A query,Greg.

I have studied your sample ballot for Northern Ontario. In Thunder Bay, should the voter vote for three regional representatives instead of one?

I notice that in your example for Thunder Bay, each local candidate is also represented as a regional candidate. Doesn’t this allow for parties to give each of their selected candidates two shots at a seat, and less of a chance for the popular will to override that of the party establishment?

eeid — no, people choose their preferred regional candidate. for parties earning regional seats, the candidate(s) with the most votes wins. i’m thinking that the parties would each offer 3 candidates in the North (since there are a maximum of 3 regional seats available). So, you decide which party you want to support, then pick which of the 3 candidates you like best. if a party wins 1 regional seat (on the basis of their total votes in the region), the person with the most votes wins. if a party wins 2 regional seats, the top two candidates are elected.

Jan — you are talking about what is called “dual-listing” (where candidates are on both local and regional ballots). Dual-listing is typical of all MMP systems. You asked “Doesn’t this allow for parties to give each of their selected candidates two shots at a seat, and less of a chance for the popular will to override that of the party establishment?” It’s a good question. Parties will have to decide whether they want to run all or none of their local candidates on the regional ballot. Obviously, if they run all their local candidates on the regional ballot then voters might think they are not being given sufficient choice. It could hurt the party’s chances.

The two ballots elect representatives at two different scales of community — one at the local level, another for the region as a whole. It may be possible for a candidate to be less popular locally than his/her local opponent, but yet be more popular across the region as a whole than the alternatives on the regional ballot. But since it is open-list, voters get to choose. It’s really up to them — if they like a candidate on the regional ballot more than another candidate, even if they lose locally, they are still preferred regionally. In my system, the decision rests with the voters.

But in the other MMP system — closed-list — I think the circumstance to which you refer is very problematic. The only thing we know is that voters rejected that candidate locally. If the lists are closed, and that person gets elected anyway because they are the top-ranked person on the list, I think there is a legitimacy question. This is not a problem in open-list, because voters have a choice whether to elect them on the regional ballot or not.

That’s why open-list is a more legitimate and more democratic system than closed-list.

Will a candidate who is elected both locally and regionally then be skipped over in the regional counting for the next-most preferred candidate of that party?

I see it as perhaps being more convenient to have primaries similar to US primary elections for regional lists than to have to do all of the selecting at one time. For example, the top three vote-getting candidates in a primary for the NDP in the North would then be the three NDP list candidates, with the highest vote-getter in the primary first, the middle vote-getter second, and the lowest vote-getter third.

Tom — here’s how I imagine it working. A party would hold a regional nomination meeting, which is, in effect, a primary for party members (all parties would do this for all regions). In the North, we know there are at most 3 regional spots available, so the party will want to nominate 3 candidates. So, in the regional nomination meeting/primary, party members vote for their preferred candidates. The top 3 vote-getters at the regional nomination are then the 3 regional candidates for that party (note they are all put before the general electorate as equals — there is no ordered list).

Then, at election time, voters choose between which of the 3 they prefer the most (basically, voters decide which party they prefer, then choose the candidate they like best of the 3 regional party candidates).

Let’s say that one of local candidates has also won nomination for the regional ballot (i.e. one of the three is dual-listed). Let’s also assume that this person wins their local seat. If this person received the most votes on the regional ballot, then the elected regional MPP will be the next-highest candidate (since the first candidate has already won their local seat).

FYI…
The Scottish Social Attitudes Survey, 2003, asked some questions of Scottish voters after two MMP elections. I’d like to point out one question in particular:

I would prefer to have been able to vote for individual candidates on the regional vote rather than for a party list. Agree 42.1%, Disagree 12.5%, Neither agree/disagree 24.8%, Don’t know or no answer 20.6%.

Greg,

Need you to clarify a fact.

Are the regional members of the Scottish Parliament elected from closed lists?

Thanks.

eeid – yes Scottish regional MPs are elected using closed lists. but as you can see from the above numbers, a substantial number of people would rather choose their regional MPs. Representations in Upper Bavaria are elected from open lists. Hence my proposal take the best practices from both Scotland and Germany.

I agree that the open lists are better (although your pictures above make it seem like a closed-list system will have like 4X more parties in it than an open-list one, which makes it seem like an even more complicated ballot than one that just says, “Liberal, NDP, Green, PC” on it (with whatever independents or whoever else as well).

But otherwise, it does make sense to have it more open. The only problems I could see is if people get confused who’s running for what since they have both local and regional people campaigning. But since I only see that as a minor problem, these propositions make sense.

Greg,

I want to make a note about the Bavarian landtag.

Has the Christian Social Union (CSU) ever been challenged there? It won the last election in 2003 with 60 per cent of the popular vote.

MMP ensures that SPD gets a foothold. Otherwise, the CSU will win majorities under FPTP similar to the Alberta Conservatives!!!!

