How MMP Lists Are Created
Wednesday August 29th 2007, 11:21 am
Filed under: - Electoral Reform, Canadian Politics

Scott Tribe, who runs Progressive Bloggers, asked me how I thought the party lists should be created under the MMP system recommended by the Ontario Citizens Assembly. He has posted my response on his blog. I thought I would just copy that response here in its entirety (note that the # of regional list nominees below are minimums, parties will probably want to nominate a couple extras per region):

***

A critical component to the proposed MMP electoral reform system for Ontario are the List MPP’s. There has been much gnashing of teeth over whether these are “democratically accountable” MPP’s or not. I will not re-visit that argument today, but what I will be touching on today is what might be the best method (from a democratic viewpoint) that parties should endeavour to use when picking people for their various lists, because that was not set in stone by the Citizens Assembly. It was left to the parties to come up with their own method, with the provision being they had to submit their lists to Elections Ontario to a) show who they picked and what order they’re in and I believe b) how they picked them.

I again talked with Greg Morrow of DemocraticSPACE.com. I chatted with him a couple of days ago about this, and he discussed with me what his preferred method would be. He has a very detailed explanation that he goes into below, so without any further ado….

Greg believed there are many good options out there to choose from, The best way, in his opinion, is to hold REGIONAL NOMINATION MEETINGS. The key is for those regions to be big enough that enough women and minorities are nominated (according to Greg, a minimum of 5 list spots would be a good number), but small enough that the candidates/regions represent a genuine community of interest.

In Greg’s setup, Party members at regional nomination meetings can decide whether they want to include candidates nominated locally on their list or not (in some cases, where it is clear a candidate will win locally, it is a waste to dual-list him/her; in other cases, it might be desirable to dual-list a candidate who faces a tough local battle; the party could set some basic parameters — for example, at least 1-2 list candidates per region must be not dual-listed, to ensure there are enough list members in case all local members are elected, or 40% of list candidates must be women, or that there be a representative number of minority candidates — i.e. roughly equal to their share of the regional population, etc).

In Greg’s opinion, allowing locally-nominated candidates to be eligible for list spots is important to generate competition within the parties (as people say: bad for politicians, good for voters), which increases the accountability of local MPPs. But, if the list is entirely made up of local candidates, you close off opportunities for women, minorities, or just candidates who bring some special skills to the table, but who are don’t have the money to run a local campaign. According to Greg, the key is to put the selection of list candidates in the hands of the regions, rather than the central party executive. If so, it’s not dissimilar from local nominations today.

However, Greg believes this process also gives the party some flexibility in deciding how order its list candidates — the regions nominate the list candidates, but the party can then decide whether to “zipper” its list — i.e. alternate man-woman-man-woman-etc, or whether to put candidates from regions where it thinks it will not be as successful locally (so as the ensure a caucus that is regionally balanced), or to ensure that enough of its top spots are minorities. It’s the best of both worlds — candidates are nominated regionally (thus accountable to the membership in a given region) but the party can be strategic in how it orders those nominated candidates.

Greg says there are different ways that you can think of the regions — you can have fewer larger regions or more smaller regions. It’s a trade-off — the smaller the regions, the better they represent regional interests, but the worse it is for women and minorities (since there aren’t enough list spots — the literature generally recognizes you need at least 5 list spots in order for women/minorities to get a fair shot). On the other hand, fewer regions means that they can all have at least 5 list spots.

Greg then sent me some very detailed options of how you could do the regional list nominations – he included 3 possible examples:

A) 6 larger regions — best for women/minorities (since all regions have 5+ list spots)
B) 9 medium-sized regions — a balance of list spots and geography
C) 11 smaller regions — better regional representation, but at expense of women/minorities (since small # of list spots)

Below is what each example might look like in Greg’s scenario:
***

Option A: 6 larger regions

1. NORTHERN ONTARIO
(Kenora, Rainy River, Thunder Bay, Cochrane, Algoma, Manitoulin, Sudbury, Greater Sudbury, Timiskaming, Nipissing, Parry Sound, Muskoka)
estimated 9 local ridings
nominate 4 list candidates

2. EASTERN ONTARIO
(Prince Edwards, Hastings, Lennox & Addington, Frontenac, Renfrew, Lanark, Leeds-Grenville, Stormont-Dundas-Glengarry, Prescott-Russell, Ottawa)
estimated 13 local ridings
nominate 6 list candidates

