Considering the 2008 Election Results: Canada’s Electoral System Needs to be Replaced

By Harold Jansen on Oct 17, 2008

I’ve been thinking over the election results since Tuesday night and I keep coming back to one thing: Canada’s electoral system needs to be replaced. I need to point out first of all that I’m not an electoral system ideologue. I don’t think there’s one abstract, theoretically derived system that works best in all countries and at all times. Canada currently has a single member plurality electoral system (often misleadingly labelled the “first past the post” system). This system works pretty well in countries where there are only two strong parties that are competitive with one in all regions of the country. That was the case in Canada from 1867 to about 1921. It hasn’t been the case for a long time and it’s a big reason why our electoral system needs reforming.

The reality is that Canada has a multi-party system and that fact isn’t about to change anytime soon. There was a lot of talk in this election about the vote-splitting on the left, which reminded me a lot of the discussions about vote-splitting on the right in the 1990s. I’ve seen some discussion about how we need to reduce the number of parties on the left, perhaps by having the Green party disappear. However, there are reasons why people started, joined and voted for the Green party rather than the NDP, just as New Democrats don’t want to be Liberals. I know plenty of former Canadian Alliance members who chafe at the principles they’ve had to see watered down as part of the Conservative party of Canada. Rather than reducing the number of parties to fit the requirements of an archaic electoral system, wouldn’t it make more sense to adopt a system that recognizes the diversity of political views that Canadians already hold.

This multiplicity of views means that the translation of votes into seats under our system is pretty erratic. Every election, I ask my students to predict how many seats each party will get in the election. I was struck by how poorly they performed this time. The reason is not that they’re not very bright, but that there’s a degree of randomness in the translation of votes to seats. As an example, I’ll pick the surprise in my own province. Conservative Rahim Jaffer lost his seat in Edmonton Strathcona. Hearing that, you might expect that he had fallen out of favour with voters. In fact, he received almost exactly the same proportion of the vote on Tuesday night (41.6%) as he did in 2006 (41.7%). What changed was the way all of the other parties’ votes divided (or didn’t). There are many examples like that to be found in the election results. It also helps to explain why small shifts in the overall popular vote can have a significant impact on seat totals, such as we saw with the Conservatives and the NDP on Tuesday night.

The other significant reality of Canadian politics is the regional disparity in support for the parties. We see this most evident with the Bloc Quebecois, which only runs candidates in Quebec. But we also see it with other parties. Because only one party can take a seat, it means that for supporters of a minority view, elections are an exercise in frustration. Ask a Liberal supporter in Alberta how satisfying it is to vote; ask a Conservative supporter in downtown Toronto the same thing. The problem with our electoral system is that it exacerbates these regional differences. Six out of every ten voters in Quebec did not vote for the BQ on Tuesday night, but they took two-thirds of the seats. About a third of Albertans don’t vote Conservative; you’d never know it looking at the election results. A different electoral system is not going to suddenly revive Liberal fortunes in western Canada, nor is it going to help voters in urban centres fall in love with the Conservatives. But there are systems out there that will more accurately reflect the true reality of party support at the regional level.

So, what’s the alternative? Well, I’d suggest a form of proportional representation, and if you want my specific recommendation, it would be a mixed-member proportional electoral system. This system was pioneered in Germany and has increasingly become the consensus choice of experts on electoral reform. MMP is seen to provide the “best of both worlds,” because it combines some of the best elements of the SMP system with the proportional overall outcomes of proportional representation. Under the system, voters elect an MP to represent their district, exactly as we do now. Voters would also get to cast a second ballot for the party they support (which may or may not be the same as the party of their preferred candidate). What we do then is give parties additional MPs from party lists so that their total number of MPs equals the number they should get based on their percentage of the vote.

How would the results on Tuesday night have been different under an MMP system. The first table has the actual results on election night.

Province

Conservative

Liberal

BQ

NDP

Green

Independent

N.L.

