Wednesday, May 18, 2016

A Moderate Mixed Member model can balance all the values needed for electoral reform.

A Moderate Mixed Member model can balance all the values needed for electoral reform.

This proportional representation model keeps our local ridings.

As a general rule, a group of eight current ridings becomes a region electing five local MPs and three regional MPs.  The vast swaths of voters who are denied their preferred MP by the single-MP riding elections elect regional MPs to top up the local results The total MPs match the vote share in the region. Almost every vote counts, moderately, without risking a multitude of mini-parties.

The local ridings become 60% larger, but in return we help elect three regional MPs accountable to our local region, as well as one local MP who will champion our own community. The government will be accountable to MPs reflecting a true majority of voters.

Law Commission of Canada

In 2004 the Minister of Justice, the Hon. Irwin Cotler, tabled the Law Commission of Canada’s Report recommending a mixed member proportional system, just as outlined above. MMP is used in Germany, Scotland, New Zealand, and other jurisdictions.

The Law Commission recommended one vital improvement: no closed lists. All MPs are elected and have faced the voters. If voters for a party are entitled to elect a regional MP, it will be the party’s regional candidate who got the most votes across the region.

You have two votes. With your first vote, you help elect a local MP as we do today. With the second, you also help elect a few regional MPs: it’s proportional. You can cast a personal vote for the regional candidate you prefer: it’s personal.

The regional candidates will be democratically nominated in 42 regions, each small enough to make them accountable, not by the province-wide lists that Ontario voters rejected in their 2007 referendum.

Fair to all voters

In the 2011 election, Canada elected a one-party government where one man had complete power supported by only 39.5% of voters. Yet voters in eight of Quebec’s ten regions had no voice in the government. While 19% of Canadians voted Liberal, Liberal voters in 24 of those 42 regions had no voice in Ottawa. Despite the NDP forming the Official Opposition in 2011, NDP voters in 13 of those regions had no voice in the House of Commons.

Never again should this be possible.

Look at the factors the government has placed before the Special Committee on Electoral Reform.
  1.  Effectiveness and legitimacy: that the proposed measure would increase public confidence among Canadians that their democratic will, as expressed by their votes, will be fairly translated and that the proposed measure reduces distortion and strengthens the link between voter intention and the election of representatives.
With this model every vote counts. Every voter has equal and effective votes. We get what we voted for.

As the Ministry notes, winner-take-all elections routinely create governments without majority popular support, and at times, with less popular support (but more seats) than the second place party. And the Australian system of preferential ballots in single-MP ridings mean voters end up with a choice between only two alternatives. Being represented by my second choice is the problem, not the solution.

As detailed below, on the votes cast in 2015, voters for all three parties would elect either local or regional MPs in almost all of the 42 regions across Canada. This keeps parties from monopolizing regions

2.    Engagement: that the proposed measure would encourage voting and participation in the democratic process, foster greater civility and collaboration in politics, enhance social cohesion and offer opportunities for inclusion of underrepresented groups in the political process;

The Ministry notesOf the 34 member countries of the Organisation for Economic Co-operation and Development (OECD), Canada is one of only three that continue to use the FPTP system to elect legislators.” The rest mostly use proportional representation and have stable majority coalition governments like Germany.

As experience shows in almost all of the many countries in Europe and elsewhere using PR, when you know that you will not able to govern alone, you learn to work together. No one man or one party can get his own way.

As former Ontario Attorney-General John Gerretsen loved to say “Nobody is ever 100-per-cent right and nobody is ever 100-per-cent wrong. Governing is the art of compromise.” That’s why he supported proportional representation.

Canada’s new Minister for Democratic Institutions, Maryam Monsef, was interviewed by Peter Mansbridge. He said "People want the system changed. How are you going to go about that?" She replied: "Yes, Canadians voted for real change, for a government that puts their needs ahead of the party's. A government that celebrates our diversity, and not takes a divisive approach. . . . I hope that Canadians will be proud of their politicians and their country because of the kind of work we do and the respectful manner that we go about it.”