3. CENTRAL ONTARIO
(Simcoe, York, Durham, Haliburton, Kawartha Lakes, Peterborough, Northumberland)
estimated 16 local ridings
nominate 7 list candidates

4. SOUTHWESTERN ONTARIO
(Essex, Chatham-Kent, Lambton, Elgin, Middlesex, Oxford, Perth, Wellington, Waterloo, Huron, Bruce, Grey, Dufferin)
estimated 16 local ridings
nominate 7 list candidates

5. HORSESHOE
(Hamilton, Haldimand, Norfolk, Brant, Niagara, Peel, Halton)
estimated 18 local ridings
nominate 8 list candidates

6. TORONTO
(Toronto)
estimated 18 local ridings
nominate 8 list candidates

***

Option B: 9 medium-sized regions

1. NORTHERN ONTARIO
(Kenora, Rainy River, Thunder Bay, Cochrane, Algoma, Manitoulin, Sudbury, Greater Sudbury, Timiskaming, Nipissing, Parry Sound, Muskoka)
estimated 9 local ridings
nominate 4 list candidates

2. OTTAWA-EAST
(Renfrew, Lanark, Stormont-Dundas-Glengarry, Prescott-Russell, Ottawa)
estimated 8 local ridings
nominate 4 list candidates

3. QUINTE-LIMESTONE-KAWARTHA
(Prince Edwards, Hastings, Lennox & Addington, Frontenac, Leeds-Grenville, Northumberland, Peterborough, Kawartha Lakes)
estimated 6 local ridings
nominate 3 list candidates

4. SIMCOE-UPPER GRAND-HURON
(Perth, Wellington, Waterloo, Huron, Bruce, Grey, Dufferin, Simcoe)
estimated 10 local ridings
nominate 5 list candidates

5. SOUTHWESTERN ONTARIO
(Essex, Chatham-Kent, Lambton, Elgin, Middlesex, Oxford)
estimated 9 local ridings
nominate 4 list candidates

6. HAMILTON-NIAGARA
(Hamilton, Haldimand, Norfolk, Brant, Niagara)
estimated 8-9 local ridings
nominate 4 list candidates

7. YORK-DURHAM
(York, Durham)
estimated 9 local ridings
nominate 4 list candidates

8. PEEL-HALTON
(Peel, Halton)
estimated 10 local ridings
nominate 5 list candidates

9. TORONTO
(Toronto)
estimated 18 local ridings
nominate 8 list candidates

***

Option C: 11 smaller regions

1. NORTHWEST ONTARIO
(Kenora, Rainy River, Thunder Bay)
estimated 2-3 local ridings
nominate 1-2 list candidates

2. NORTHEAST ONTARIO
(Cochrane, Algoma, Manitoulin, Sudbury, Greater Sudbury, Timiskaming, Nipissing, Parry Sound, Muskoka)
estimated 6-7 local ridings
nominate 2-3 list candidates

3. OTTAWA
(Ottawa)
estimated 6 local ridings
nominate 3 list candidates

4. EASTERN ONTARIO
(Prince Edwards, Hastings, Lennox & Addington, Frontenac, Renfrew, Lanark, Leeds-Grenville, Stormont-Dundas-Glengarry, Prescott-Russell)
estimated 7 local ridings
nominate 3 list candidates

5. CENTRAL ONTARIO
(Simcoe, Haliburton, Kawartha Lakes, Peterborough, Northumberland)
estimated 7 local ridings
nominate 3 list candidates

6. HAMILTON-NIAGARA
(Hamilton, Haldimand, Norfolk, Brant, Niagara)
estimated 8-9 local ridings
nominate 4 list candidates

7. UPPER GRAND
(Wellington, Waterloo, Dufferin)
estimated 6 local ridings
nominate 3 list candidates

8. SOUTHWESTERN ONTARIO
(Essex, Chatham-Kent, Lambton, Elgin, Middlesex, Oxford, Perth, Huron, Bruce, Grey)
estimated 10-11 local ridings
nominate 5 list candidates

