Ziu Reform Possibilities

[Baron Alexandreu Davinescu]

If possible, I'd like to open up some discussion on Ziu reform.  This has been a topic of perennial interest to many, and I think it might be smart to have a good-faith discussion about some possibilities that might be useful for us.  We just recently switched to a 20-seat Cosa instead of a 200-seat Cosa, but that doesn't fundamentally change that much.

A few things from my perspective, to begin:

• We should probably err on the side of preserving things like the Senats, as an important institution that has often been important to slowing down the pace of very significant legislation (without ever actually becoming an undemocratic block on it forever) and as a provincial representative.  So at least for me, I wouldn't want to make changes that would just be an obvious set-up to eliminating the Senats. 
• There are numerous good things we should try to achieve: proportionality, direct accountability, low barriers to entry, simplicity, intuitiveness, and ease of implementation.  Not all of these are equally important, but we should be honest about trade-offs.
• While we're a highly educated people, with 71% of our citizens holding some form of college degree, we have a majority of American-Talossans, and so most of our electorate is going to be familiar with our current system and then their own American one.  In recent elections, many people failed to use RCV even when directly and repeatedly instructed to do so.  Simplicity is going to be key!

Some people have been talking about Mixed-Member Proportional systems.  As best I understand it, this entails people voting for a candidate for their area and also for a party.  Each candidate wins based on the local vote, and then a big chunk of seats are divided up according to the national party vote.  Sometimes it's always the same number of seats, and sometimes it's a fluid number that varies depending on how many additional seats are needed to make the results very proportional.  There's also a version where the plurality party gets a "majority bonus" of extra seats.

So what are some ideas people might have about this system, or a different system that might work better?

[Miestră Schivă, UrN-GC]

My country has used Mixed Member Proportional for more than 30 years and I think it works pretty well, ask me anything.

The simplest way to work it in Talossa would be to have one Cosa seat elected by each Province (8 seats) then the other 12 seats "topping up" to proportionality. So: say a party wins 2 provinces, but 40% of the vote. 40% of the vote entitles you to 8 seats. So: they also get 6 seats from the top of their party list.

There is a quirk called "overhang" which happens if a party gets *more* local seats than their party vote would justify - for example, if an Independent who didn't have a party won a province. Then that local seat is ignored when the party seats are handed out; which means there end up being "extra" party seats, in our case, more than 20. Not a big problem IMHO.

The only real issue I see here is that if you want to keep the Senäts as is, then you have each province electing an MC *and* a Senator, which seems otiose and redundant. The clear compromise here is: elect the Senäts differently. The best alternative I've hard so far is Lüc's suggestion of electing 4 Senators from the whole Kingdom at large every term.

[Marcel Eðo Pairescu Tafial, UrGP]

My country pioneered MMP in 1949, ask me anything as well.

As much as I think the system has served Germany well, it is a bit difficult to understand at times, especially with regards to overhang seats and independents.

For instance, in the German way, independents who win a local seat do not constitute an overhang, and instead reduces the total number of seats taken in account for top-up purposes by one, i.e. if an independent wins their district in a 20-seat Cosă, the total number of seats stays at 20. However, because independents don't affect the party-based distribution of the remaining seats, people who voted for a winning independent have their party vote ignored, lest they have double the influence on the result as other voters. This step is necessary because otherwise, local candidates could all choose to run as independents while the party vote stays unaffected, transforming Mixed-Member Proportional into Mixed-Member Majoritarian à la Japan.

[Miestră Schivă, UrN-GC]

Yeah; while I like the MMP system there are simpler forms of personalised proportional representation. But, any Ziu reform which means actual control by voters over who sits in the Cosa would be great.

[Breneir Tzaracomprada]

I know that we keep citizenship eligibility at 14 for historical reasons but we should consider a minimum age limit for being an MZ. I would suggest 18 or 21 (if this does not run afoul of the OrgLaw). If we are serious about making  structural improvements to the Ziu then it might be worth considering whether 14 or 16 is generally just too young to effectively participate in national decision-making.

ADDED: I mean no offense to our youngest citizens but wanted to broach the topic.

[Mic’haglh Autófil, O.Be]

As someone who has advocated for MMP before, I'm certainly glad to see folks starting to soften to the idea. Even just discussing it is a good start.

I do agree that maintaining some sort of provincial representation is worth doing, but that having provincial representation in both houses goes too far. If we wanted to move the Cosa to an MMP setup, then I would agree that implementing something like Sir Luc's idea would be a good change for the Senats.