0

6

 

1

0

0

PEI

1

3

 

 

0

0

NS

3

5

 

2

0

1

NB

6

3

 

1

0

0

QUE

10

13

50

1

0

1

ONT

51

38

 

17

0

0

MAN

9

1

 

4

0

0

SASK

13

1

 

 

0

0

ALTA

27

0

 

1

0

0

BC

22

5

 

9

0

0

NORTH

1

1

 

1

0

0

TOTAL

143

76

50

37

0

2

The following table shows the hypothetical results of the election under an MMP system. If you want the assumptions and technical details I used in making the calculations, you can consult the fine print at the end of this post.

Province

Conservative

Liberal

BQ

NDP

Green

Independent

N.L.

1

4

 

2

0

0

PEI

2

2

 

0

0

0

NS

3

3

 

3

1

1

NB

4

2

 

3

1

0

QUE

16

18

29

9

2

1

ONT

42

36

 

19

9

0

MAN

7

3

 

3

1

0

SASK

8

5

 

3

1

0

ALTA

19

3

 

4

2

0

BC

16

7

 

10

3

0

NORTH

1

1

 

1

0

0

TOTAL

119

81

29

57

20

2

The result is a House of Commons that more closely corresponds to how Canadians actually voted on election night. The biggest change comes for the BQ; no other party in Canada benefits more from the distortions of our SMP system than them. The seats they would lose in Quebec would be divided among the other parties, but the Conservatives and NDP would especially benefit in that province. Overall, the NDP would be a net beneficiary of the system with better representation across Canada, but especially in Quebec. The Green Party would have elected 20 MPs from every province except Newfoundland and Labrador and Prince Edward Island. Those gains obviously come at the Conservatives’ expense as the party now gets the typical SMP leading party bonus. For most of Canadian history, though, the Conservatives have been the second place party and the electoral system has not been especially kind to them. The other key thing about this for the two major parties is that they are decidedly national, with MPs from every region of the country.

There are many criticisms of proportional representation systems, but I’d argue that most of them either do not hold up well to critical scrutiny or are problems we already face in Canada. One common criticism is that we would never see majority governments under PR/MMP and that all governments would be minorities or coalitions. How is that different than what we have right now? A lot of people are arguing that as long as the Bloc is getting 40-50 seats, a majority government is almost unattainable. Furthermore, if Canadians are divided over the direction of their government, why should we artificially choose one of the parties to be able to direct the government in its entirety? Another related criticism is that we would see more frequent elections under a PR system. Again, I can scarcely see how we would do worse than how we’re doing now. We have had three elections in four and a half years and will likely see another one in two-three years. This last election was triggered because Stephen Harper and the Conservatives thought they could get a majority if they called it when they did. Under a PR system that incentive would be gone. The Conservatives would have gained around four seats over last time, hardly worth bringing down the government for. The flaw with all of these arguments is that they equate minority or coalition governments and more frequent elections with effective government. The only really good study of this found this not to be the case. Another frequent criticism is that extremist or anti-system parties can basically use their position to get what they want in order to keep a government in power. Again, that’s what we have now. I would define a party dedicated to breaking up Canada as an extremist party and our system right now gives the Bloc Quebecois inordinate influence in the House of Commons.

In other words, we’re getting all of the supposed disadvantages of PR without any of its advantages, such as higher voter turnout (although there are many causes of turnout decline that are unrelated to the electoral system) and greater representation of women. Canada’s society and party system would more appropriate be served by a proportional system. I’d argue that’s been the case since 1921 and the situation is only worsening.

What are the prospects of electoral reform? Probably not great, since the parties who control our choice of electoral system all have a vested interest in the system. The only hope for a Conservative majority is through the assistance of the electoral system. For the last six elections, the BQ has been the biggest beneficiary of the system. Historically, no party has been helped by the system more than the Liberals. They’ve governed most of the time since 1921, but have been doing so on an increasingly small proportion of the popular vote, as Richard Johnston has noted (PDF link). All of these parties have a vested interest in keeping the system how it is. No party has been hurt more by the electoral system than the NDP, but the party has been incredibly ambivalent in its commitment to reform. No NDP provincial government has ever introduced electoral reform. When the Liberals have relied on the NDP in a minority government situation, the NDP has never demanded movement towards electoral reform as its price for its support. Why? When the party’s leader says, “I’m Jack Layton and I’m running for Prime Minister,” he knows that the only way that’s ever going to happen is with a big helping hand from the electoral system.