With two votes, you can vote for the party you want in government. And you can also vote for the local candidate you like best regardless of party, without hurting your party, since it's the second (regional) ballot that determines the party make-up of the legislature. About 32% of voters split their ballots this way in New Zealand with a similar system. MPs can thus win support from constituents of all political stripes, enabling them to act more independently in the House of Commons.

3.    Accessibility and inclusiveness: that the proposed measure would avoid undue complexity in the voting process, while respecting the other principles, and that it would support access by all eligible voters regardless of physical or social condition;

Today many voters feel excluded, if their favourite party or candidate has no chance of election. Turnout is 5 to 6 points higher in countries where the electoral system is proportional, says research published by Elections Canada.

Complexity and accountability are trade-offs. The system used in Germany’s federal elections is very simple: you cast one vote for a local candidate, and one for a party’s closed list. The variation used in the German province of Bavaria has larger ballots, because you can vote for your preferred regional candidate of your party, just as the Law Commission of Canada recommended. Every MP has faced the voters, but in return, you have a bigger ballot. A good trade-off.

4.    Integrity: that the proposed measure can be implemented while safeguarding public trust in the election process, by ensuring reliable and verifiable results obtained through an effective and objective process that is secure and preserves vote secrecy for individual Canadians;

Never again will one man gain absolute power with the support of 39.6% of the voters. The public will know that every vote is effective.

5.    Local representation: that the proposed measure would ensure accountability and recognize the value that Canadians attach to community, to Members of Parliament understanding local conditions and advancing local needs at the national level, and to having access to Members of Parliament to facilitate resolution of their concerns and participation in the democratic process.

This is where this model shines.

Voters are guaranteed two things which equal better local representation:
1.         A local MP who will champion their area, accountable only to local voters.
2.         An MP whose views best reflect their values, someone they helped elect in their local riding or local region.

The best of both worlds. By comparison, today the voter faces the party’s only candidate.

It is not just northern ridings which need single MPs. Canada has at least 126 present ridings whose communities rely on their local MP to champion their community. And the claim that large urban areas do not want their borough, ward or local town to have its own MP is over-rated.

The region is small enough that the regional MPs are accountable. Every region will have fair representation in both government and opposition. 

Yes, the riding is larger, but in return, voters with their diverse voices will have diverse MPs, just as they do in Scotland and Wales.

MPs will be even more accountable if they must be democratically nominated, which would be a great improvement under any voting system.

What would this model look like?  

The Appendix below lists the regions I have used for this projection of how it would work. In reality, the usual impartial Boundaries Commissions would define the ridings and regions after the usual public hearings. Canada’s Chief Electoral Officer Marc Mayrand has assured Canadians that this can be done even if they have only 10 months, but I expect they will have more.

For example, the BC Interior and North has nine MPs. That would become six local MPs and three regional MPs (but in a metropolitan region that might become five and nine.) Last October Conservative voters elected five MPs with only 37% of the vote, the NDP three and the Liberals only one. With this model let’s assume the Conservatives win three local seats, the NDP two and the Liberals one. Liberal voters elect two regional MPs, and NDP voters one. As compared with the actual result last October, the Liberals have taken two seats from the Conservatives.  

For another example, look at the nine MPs in the northern suburbs of Montreal, the North Shore, the Laurentides—Lanaudière region. The Bloc swept six of those seats with only 31% of the vote, in four-way races. Let’s assume five local MPs (three Bloc and two Liberals). Liberal voters elect a regional MP, NDP voters two, and Conservatives one. Every vote counts.   

Moderately proportional

By using regions smaller than Scotland’s 16-MP regions, we get moderately proportional results. This was the goal of the UK’s Jenkins Commission when they too recommended 8-MP regions. It was also the goal of Stephane Dion’s P3 model.

If we had used province-wide perfect proportionality for 338 MPs, the results would have been: Liberal 137, Conservative 109, NDP 67, Bloc 15, Green 10. With this model's simulation, I calculate the results for 338 MPs are: 145 Liberals, 106 Conservatives, 70 NDP, 14 Bloc, and 3 Greens.