Overhang seats can of course be a concern, but as Miestra and Sir Marcel have both already pointed out, there are methods to reduce these. Though I would prefer a fixed size to the Cosa, I don't mind the method where seats are added to negate overhang seats. Especially if we take the German method of excluding party list votes for successful independent voters, that likely minimizes the number of overhang seats anyway, so one every few elections isn't objectionable.

One thing I don't really support is the majority bonus or majority jackpot system; yes, not using them makes government formation and stability a bit harder, but that just means you might have to actually negotiate things on occasion. C'e la vidă.

One of the nice things about MMP, in my opinion, is that it seems to be fairly flexible in the details of how you choose to implement it. As an example, there is nothing preventing us from still electing the party seats via an open-list method, or from using ranked-choice voting for provincial seats.

[Baron Alexandreu Davinescu]

Okay, so it seems like there's a lot of offers of help with figuring this out.  I think I have a sense of my caucus, and I think I can say that there would be flat-out no real support for anything that gets rid of the Senats unless it was demonstrably better in very specific and practical ways.  My strong suggestion is that we focus on versions of this that would work with the Senats, not eliminate it, if we want to come up with something that has broad support.

So looking at the nuts and bolts here, we could:

• Elect a senator from each province at the current interval.
• Elect an MC from each province every election.
• Choose a party as well.

It doesn't seem like this would interfere with RCV for the Senats at all, and it seems like we'd have enough people for this, right?

Miestra has proposed that we'd add however many seats necessary to ensure that the results would be proportional to the vote.  But that's a question of degree, right?  Like... let's say that Party A won 50%, Party B won 25%, and Party C won 25%.  In a 20-seat Cosa, that's 10 seats for A, 5 seats for B, and 5 seats for C.  But if B won six of the provincial seats, then they'd have 30% (more than their share of the national vote) and so A and B would need extra until things were proportional.  How does that look?  How close would things need to be?

[Miestră Schivă, UrN-GC]

let's say that Party A won 50%, Party B won 25%, and Party C won 25%.  In a 20-seat Cosa, that's 10 seats for A, 5 seats for B, and 5 seats for C.  But if B won six of the provincial seats, then they'd have 30% (more than their share of the national vote) and so A and B would need extra until things were proportional.

You could do it that way, but that's the harder way. The simpler option would be to just live with the overhang. Let's say A and C both won 1 province. Then, Party A gets 9 party list seats (to sum up to 10) and Party C gets 4 seats (to sum up to 5). So that's a total Cosa of 21 seats. Party B gets a small bonus.

[Marcel Eðo Pairescu Tafial, UrGP]

let's say that Party A won 50%, Party B won 25%, and Party C won 25%.  In a 20-seat Cosa, that's 10 seats for A, 5 seats for B, and 5 seats for C.  But if B won six of the provincial seats, then they'd have 30% (more than their share of the national vote) and so A and B would need extra until things were proportional.

You could do it that way, but that's the harder way. The simpler option would be to just live with the overhang. Let's say A and C both won 1 province. Then, Party A gets 9 party list seats (to sum up to 10) and Party C gets 4 seats (to sum up to 5). So that's a total Cosa of 21 seats. Party B gets a small bonus.

Yeah, that's the original MMP system that is still in use in New Zealand. What AD mentioned are called compensatory seats and Germany moved to a system with compensatory seats because the Constitutional Court ruled that the old system was unconstitutionally disproportional (mainly Bavaria's fault), which in turn lead to ever bigger parliament sizes... the system we use now caps parliament size at 630, and abolishes overhang seats entirely: now, if a party wins more constituencies than their party vote would justify, the candidates that won by the narrowest margin simply don't get in, meaning some constitutuencies (IIRC again mainly Bavarian ones) not having a local representative.

[Mic’haglh Autófil, O.Be]

I'm of the opinion that the former of the two methods (the one Miestra describes) is better. As much as Germany's newer method fixes the size of the legislature, and while I did mention that I like that sort of thing on principle, I don't like the idea of a constituency being represented by someone who didn't really win it.

An easier way to describe that method might be to say that the size of the Cosa is increased by a number of seats equal to the total number of overhang seats won. In this example, Party B wins one overhang seat, and is the only party to do so; the Cosa is therefore 21 seats for the term.

I should note that using ranked-choice voting for the provincial seats instead of first-past-the-post would also reduce the odds of overhang seats occurring. A candidate that secures a majority of the vote in a province is more likely to come from a party that has a broader base of support nationwide.

I do notice that in his most recent post, the Baron appears to have glossed over the issue of double-representing provinces in the Ziu. While I would like to give him the benefit of the doubt and assume this was an oversight on his part, in the event it was intentional, may I reiterate for the benefit of him and his caucus that the method proposed by Sir Luc does not eliminate the Senats' existence or even alter its powers, simply the method by which it is elected.