Legal Beagle by Graeme Edgeler

What's the practical implication of not having an overhang? If we had none at the last election, would a list seat have been taken from the largest party or the last elected list MP?

What's the practical implication of not having an overhang? If we had none at the last election, would a list seat have been taken from the largest party or the last elected list MP?

The last elected MP. If the overhang was going to be one, then the Sainte-Laguë apportionment would only use the top 119 quotients, if two, then top 118 etc. This is already the system used if an independent, or a candidate from a party which did not run a list, wins a seat.

In 2011, National would have been down one seat.
2008 - National and the Greens each down one.
2005 - National down one.

Except in 1996, National has had the 120th quotient at every election.

While worthy all the changes you mention dont 'enhance the overall experience' as marketeers might say.

So that brings me to my main point, 'you may get what you want but lose what you had' effect. For an electoral system the tinkering is never done. Is this what we really want ? I notice from the Germans experience of MMP, which ours is most closely aligned, they have ended up by small changes with quite a different system than what they started with.

As for National having the 120th quotient at every election bar one, I would have thought that the maths would point to astronomical odds on that happening. Has David Farrar reverse engineered that result from the way they chose electorate MPs who go on the list as well.

While worthy all the changes you mention dont ‘enhance the overall experience’ as marketeers might say.

No. I argued for a much lower threshold as the only change likely to make very much of a (EDIT: in my opinion positive) difference at all (I don’t see anything higher than 3% leading to much appreciable change, but would welcome even lower).

Some of the more technical changes I recommend have the prospect to improve things in very rare circumstances, but all up I am not expecting something revolutionary.

Look at it as a lottery. (See here). The quotas are ranked in order, and each quota is like a lottery ball.

The more votes a party gets, the more balls. Therefore the chances of their "winning" the 120th quotient increases. National were largest party in 3/6 elections since MMP was introduced.

I could work out a formula for the probability, but won't, as I haven't got that much spare time.

I could work out a formula for the probability, but won't, as I haven't got that much spare time.

I'm guessing it's going to be close to 1/2^(number of elections). So getting a run of 5 is 1/2^5 = 1/32. Not that astronomical, but still unlikely. Interesting that it happened, but well within "these things happen".

ETA: Rough reasoning is that most of the time, National's got about a half in half chance of being the once benefiting from the rounding. So it's much like a string of coin flips.

ETA2: Hang on, the chance is more like 1/(number of parties)^(number of elections). Hmmm, yes, that's much more astronomical.

ETA3: Or something....I'm not really sure what the odds of being the 120th quotient is each time is. But how they stack up to a probability over a series of independent elections is pretty straightforward. Is their chance proportional to (their size)/(the total size of all parties)? If so, my first guess wasn't too far wrong.

I accept that being a larger party National ( or Labour) has more chances but apparently the Sante Lague method is chosen so as not to favour larger parties ( as opposed to D'Hondt method gives similar results too, but favours larger parties)
Here is actual quotients for 2002 when National got a very low 22% but still snagged the final seat.
http://www.electionresults.govt.nz/electionresults_2002/e9/html/e9_part2.html

We can see Labour had 9 quotients from 100+ while National had 5. The party to miss out in this election was United Future . Something like 25 votes making the difference

“I’m going to go along with everyone else here, and predict that the Commission will propose dropping the party vote threshold to 4%, and removing the one-seat exception. I don’t rule out its going to say 3%, but would be very surprised with anything lower than that.”

Pretty much agree. But I also think it may recommend that the best idea would be to abolish the electorate lifeboat and go down to 3% if voters agree to do so at a referendum – with a fall back of abolish the electorate lifeboat and go down to 4% if the government doesn’t want to have a referendum.

“The Commission will, however, recognise public concern about centralised control of party lists, which I believe it will seek to allay by proposing strengthening the requirement for internal party democracy. It will suggest that at the very least parties should hold internal votes of all (or close to all) their members, in the process of list-ranking (although the result won’t be binding).”

Oh God, I hope not. After all, Labour has just gone through a quite extensive internal review and reworking of its constitution/candidate selection process that would, under such a rule, be unlawful (or, at least, would have to be added to in order to comply with the law). All in order to produce a (for the moment … because you can be sure there’ll be future pressure to up the vote’s status) non-binding indication of whether a candidate should be listed at number 25 or 26 on the party list.