In 2015 the Green Party got only 3.5% of the vote. Some PR systems would use a threshold of 4% or 5% and give it no MPs beyond Elizabeth May. Even without a threshold, this model would let those Green Party voters elect only three MPs, all from British Columbia, not the 10 they deserve.

But, as Prof. Dennis Pilon says in this video "Now keep in mind that, when you change the voting system, you also change the incentives that affect the kinds of decisions that voters might make. For instance, we know that, when every vote counts, voters won't have to worry about splitting the vote, or casting a strategic vote. Thus, we should expect that support for different parties might change."

Green Party leaders know that, when more voters vote, their vote would double. In that case, this moderate model would let Green Party voters elect 15 MPs: six in BC, five in Ontario, and one in each of Quebec, Manitoba, Nova Scotia and New Brunswick. No, not the full 22 MPs that perfect province-wide proportionality would give them, but pretty good for a moderate model.

Similarly, it would give almost perfect results to voters in every province. It would leave the Conservatives short one MP in New Brunswick, where the Liberals swept the province with only 51.6% of the vote, and short two MPs in Toronto, where the Liberals swept the city with only 51.9%. Along with the Greens missing seven MPs and the Bloc missing one, this would leave a bonus of eleven MPs, which happens to become eight Liberals and three NDP. In several provinces the luck of rounding off fractions in regions will give a party one seat here and cost them one seat there, cancelling out across Canada.

How would regional MPs operate?

Most regional MPs would each cover several ridings. Take Saskatchewan as an example. On the votes cast in October 2015, Liberal voters in Saskatoon and North Saskatchewan would have elected two regional MPs. They might be based in Saskatoon and Prince Albert, but they would likely have additional offices in North Battleford and elsewhere, just as MP Kelly Block has offices in Martensville, Humboldt and Rosetown. This is just the way it’s done in Scotland, where two regional MPs from a party will normally split the region between them for constituency service purposes, and hold office hours rotating across their region or their part of it. 


What if we used the preferential ballot to elect local MPs?

If the preferential ballot were used to elect local MPs, Eric Grenier and others have shown it would have elected more Liberal MPs. But an MMP model might compensate for the disproportional local results, if they are not too extreme.

My simulation shows this model is not quite proportional enough to cope with the increased local Liberal sweeps, but close. The result in the West and Atlantic Canada is unchanged.

In Quebec, this model had already given the Liberals a bonus of three MPs and the NDP a bonus of one: two from the Greens, one from the Bloc, one from the Conservatives. With the preferential ballot the Liberals end up with a further bonus of another two MPs, one from the Bloc, one from the NDP.

In Ontario, this model had already given them a bonus of three MPs and the NDP a bonus of one: three from the Greens, and one from the Conservatives. With the preferential ballot the Liberals end up with a further bonus of another five MPs, three from the Conservatives, and two from the NDP.


Can we make this model more proportional, to cope with the increased disproportionality? Three of the newly disproportional regions in Ontario would need 50% of the MPs to be regional, and another would be equally unlikely. I suspect that Canada would just have to live with the model being less than perfectly proportional.

Technical notes

The Law Commission recommended Canada elect at least one-third of the MPs in each region as regional MPs. That number was based on a 1989 study. Since then, we have seen the Wales Assembly have very disproportional results despite one-third regional MPs. No other mixed member model in operation uses such a low ratio. However, 50% regional MP would mean doubling the size of local ridings.

The Law Commission did not specify the region size. If we make the regions small enough that the regional MPs are accountable, we need a larger ratio such as 40%. If eight ridings become five new local ridings, the regional MPs are 37.5% of the total, and ridings are 60% larger. In other regions, it might be ten become six, 67% larger. But in a few rural areas it will be four local of six, or six local of nine. My current simulation ends up with an average of 37.8% regional MPs.

For a method of rounding fractional results, the Law Commission did not expressly recommend the “highest average” calculation used in Scotland, but they used it in their demonstration. Since then, we have seen that calculation give the Scottish National Party a majority government on 45.4% of the votes. Several other calculations can be used.  One criterion is simplicity and transparency. The simplest is “largest remainder” which Germany used to use, which they also claimed offset the high 5% threshold by being generous to smaller parties. If voters for party A cast enough votes for 4.3 MPs, Party B deserves 2.3 MPs, and Party C deserves 1.4 MPs, which party gets the eighth MP? Party C.