9. YORK-DURHAM
(York, Durham)
estimated 9 local ridings
nominate 4 list candidates

10. PEEL-HALTON
(Peel, Halton)
estimated 10 local ridings
nominate 5 list candidates

11. TORONTO
(Toronto)
estimated 18 local ridings
nominate 8 list candidates

——

There’s a lot there, but its important to show that there are models out there that can be used to ensure the candidates for the lists are picked in a democratic manner similar to what we have now in our different parties nomination meetings. One other thing Greg mentioned is this: The Labour Party in New Zealand (which Greg would argue is the Liberal Party equivalent) creates its list in exactly the manner as he described above — list candidates are nominated democratically at regional conventions and the party then orders the nominees in order to meet its goals (for example, to compensate for weakness locally in certain areas, they put candidates from areas where they are traditionally weak near the top of their list). But this is done fairly — balancing men, women, and minorities (Maori) and across all regions.

As Greg says and as I’ve said before as well, there is no need to reinvent the wheel here. Being able to demonstrate exactly how it works in the same MMP system that we’re considering is the best bet of getting buy-in from the voters of Ontario. I also think that if this system was pushed for by the Party grassroots, the ongoing fear of the central executive manipulating these lists would be relieved by the implementation of this system.

Some will argue there’s no guarantee the various parties will pick this. Well, its up to the various party activists and grassroots to push for this in their respective parties. I again will argue that if a party is seen by voters to be manipulating the lists as opposed to others who are using the above or similar methods to pick their list, the offending party will almost certainly get hammered on the hustings over it, and will probably pay a pretty heavy electoral price. Between that, and with other examples out there of democratic ways to pick these list MPP’s, I think the pressure to do so here as well would be pretty intense. I dont think the various parties would have any choice but to do so. The Party Executives may try to argue that they need to retain power of appointment in certain ridings (as they do already), but again, a determined Party membership will either limit that power or prevent it from happenning.

This “fear” is the least of the arguments against MMP, in my view.

Update:
As Wilf suggested below a variation on Option C would be to put Perth, Huron, Bruce, Grey with Waterloo, Wellington and Dufferin. Let’s call this Option D.

Option D: 11 smaller regions

1. NORTHWEST ONTARIO
(Kenora, Rainy River, Thunder Bay)
estimated 2-3 local ridings
nominate 1-2 list candidates

2. NORTHEAST ONTARIO
(Cochrane, Algoma, Manitoulin, Sudbury, Greater Sudbury, Timiskaming, Nipissing, Parry Sound, Muskoka)
estimated 6-7 local ridings
nominate 2-3 list candidates

3. OTTAWA
(Ottawa)
estimated 6 local ridings
nominate 3 list candidates

4. EASTERN ONTARIO
(Prince Edwards, Hastings, Lennox & Addington, Frontenac, Renfrew, Lanark, Leeds-Grenville, Stormont-Dundas-Glengarry, Prescott-Russell)
estimated 7 local ridings
nominate 3 list candidates

5. CENTRAL EAST
(Simcoe, Haliburton, Kawartha Lakes, Peterborough, Northumberland)
estimated 7 local ridings
nominate 3 list candidates

6. CENTRAL WEST
(Wellington, Waterloo, Dufferin, Perth, Huron, Bruce, Grey)
estimated 8-9 local ridings
nominate 4 list candidates

7. SOUTHWESTERN ONTARIO
(Essex, Chatham-Kent, Lambton, Elgin, Middlesex, Oxford)
estimated 9 local ridings
nominate 4 list candidates

8. HAMILTON-NIAGARA
(Hamilton, Haldimand, Norfolk, Brant, Niagara)
estimated 8-9 local ridings
nominate 4 list candidates

9. YORK-DURHAM
(York, Durham)
estimated 9 local ridings
nominate 4 list candidates

10. PEEL-HALTON
(Peel, Halton)
estimated 10 local ridings
nominate 5 list candidates

11. TORONTO
(Toronto)
estimated 18 local ridings
nominate 8 list candidates

8 Comments

8 Comments/commentaires
Leave a comment/Enregistrer un commentaire

As a member of the NDP, I believe as much control as possible should be with individual members, rather then the party elite.

The first person on the list would be the leader, who would be elected by OMOV.

The rest of the list would be fill in during a campaign conducted one year before the election is expected, where each member would vote one person. The list would be ordered by the number of votes recived.