The tendency to favour larger parties refers to the number of allocated seats, not to who gets the last one. I'm hoping someone on here has stats students who could be given the problem of working out the probability.

Ive looked at each election since 1999 and gave probabilities based number of final quotients between 110 and 120 allocated to National.
Thus the result is 0.4 x 0.4 x 0.5 x 0.6 x0.6 = 0.036 , which is the chance of it happening in 5 independent elections.
Roughly the same as 5 heads in a row. ???

Roughly the same as 5 heads in a row. ???

In other words, you notice it when it happens, but it's not actually the biggest deal.

But I also think it may recommend that the best idea would be to abolish the electorate lifeboat and go down to 3% if voters agree to do so at a referendum – with a fall back of abolish the electorate lifeboat and go down to 4% if the government doesn’t want to have a referendum.

I suspect they’ll leave the fallbacks to the Government. I reckon if they think it should be 3% they’ll just say so, referendum or not.

And now I’m just trying to work out how this differs my from view that there may be contingent recommendations (e.g. remove the one-seat rule, but only if you lower the threshold), and I’m not sure there is :-)

Oh God, I hope not.

I am more confident of the former part (more party democracy) than the latter (the specific mechanism I mention).

And, of course, the Government and Parliament have to agree before any changes are made!

I suspect they’ll leave the fallbacks to the Government. I reckon if they think it should be 3% they’ll just say so, referendum or not.

Depends how the Commission handles its brief. If I were them (which I'm not ... but then neither is anyone else ... except them, of course) I'd be thinking on two tracks. One is, what do we think the "best" form of MMP would be, given all the submissions we've received? (Note that this isn't a counting opinions exercise - it involves evaluating the strength of the arguments made/evidence provided for the various policy options.) The second is, what changes can actually get made? (This is where the counting of heads matters - given the public sentiment as expressed through the submissions process, what ideas (good as they may be) simply don't have a chance of being accepted into law?)

Given these two-tracks, there may be a gap open up between the ideal and the possible in that the Commission believes there are good arguments for having a lower threshold than the bulk of submissions to the review supported (my impression flicking through these is that the majority favoured either no change or 4%). Hence the possibility of a report that says something like "there are very good reasons for having a threshold at 3%, but this would be such a significant change to MMP (and one that was not broadly supported in public submissions received) that it should only be adopted if the voters agree to it. However, the strength of the arguments for lowering the threshold from 5% combined with the support from submissions would justify Parliament moving to a 4% threshold without the need for further voter approval."

As you say, this may not differ from your view that there may be contingent recommendations ... it's just a particular prediction of such a recommendation.

And, of course, the Government and Parliament have to agree before any changes are made!

True. But I'm still not sure what "problem" the Commission would be addressing by recommending "more party democracy" ... aside from the public's view that they don't really like or trust political party "bosses". And, frankly, I'm not sure there's anything the law can do to fix that issue (without creating a range of other potential problems, including unleashing the lawyers on internal party faction fights). I just hope the Commission thinks the same way!

unleashing the lawyers on internal party faction fights

Well, as a Green supporter, can I be forgiven a bit of schadenfreude at that prospect. (I suspect it would afflict National, ACT and Labour. Peter Dunne is unlikely to be suing himself, and the Maori Party list is an irrelevance).

Hence the possibility of a report that says something like “there are very good reasons for having a threshold at 3%, but this would be such a significant change to MMP (and one that was not broadly supported in public submissions received) that it should only be adopted if the voters agree to it. However, the strength of the arguments for lowering the threshold from 5% combined with the support from submissions would justify Parliament moving to a 4% threshold without the need for further voter approval."

That’s another of the benefits of the proposal paper/final recommendations model: the proposal paper could come up with a suggestion and then see how people react to it, given the arguments put forward.

I’m still not sure what “problem” the Commission would be addressing by recommending “more party democracy” … aside from the public’s view that they don’t really like or trust political party “bosses”.

The perceived problem is: “list MPs are just there to do what they’re told by party bosses, and aren’t there to stand up for ordinary people. When push comes to shove, they’ll just do what they’re told and not make waves for fear of losing their list position.”. The better 'solution' to high party discipline may however be a vast increase in the size of the House.