Appendix - the regions:

My simulation using this model lets voters elect 209 local MP in the ten provinces, and 126 regional MPs in 42 regions.

In Atlantic Canada each province is one region. The other regions are:

Ontario:
Ottawa—Cornwall 10, East Central Ontario (Kingston—Peterborough) 9, Durham—Rouge Park 6, York Region 10, Scarborough—Don Valley 8, Toronto Central 8, North York—Etobicoke 8, Brampton—Mississauga North 7, Mississauga—Halton 8, Hamilton—Niagara—Brant 11, Central Ontario (Barrie—Owen Sound) 6, Waterloo-Wellington-Dufferin 8, London-Oxford-Perth-Huron 7, Windsor—Sarnia 6, Northern Ontario 9.

Quebec:
Montreal-est 8, Montreal-ouest 6, Montreal-nord—Laval 8, Laurentides—Lanaudière 9, Outaouais—Abitibi-Témiscamingue—Nord 6, Longueuil—Suroît 10, Montérégie-est—Estrie 6, Mauricie—Centre-du-Québec 6, Quebec City—Saguenay-Lac-Saint-Jean—Côte-Nord 11, Chaudière-Appalaches—Bas-Saint-Laurent—Gaspésie 8.

Manitoba:
Winnipeg 8, Manitoba North and South 6

Saskatchewan:
Regina—South Saskatchewan 6, Saskatoon—North Saskatchewan 8

Alberta:
Calgary 10, South & Central Alberta 8, Edmonton 10, Northern Alberta 6

British Columbia:
Vancouver-Richmond-Delta 9, Burnaby—North Shore—Coquitlam—Maple Ridge 8, Surrey—Fraser Valley 9, BC Interior and North 9, Vancouver Island 7

Voters for all three parties would elect either local or regional MPs in each of the 42 regions across Canada, except no Conservative in 8-MP Montreal-est and in 6-MP Outaouais—Abitibi-Témiscamingue—Nord.  Bloc voters would elect MPs in each of Quebec’s regions except 6-MP Montreal West. The results for 338 MPs are: 145 Liberals, 106 Conservatives, 70 NDP, 14 Bloc, and 3 Greens.

Who should have been elected?

Canada's Missing MPs.

Wednesday, April 6, 2016

Alex Himelfarb to Tasha Kheiriddin: “it’s false because it’s not true.”


That was a great debate April 2 between Andrew Coyne and Alex Himelfarb, advocating for proportional representation, and Conservatives Tasha Kheiriddin and Michelle Rempel.  

My favourite moment was at 38:45.


Tasha Kheiriddin had not done her homework.

“There would be a list. The parties would choose that list. You would not vote for a local MP. The party could re-appoint people. The party would choose . . .”

Moderator Even Solomon interrupts: “Some people are yelling out “false.” And so was Alex Himelfarb, former Clerk of the Privy Council, the very top position in Canada’s civil service.

Tasha Kheiriddin: “I don’t see why it would be false.”

So Evan Solomon calls on Himelfarb. 

“It’s false because it’s not true” say Alex Himelfarb, to great laughter and applause. 

“Almost invariably what the Commissions have recommended has local MPs . . . There are systems which are designed otherwise, but not for Canada.”

Andrew Coyne spelled it out: “Not all PR system use lists. The members on the lists can be elected by the members at large. People can choose names off the lists, voters can vote directly off the list. Lists are not the caricature that’s being presented.”

Worth sharing.



Multi-MP ridings in big cities, proportional single-MP ridings in the rest of the country

Note: this post was revised and updated Sept. 27, 2016.