Members could vote for multiple list candidates, but that would encourage slates (where people would be given lists of candidates by factions in the party, and the most popular faction would control the whole list. (However major party factions will likely form thier own parties once MMP is implemented)).

For those who want gender parody members could vote one female and one male list candidate and the list would alternate between genders.

Comment/commentaire by Darwin O'Connor 08.29.07 @ 12:49 pm

While this answer is quite thorough, there is still a flaw with the regional representation here. That is simply it is left up to the parties to decide if regions get those MMP seats, it’s not legislated that those seats are given to specific regions. So the representation of smaller, lower-populated regions is left up to the whim of the parties. Under that set up, the historical trend of ignoring certain regions would be more likely to continue because there would be less of a punishment for doing it.

If these MMP seats were formally attatched to these regions, regardless of the set-up, I would more than likely be able to get behind it. But sadly, this proposal is not set up that way.

Comment/commentaire by Cam Holmstrom 08.29.07 @ 2:40 pm

Cam — yes, there are certain advantages of having proportionality calculated, and thus, list representatives elected at the regional level as opposed to province-wide. But it’s a trade-off. If we had regional MMP, results are less proportional, and women and minorities face similar barriers to entry, because the regional lists are too small.

Even still, parties aren’t stupid. If they don’t have balanced regions, then they risk being pegged as either pro-one region or anti-other regions (because their regional bias is made explicit with their lists). That is going to impact their ability to not only win party votes, but local seats too.

Parties could weigh the advantages and disadvantages for playing one region against another and decide it’s in their interest to do so. But its unlikely that any of the 3 major parties would do so, because they risk alienating whole regions, and relegate themselves to the status of a regional party. Only a small party looking to establish a base of support would dare risk blatantly favouring one region over others.

Comment/commentaire by democraticspace 08.29.07 @ 3:26 pm

“Parties could weigh the advantages and disadvantages for playing one region against another and decide it’s in their interest to do so. But its unlikely that any of the 3 major parties would do so, because they risk alienating whole regions, and relegate themselves to the status of a regional party.”

I have to take issue with this because this is something that already happens in Ontario. Ontario has a population that is far from being spread out evenly, so it’s not like most of the regions have the same kind of voting power. Basically, most of the provinces voting power lies in and around the GTA, while the North and rural areas have very little power. So there is not really a risk of alienating those voters in smaller areas because they simply do not need them to win as it stands right now. That will only get worse with this version of MMP, because all of a sudden those 39 seats are determined province wide. So the voters in those areas that have the most power already get another vote, and there are many more of their votes than there are in those lower populated regions. That means that there is a whole new pool of seats to go after that have no consideration given to regions, therefore the larger number of votes in those highly populated areas will swamp those from the lower populated areas.

So, if you’re really good to that highly populated area, you stand to get 2 votes, and there are so many more people to give you those votes in those areas. This will only make this that much worse.

As for the smaller party argument, this one just does not wash for me. I’ve been told that areas, like Northern Ontario, could form their own parties, and go for that 3% number. Well, problem there is that the Northern Ontario ridings only makes up 5% of the total votes. So that party would have win 60% of the votes in those Northern ridings to reach 3%. If they could do that, they’d be winning riding seats and would get any MMP seats. In fact, it could be argued that because there are fewer ridings seats thanks to MMP, that this small party would be hurt by MMP.

I would love to think that the parties would “dare risk blatantly faviouring one region over others”, but look at the political history of this province and country, and you see tonnes of examples of parties doing just that.

Comment/commentaire by Cam Holmstrom 08.29.07 @ 3:48 pm

The question is Cam… does the proposed new system of MMP appear to be better then the current one, despite the fears and nit-pickling of some folks?

I say if it does (and I believe it is) then we should stop trying to be perfect and go with the improvement.

Comment/commentaire by Scott Tribe 08.29.07 @ 8:41 pm

Hi Cam, I think I’ve talked with you about this on a number of differene venues.

Proporational systems are better for representation of any sort of smaller group of people, whether it be geographic, or idealogical. Thus as the environmental vote can go to the Greens if the large parties give that sentiment short shrift so too can the Rural Vote. Plus PR creates legislatures that encourage cooperation so even smaller parties can be productive and valuable members of government. This is unlike FPTP where the winner take all mentality is ingrained.