It’s going to take citizen education, action and pressure to accomplish change. And it’s going to be a long and uphill struggle. But our political system is showing the cracks of an electoral system that no longer serves its needs.

The fine print: My MMP projections are based on the assumption that roughly half of the seats in each province are elected in single member districts and their breakdown is the same as what we saw on election night. I decided that in the case of the two independents, that they would have had enough appeal to have been elected even under the larger districts in an MMP system. I also assume that there is no “split ticket” voting; in other words, the voter would support the same party as the candidate they supported. Obviously, the assumption is also that people would vote the same under the PR system as they voted under SMP, which is likely true for the vast majority of people, but almost certainly not true for all people. We lack any data to assume or project other behaviour. The overall results treat each province as a district (I combined the North into one district) and allocates the seats using a largest remainders formula and a Droop quota.

Comments

Comment viewing options

Select your preferred way to display the comments and click "Save settings" to activate your changes.

Calculations

So are you proposing separate nomination lists for each province? ie. PR seats in a province are allocated using the Pop. voet from that province

I guess that would give PR MP's some sort of regional standing, one problem with PR being that MP's don't have a riding to represent. If PR seats are allocated using regional pop.vote tally's, then at least you would be able to say that this PR MP represents Alberta and this other PR MP represents Ontario...

I'm down with MMP then I guess. THe only problem of course is that we would need a majority in the parliament to put this to a referendum. As long as the majority in the house stands to lose from this proposal, it will never happen. (CP + Bloc) = 193 In fact, I think it is a practical impossibility for the majority of MP's in the house to stand to gain from MMP. Basically you would have to rely on the good graces of the majority to wave their own personal interests in lieu of the greater democratic good, and we know how likely that is... (although it happened in Ontario...)

Yes, separate lists per province

I'm not sure that nation-wide lists would be an option in Canada. Representation in the House of Commons is constitutionally divided between provinces and I don't see how we could have national lists without a constitutional amendment. I'm not a consitutional lawyer, though, but that's how I read it. I also think it makes sense politically.

I agree: this is a tough sell in the current Parliament. Unless politicians can benefit from being seen as an initiator of reform (which is what I think McGuinty was trying to do), there isn't much incentive to move forward. I suspect as well that the gains from that wouldprobably be outweighed by party self-interest. But change does happen, so I'm goingto keep making the case for this.

MMP improvements

This article is a very fine summary of the case for electoral reform.

An important question is "who are compensatory MPs accountable to?" I think it's becoming clear that most Canadians don't like closed province-wide list components in an electoral system. Luckily, MMP models with regional open lists are available. This means that, on one side of the ballot, you vote for your local MP, and on the other side you vote for your party and one of its regional candidates you prefer. The German province of Bavaria has used this model for 60 years. It has also been recommended for Canada by the Law Commission of Canada, and is being studied in Quebec following the report of Director-General of Elections last December, which recommended MMP with nine regions for compensatory MNAs in Quebec.

Federally, you might see seven regions in Ontario, five in Quebec, and two each in BC and Alberta. This would create both reasonable proportionality and regional accountability. Northern Ontario voters would still elect the same number of MPs, and Northern votes would elect Northern MPs. The Greens would win seats almost everywhere. In Quebec outside Montreal, where they got only 3.2% of the vote, these regions would be too small to give them a seat, but frankly that's a pretty small percent.

Wilf Day Port Hope, Ont.

Good point

Excellent comment, Wilf. I basically did the provice-wide list method for simplicity. This wasn't meant to be a full-fledged MMP simulation, just a quick half hour with me and my copy of Excel. For the larger provinces, there is definitely merit in sub-provincial lists. The benefit of this is that it would prevent the Ontario list from being dominated by Toronto area candidates, for example. And I agree, it would be much more easy to accept. We could also get into lengthy discussions over closed vs. open lits and all of the other options. But those are details. Although I prefer almost any kind of MMP to the other alternatives (as long as there are enough list seats to fully compensate), I'd happily support list PR or the single transferable vote as an improvement over what we've got.

Thanks

One of the best posts on Mapleleafweb to date. Thanks Harold

Syndicate content