In the excellent debate on proportional representation April 2 at the Broadbent Institute’s 2016 Progress Summit, Andrew Coyne said most of our population live in large-urban areas, so you could have multi-MP ridings quite easily in those areas. “You may wish to have some kind of hybrid where the other areas could have one MP per riding. That’s going to be part of the debate.” See 46:55:


This echoes the similar suggestion by Jean-Pierre Kingsley, Canada’s Chief Electoral Officer from 1990 to 2007, who has gotten great interest by suggesting an urban/rural hybrid system.

So some of us started to discuss an urban+rural model for proportional representation. 

Rural-Urban Proportional Representation

This has become the Rural-Urban PR model now being proposed by Fair Vote Canada, along with the other models Fair Vote Canada also proposes. Details are here. I will now outline the Kingsley-inspired example.

Multi-MP ridings in big cities, single-MP ridings in the rest of the country.

Large urban centres over 100,000 people contain 60.0% of Canada’s population, says Statistics Can from the 2011 census. A couple of those are cities with a bond to the single MP: Kingston just elected its former mayor Mark Gerretsen as its MP, while Guelph elected Lloyd Longfield. On the other hand, a couple of centres under 100,000 – like Saint John (N.B. – could be in a multi-MP riding. So could some people living near those large urban centres.

Still, that leaves about 86 MPs from the other ridings (see list below). In Ontario and the West the majority of those seats are safely held by Conservatives. No other party will agree to hand them these seats.

Whichever parties hold those 86 seats, those voters are the ones most in need of having their voice no longer silenced. In Quebec, 428,000 Conservative voters cast ineffective votes last October. In Atlantic Canada, everyone but Liberals was shut out. So we have to give unrepresented voters in those 86 ridings representation too.

Top-up seats - Regional MPs

Sweden elects 11% of their MPs as top-up MPs, because their multi-MP districts have a range of sizes, one with only two MPs, and four more with 4, 5 or 6 MPs. So one way to look at this model is, Sweden's model, but with about 25% of the MPs from our present single-member ridings. Since so many of the MPs come from multi-MP ridings, it will be easy to add a few regional MPs to compensate for any disproportional results. 

In my simulation, we keep 23% of MPs from single-MP ridings, and need only 15% of MPs in regional top-up seats. To keep the House of Commons the same size, the single-MP ridings would become about 17% larger.  That’s better for those communities (and their MPs) than having the ridings become 56% larger under an MMP model.

The typical top-up region might contain 20 MPs: 13 from multi-MP districts (typically with four or five MPs each) and four from single-MP ridings, all topped up by an additional three MPs (15% of the region). For the six smaller provinces the province is the top-up region. See details of the 18 regions below.

So who would elect the 51 regional MPs?

Swedish option

They could, as in Sweden, be the best runner-up candidate in the region where that party’s voters were the most under-represented. Some people fear that MMP would mean list MPs from large urban centres would swamp the small-urban and rural communities. They prefer the Swedish model.

The three top-up regional MPs would be the party’s candidate in the region who was the best runner-up in the area where that party’s voters were the most under-represented; either a multi-MP district, or an aggregation of single-MP districts. 

Two-vote option

Or they could be elected from a regional ballot, where voters have two votes, as in MMP.

The three top-up regional MPs would, just like open-regional-list MMP, be the regional candidate of the party who got the most votes (after skipping over candidates who had already won a local seat).

One way to limit the size of the ballot, and eliminate the risk of a party trying to copy the Berlusconi trick (running a twin "decoy" list of regional candidates), would be to copy Bavaria and put only local candidates on the regional ballot.

How would the Swedish option actually work?

Take Saskatchewan. It has 14 MPs. Four will be elected from the Saskatoon region, three from the Regina—Moose Jaw region. Five single-MP ridings are only about 17% larger than today. That lets Saskatchewan have two regional MPs.

On the votes cast in 2015, the Conservatives win two seats in the Saskatoon district, the Liberals and NDP one each. Each party wins one seat in the Regina—Moose Jaw district. The NDP still wins the northern single-MP district, and the Conservatives win the other four. The two parties under-represented are the Liberals and NDP. Therefore, they each get a regional (provincial) MP.