The political history of Ontario and Canada is one of FPTP where vote rich swing ridings and false majorities are what matter. This would not be so in MMP.

Could you tell me what the difference would be between a possible Northern Ontario Party and a Rural Ontario Party?

Comment/commentaire by AamirHussain 08.31.07 @ 2:15 pm

I have no doubt most parties would use regional conventions to nominate list candidates. Certainly the Citizens’ Assembly thought this was a reasonable approach. They said:

“In New Zealand, the Labour Party and the National Party determine their lists at regional conventions. The lists are then assembled by a special national committee of each party.

“The more common practice in MMP systems is for list candidates to run locally as well. In the 2002 New Zealand election, 84% of list members also ran locally. This gives these candidates more visibility and strong connections to particular areas or regions.

“Permitting dual candidacy recognizes that there can be only one winner in local ridings under a Single Member Plurality system. Candidates who have strong public support can lose local races. For example, in the 2003 Ontario election, the winning candidate in one district received 35.87% of the vote. In another district, a losing candidate received 45.16% of the vote. As this example shows, candidates who lose can actually have more support than other candidates who win.”

The Citizens’ Assembly considered requiring parties to nominate candidates in some particular way, but not many democratic countries have laws governing how parties nominate candidates: some parties are different from others, in a democracy. So there is no legal guarantee stopping a party putting rich white males from Toronto in the top 50 list positions. Just political reality.

As to your sample regions, the Ontario PC Party has nine regional vice-presidents from nine regions, but I can’t find their definition. I like your Option C, except I would ask Perth, Huron, Bruce, Grey whether they want to go with Upper Grand rather than the Southwest.

Comment/commentaire by Wilf Day 09.02.07 @ 10:12 pm

I honestly think your proposed regional schemes greatly underestimate the size of the lists, mainly because of dual candidacy. If I tallied them up correctly, your suggestions each tended to have 40-50 list spots.

Realistically, with dual candidacy allowed you can bet that the riding candidates will push hard to have that extra shot at winning a seat that comes with a list candidacy, and there’s no good reason to deny a supposedly qualified candidate at least some position on a list, even it it’s not a particularly good position. A party doesn’t pay extra to have a longer list, so why not include all 90 of your riding candidates, plus anyone that the party feels should be in a seat despite not winning a riding nomination?

(Granted, smaller parties that expect few or no riding seats probably won’t bother with any more than 39 list spots, but I’m assuming we’re talking about the big 3 here that can be expected to run full slates in the ridings.)

One real-world place to look to is Germany. Their lists are pretty long, and it’s hard to see why Ontario’s would be much different. I would speculate that an Ontario ‘full-slate’ list would run from 90 to the full 129.

On the regional balance question, if all (or nearly all) candidates are dual candidates, parties [i]could[/i] opt to promote an equitable distribution by using a fairly straightforward recipe for filling the list:

Step 1. Within each region, rank your riding candidates by likelihood of winning, i.e. best chance of winning goes first. (I know it won’t be perfect, but the honest best guess is probably good enough.)

Step 2. Fill your list alternating through the regions, picking the candidates in the order that you just established.

So if you had, say 9 regions, then spots 1-9 would be the best bets of each region, followed by each region’s 2nd best bet in spots 10-18, then each region’s 3rd best in spots 19-27, etc.

The party’s resulting seat distribution after the election should be as closely balanced as can be accomplished with whatever number of list seats the party wins. If the party is fairly realistic about who is most likely to win, then winning riding candidates should generally be those near the top of the list. Regions that are underrepresented in the ridings would therefore be the first to benefit from any list seats.

For smaller (list-only) parties it’s even easier to balance the regions – just alternate throughout the list.

So it can be done – it just takes internal willingness on the part of the parties. (Although personally I’m too cynical to believe that the bigger parties won’t just use the list for some thinly-disguised cronyism. Sorry, had to be said.)

Comment/commentaire by Jamie Deith 09.20.07 @ 9:52 pm

RSS feed for comments on this post. TrackBack URI


Leave a comment/Laisser un commentaire
E-mail address never displayed/Votre adresse email ne sera jamais publiee. HTML: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>

(required/requis)

(required/requis)



If your comment doesn't appear, it is because our automatic anti-spam software is blocking it. If so, just send us an email and we will post it for you.