You used the SPD sample above to highlight voter choice and accountability with the open lists. If the results for the CSU are the same, I will argue that the result of open lists in Bavaria can lead to the emergence of populist sentiments among regional list candidates. Thus, the possibility of anti-immigration regional candidates being elected for the CSU exists, since the party leaders cannot rein them in. The same can also apply for fringe anti-capitalist candidates for the SPD in Bavaria.

But, eeid, even if the candidates are “extreme”, if the people support them, they should be elected whether the party wills it or not.

I really appreciate the nuanced discussions taking place. But I also see an underlying dedication to electoral reform away from FPTP to something more equitable. With that in mind, I’m surprised the required “supermajority” threshold of 60% hasn’t been the subject of more discussion. Personally, I’m outraged that the Ontario Liberals unilaterally imposed this condition on electoral reform. It’s an insult to the Citizen’s Assembly, and an insult to the people of Ontario. While I commend these Liberals for running for election in 2003 on electoral reform and opening up the possibility of such change, I cannot but condemn these self-same Liberals for virtually making it impossible to pass such reform. Lately McGuinty claimed a “neutrality” on the subject wanting rather the voice of the people to be heard. He wants the people to decide but is willing to accept their decision only if extraordinary conditions are met. I sat in the gallery of the legislature and watched as McGuinty shocked everyone, even members of his own caucus (who evidently also remained unawares of this stipulation and showed their disgust by walking out), by imposing this nearly unprecedented prerequisite of a “supermajority” on any referendum on electoral reform. It’s classic run from the left govern from the right Liberal politics. It’s also a mere pretense of democracy.
From Fair Vote Canada: http://www.fairvotecanada.org/
Dr. Dennis Pilon, University of Victoria political scientist and FVC National Council member:
“Apart from the recent PEI and BC referendums, no voting system change decision in Canada was ever subjected to a super-majority rule. In fact, the establishment of all Canadian federal and provincial voting systems was by a simple majority vote of the designers…[and] all western countries have seen the establishment of their voting systems or any changes in their voting systems handled either through a simple majority vote of parliament or a simple majority vote in a referendum.”

Why does the local riding seats have to be dropped from 107 to 90? Why can’t they keep the 107 seats we have right now plus 1 extra seat to make majorities possible and add the 39 regional seats?

Hi Greg,

I admire your passion for this issue and the work you are doing to get PR adopted in Ontario. I’m a supporter of PR, but consider myself an “average” voter who is not yet familiar with the minutiae of the proposed system. I have a few points:

1) This post, moreso than those you have written previously, strikes me as “spin”. The poll is a weak supporting argument (which you acknowledge) yet you use it to suggest that the system won’t pass without amendment. Makes me question the motivation.

2) I appreciate that you think open lists are legitimately better for the people of Ontario, and personally bow to your expertise in the area. However, the OCA democratically decided to propose closed-lists. I think we have to respect that. That’s what this whole process is about. IMHO, this post somehow seems to have crossed over from expert opinion to biased politicking.

3) I doubt, as an average voter, that the ballot question will come down to closed vs open lists. Most voters won’t bother to understand the difference. The decision will rest on whether we want PR and the potential for perpetual minority governments. Stability will be the deal breaker, if anything. My two cents.

4) Browsing the comments above, I note that people are more confused by your sample open list ballot than by the closed list one. I had the same questions: Should I be voting for more than one regional rep? Can I / should I vote for the same person twice? Assuming my chosen candidate gets elected locally, does that nullify my regional vote for that person? — I don’t think we can assume that the average voter will have researched these questions in advance.

5) The closed list ballot is clear to me. I choose my candidate and my party, and they don’t have to be the same thing. I can be a McGuinty Liberal supporter, but still vote for my local Green candidate. I think this is a simple message with a benefit that is clear to average voters (assuming, of course, that I am indeed average and not completely out to lunch!).

In terms of winning over the masses, I agree with the OCA’s recommendation.

The biggest problem with these long lists is that currently you have an average voter who can’t tell you much of anything about the 3 or 4 main candidates beyond the imcumbant. Try getting them to become knowledgable about two dozen to make an informed decision. My guess is that you will have people picking someone from the party they supported for the local vote by default and due to conventional human behaviour when they are cluless about something, the person that comes first will win.
It happened in Alberta with our senator ‘elections’. The top 3 candidates on the ballot won in addition to the only woman. You’ll end up with a bunch of people names Allen and Brown in the legislature when guys like Smith and Taylor don’t have a change. Stupid.

