange for additional seats seemed a good way
to maintain a fair representation, at least in the (time x representatives)
space.
Adding a dimension to create an invariant is very common in math & physics...

<snip>

Stephane