The Liberals are most under-represented in the group of five single-MP ridings. Their best candidate in those ridings was aboriginal leader Lawrence Joseph, who came close to winning in the Northern riding of Desnethé—Missinippi—Churchill River.

The NDP are most under-represented in the Saskatoon Region. If current MP Sheri Benson still won the first NDP seat there, those voters elect a second Saskatoon Region MP such as Claire Card.

Lawrence Joseph and Claire Card have dual mandates, although this is less obvious than under MMP. They were each ranked as a near-winner by voters in their district, but were elected thanks to voters in the whole of Saskatchewan. So they are Regional MPs (or in this case Provincial MPs). They would make themselves available to voters across the region not already served by a local MP of that party. They would be allowed to open satellite offices where needed across the region, just as Conservative MP Robert Kitchen has offices today in Estevan, Weyburn and Moosomin. Just the way it's done in Scotland.

It is proportional

On a simulation from the votes cast in 2015, the total results are (compared with perfect province-wide proportionality) Liberal 136 (137), Conservative 109 (109), NDP 69 (67), Bloc 15 (15), Green 9 (10). Little more than the inevitable rounding anomalies.

But how robust is this model? On a simulation from the votes cast in 2011 (transposed to the 2015 ridings), the total results are (compared with perfect province-wide proportionality) Conservative 146 (140), NDP 100 (103), Liberals 62 (64), Bloc 18 (18), Green 12 (13). This model is robust enough to give almost perfectly proportional Quebec results despite the Orange wave of 2011 (a Conservative bonus of 1). In a few regions of Ontario and BC a local Conservative sweep creates small bonuses.  

If the Green vote doubled (from non-voters, and except in Saanich—Gulf Islands), can it handle such a shift? The results are (again compared with perfect proportionality) Liberal 132 (130), Conservative 107 (105), NDP 65 (65), Bloc 14 (16), Green 20 (22). Again, the differences are rounding anomalies.

Would the Green regional MPs all come from the large-urban areas? In my simulation with the doubled Green vote, the 16 regional Green MPs are ten from large-urban areas, and six from single-member districts, added to the five directly elected from multi-member ridings.   

We have used only six 2-MP ridings across Canada, and only when necessary, since voters generally want to have representation from more than two parties.

Accountable and proportional

The result of the simulation shows that this model lets voters elect MPs who are both highly accountable to the small regions or local ridings that voted for them, and very proportional to the votes cast.

Summary

The 86 present ridings become 76 single-MP ridings. The 249 larger-urban MPs become 211 MPs from multi-MP ridings.

The large-urban multi-MP ridings could use any good PR system. (The Swedish open-list system, STV, or Dion’s P3.) For the simulation we used the Swedish system because it is the simplest calculation (it requires no assumptions about second choices).

The 76 single-MP ridings use a winner-take-all ballot: FPTP or AV.

This model lets communities under 200,000, and almost all medium-urban, small-urban and rural communities, keep their local MPs in ridings only about 17% larger than today, not 60% larger. This will not only please those voters, it will reassure many current MPs.

The seat totals are the same under the Swedish option or the two-vote option. Those options change who fills the seats, not the party totals.

Regions all keep their number of MPs

The regions will ensure that, for example, only Northern Ontario votes help elect Northern Ontario MPs, GTA votes elect GTA MPs, BC Lower Mainland voters elect Lower Mainland MPs, and all regions keep their present numbers of MPs.

Will voters in big cities accept multi-MP districts?

Take Toronto. The multi-MP districts would not include the whole megacity. They might be six MPs from the area served by the “Community Council” for Toronto and East York, five from each of the others (North York, Scarborough, and Etobicoke—York). If the GTA’s 55 MPs are in one “top-up region” it has eight regional MPs. So, no MP specifically from Beaches—East York. But a candidate who was already, for example, a ward councillor might get enough local support, added to support from the rest of the larger district, to elect him or her as one of those six MPs.

The BC Lower Mainland might have districts ranging from 3 MPs to 7 MPs.

Option: A bigger House 

If MPs prefer, these 51 additional MPs could be added to the House by expanding its size, so that the 338 present ridings would continue either as multi-MP ridings or single-MP ridings. It has often been said that "it is easier to have a more representative system if you have more representatives."