Stereo Type —

What is unusual here is that I’ve combined reporting of findings with advocacy; ideally these two should be separate functions. If you have followed my efforts of the past many months, you will know that I have revised my models numerous times in order to seek consensus (that is the model with the broadest possible appeal). That is, I respond to what I hear and am willing to let go of personal preferences to accommodate the most people. My advocacy for regional over province-wide and open vs closed lists should be seen in this light. A content analysis of the submissions and reports of the consultations shows that the vast majority of Ontarians who commented — and I’m talking about 80% here — wanted open-list. Most also felt that list MPPs should also be accountable to a region. So, I am advocating for the consensus position.

I don’t believe the CA has debated the question about open vs closed or province-wide vs regional sufficiently in light of the current 90-local seat, 30% list-seat model. As I’ve stated, I believe the decision to endorse province-wide lists (which necessitates closed-list) was made in the initial context of keeping all 107 local ridings and have only 25% list seats. With such a small share of list seats, it was clear that to get the best proportionality required province-wide lists. That made sense given the other decisions they had made to that point. However, one group (I believe led by Hal Willis) took it upon themselves to study a model with fewer local seats and fewer overall seats (they were concerned about exceeding the previous 130 total seats and were concerned that 25% list seats would not allow enough opportunities for women/minorities). All true. I don’t know how they picked 90 seats, but I do know that I emailed the academic advisor detailed maps that I created for a 90-seat model. So I think lowering the local seats to 90 and bringing the total seats was the right decision. Now what should have happened at that point (with the local seats now at 90 and increased the list seats to 30%) was to re-evaluate the decision to go with province-wide and closed vs regional and open. Remember, Ontarians made it quite clear at the consultation meetings that they did not want closed-list, so that is something that the CA should have been concerned about.

From my emails with the academic advisor, I had the sense quite early on that the CA might keep most of the local seats, go with MMP-lite and thus need province-wide closed list. So I took it upon myself to conduct a survey to find out if what we heard in the consultation meetings (i.e. that most people objected to closed-list) would influence whether they would support change or not. Because of the caveats I mentioned in the post, I did not publish these findings, instead preferring to make other arguments on this site and through emails to various people. When that proved ineffective (the argument not being that closed-list is better, but that it was “too late” to change it), I decided reluctantly to publish the findings, hoping to add some urgency in advance of this weekend’s meeting. So that’s the genesis of the this post. Hopefully that gives you some idea of my thinking.

I take spin to be making falsehoods to cover truths. Since I am reporting what I believe to be true, I am certainly not engaging in spin. Am I politicking? You bet. I am fighting for what I believe the Ontarians want. And I make no apologies for that.

I must say that I find your sample ballot to be confusing, Greg. I think I understand how it works, but I’m not sure if it would be the case of all voters. As Stereo Type pointed out, I think you’d find a lot of cases of people voting for the same person as local representative and regional representative (in cases of dual candidacy such as the ones in your sample). I should be made clear that while this is allowed, it is not mandatory.

Also, I think you underestimate the size that your ballots could have. You showed us a ballot with only four parties and a list of five candidates for each party. In regions such as Horseshoe and Toronto, parties would have to nominate at least eight candidates (and maybe more, to account for dual candidacies) or run the risk of having underhangs. Having more parties on the ballot would also make it balloon. The “produce” ballot has no less than 25 parties on it (but of course it serves to elect a statewide list; in Ontario it’s likely that minor parties wouldn’t present candidates in all regions).

This said, I do agree with you that open lists are preferable to closed lists, but they come at the cost of increased complexity, and like Stereo Type, I’m not certain the average voter really cares about the difference (your survey notwithstanding). What I did like, though, is the fact that you allow independents to run as regional representatives. I never thought of that, but given that your proposal puts a lof of weight in calculating the compensation seats regionally, it makes sense.

I take spin to be making falsehoods to cover truths. Since I am reporting what I believe to be true, I am certainly not engaging in spin. — Comment/commentaire by democraticspace 04.11.07 @ 2:26 pm

Hi Greg – thanks for your response.

I don’t think ’spin’ implies falsehood. In fact, I think it is just the opposite: presenting factual truths in a way that suits your cause. My intention was to point out that the poll data, while true, are weak. Using the poll weakens your entire argument by making it seem as though you are grasping at straws. Forced to use a weak argument as a last-ditch effort to make the OCA see your point of view. I didn’t mean any offense, just that I found your previous logical and well-reasoned expert opinions to be far more convincing.

The main point, however, remains: the closed list ballot is simple and easy to explain to average voters who will probably only see this decision at the level of FPTP vs PR.

Stereo — by your logic, then it wouldn’t matter what the CA recommended (i.e. it would still be FPTP vs PR). I give Ontarians more credit than that. People will take a look at what is on offer and decide. They won’t get into the details. But if the NO campaign can make simple arguments such as “closed list means no accountability” then it will impact how people vote. What’s the argument against open-list? That it gives too much choice? I guarantee that the former argument will resonate much more. And who is going to argument against giving their region a voice in Queen’s Park vs. a third of the MPPs not tied to any community? Again, which argument will resonate more? The CA is taking a huge and unnecessary risk in proposing province-wide closed list, especially since the reasons for initially proposing it have gone away now that they reduced the local seats to 90 and upped the list seats to 30%. Many acknowledge the benefits of regional/open, but the argument at this point is that it is too late, they’ve run out of time.