What would the map look like?

Of course, the actual multi-MP districts and single-MP districts would be set by the normal Boundaries Commissions process.

Still, here’s the map I used for my simulation.

A group of seven present ridings becomes (in large urban areas) a 6-MP district, or (in other areas) an aggregation of six single-MP ridings. Eight becomes seven, nine becomes eight, six becomes five, five becomes four, and four becomes three.

These multi-MP districts are from metropolitan areas with at least 200,000 residents, sometimes including adjacent communities. They also include, of necessity, St. John’s, Charlottetown, Barrie, Kelowna, Sudbury, Saguenay, Trois-Rivières, Moncton, Saint John, and Thunder Bay. 

Newfoundland & Labrador (2+4+1):
St. John's 2                                      
Singles 4
Nova Scotia (4+5+2):
Halifax 4
Singles 5
PEI (2+1+1)                                                                              
Eastern PEI 2
Single 1
New Brunswick (6+3+1):
Saint JohnFredericton 3      
Moncton—BeauséjourMiramichi 3
Singles 3
Eastern QuebecMauricieCentre-du-Quebec 25 (13+8+4)     
Quebec City 6                                                          
Saguenay-Lac-Saint-Jean 2                                 
Lévis—Lotbinière 2           
Mauricie 3
Centre-du-QuebecChaudière-AppalachesBas-Saint-Laurent—Gaspésie—Côte-Nord singles 8
Montreal-estMontérégieEstrie  27 (20+3+4)
Montreal-est 5                                                         
Montreal-nord 4                                                       
Longueuil--Richelieu 6                                                   
Roussillon—Suroît 3                                                            
Sherbrooke 2 
Singles 3
Quebec-ouest (Montreal-West, Laval-Lanaudiere-Laurentides and west) 26 (19+3+4)
Montreal West 6 
Laval 3 
Lanaudière 3
Laurentides 4
Outaouais 3                            
Singles 3
Eastern Ontario (19: 8+8+3)
Ottawa-East—Prescott-Russell—Cornwall 4 (the bilingual district)
Ottawa West 4 
Singles 8 (from Renfrew and Leeds—Grenville to Haliburton—Kawartha Lakes)
Greater Toronto Area 55 (47+8)
Toronto and East York 6
Etobicoke—York 5
North York 5
Scarborough 5
Vaughan—Richmond Hill 4
Markham—Aurora—Newmarket—Georgina 4
Durham Region 5
Brampton—Caledon 5
Mississauga 5
Oakville—Halton 3
South Central Ontario (Hamilton—Niagara—Waterloo) 19 (12+4+3)
Hamilton (incl. Burlington) 5                                  
Niagara Region 3
Waterloo Region 4  
Singles 4
Western Ontario (Barrie--London—Windsor) 18 (11+4+3)
BarrieSimcoe 3       
London—Elgin 4
Windsor-EssexChatham-Kent 4
Singles 4
Northern Ontario 10 (5+4+1)
SudburySault Ste. Marie region 3                
Thunder Bay region 2                     
Singles: 4
Manitoba 14 (7+5+2)
Winnipeg 7
Singles 5
Saskatchewan 14 (7+5+2)
Saskatoon 4 (incl. N½ of Moose Jaw—Lake Centre—Lanigan)
Regina—Moose Jaw 3
Singles 5
South Alberta 17 (9+5+3) 
Metropolitan Calgary incl. Banff—Airdrie 5 (North) & 4 (South)
South-Central Singles 5
North Alberta 17 (9+5+3)
Metropolitan Edmonton 5 (North) & 4 (South)
North Singles 5
British Columbia Lower Mainland 26 (22+4)
Vancouver and Vancouver North and West 7
Surrey—Richmond—Delta 7 
Burnaby--Maple Ridge 5
Fraser ValleyLangley East 3

BC Interior and North and Vancouver Island 16 (8+6+2)
Kelowna—Okanagan 3
VictoriaNanaimo 5
Singles 6
Territories 3 unchanged