Mark:
You’re correct that when there are many unrecognized candidates on a list, voters tend to prefer the first one. This can be mitigated by having Elecitons Ontario randomly choose before the election the order in which the parties are listed on the ballot for each riding, and the order in which each party’s candidates are listed.

If we see consistent and troubling bias shown towards the first candidate, then we can take the further step of separately randomizing the order on each ballot. Obviously counting is more difficult when every ballot looks different, but this has been done successfully for STV elections in Australia (Robson Rotation)

Greg:
Thanks for your continued election blogging!

I agree that more voter choice in party lists would be preferable. With a closed list there’s the potential for party flacks to be placed in high positions on the list, getting elected despite no personal support among the electorate. Also, I’m in favor of more intra-party competition and slightly weaker parties, so I’d be happy to scrap closed lists.

However, I have three concerns about the sort of open list you propose:

1. Women and Minorities: Canada has a deplorably low level of political representation by these groups, as I’m sure you’re aware. Unfortunately, completely open lists don’t usually tend to improve this. (If I could conjecture why, it probably involves a conjunction of stereotypes and long lists of unfamiliar names. But the reason why this occurs isn’t critical–the point is that it happens.)

The OCA seemed to consider this one of the more important reasons for closed lists, and I feel that choice should be respected. That doesn’t mean we can’t propose changes to their decisions, but it does mean we should try to ensure that our changes don’t belittle something they considered important.

2. Punishing bad behavior: Another problem with open lists is that they make it difficult to stop the election of candidates who aren’t well-liked by most people, but have a small core of supporters.

Usually the vast majority of list votes are spent (one might say wasted) on the party’s leaders, or high-profile candidates. The following positions are then determined by very few votes, far less than a quota’s worth. A solid core becomes much more valuable than being popular among the electorate at large. So candidates who are generally disliked for their extremism, past scandals, or patronage are elected relatively easily.

This can be a problem in a closed list as well, of course! However a closed list offers more opportunities for a party to control its candidates, lowering the list position of those that would harm the party as a whole. More generally, I think that both a party’s control of its list and the ability of the electorate to influence list ranking are useful tools. I would prefer a system that utilizes both of these tools, rather than just one.

3. Voter confusion: Finally, forcing the voter to choose an individual candidate for the list vote could have some unfortunate side-effects. As was mentioned above, it will make the ballot quite long–it’s easy to envision forty candidates in the large metropolitan regions.

But more importantly, I don’t think voters will understand that the “regional vote” is mostly about determining party seat share. If a voter strongly prefers the Apple Party, but recognizes none of that party’s candidates, I fear he will vote for another party, which defeats the purpose of list PR. A blurb of explanation on the ballot will not likely be enough to prevent these sorts of misunderstandings.

There are ways that these concerns could be addressed in whole or in part, while still maintaining voter choice between candidates of the same party. Some options:

- Flexible lists, in which a voter can make an “above the line” vote for a party, or vote “below the line” for an individual candidate. This lets voters focus on parties, and gives parties a certain amount of control over their lists–but still gives the voters the final choice.

There are many potential variants of flexible lists. Some require a threshold of below the line votes before allowing the list order to be changed. Others allow the voter to choose multiple members of the same party, or (equivalently) vote for a party but strike out the names of candidates she doesn’t like. And these multiple choice variants can be normalized to one vote, or not. One has to be careful not to get too complicated, so I’m not sure which variant I’d prefer.

- Multiple lists, in which each party presents more than one list to voters. Each party is allocated a number of regional seats based on the total vote of all their lists; then those seats are allocated proportionally to each of the lists.

Perhaps three lists per party would be a good number. This would considerably shorten the ballot, but still provide the voter some choice of which candidate(s) she prefers from a given party.

- Legal constraints on how parties choose their lists. If we demand some sort of democratic primary to choose party lists, but keep closed lists, that could be “open enough” to satisfy many voters. One can imagine further constraints on how primaries work, but I’m not sure it’s worth

As far as regional lists are concerned, I’m not really thrilled at the idea, but I’d be willing to accept them. They raise the effective threshold, although I prefer low thresholds. And I’d rather have the best list representatives I can than have them hail from the same region as myself. (What’s the motivation for segmenting the electorate by region, as opposed to, say, profession, or age?) However, I recognize that we need a limit on district magnitude if we’re going to have non-closed lists, and that’s more important to me. So I can live with the regions.

I hope to hear your thoughts!