Over the past year, a bipartisan commission embarked on a cross-province tour, tasked with consulting on where, exactly, it should draw lines on a map.
This group, Alberta’s electoral boundaries commission, re-forms every eight to 10 years.
Its mission is to sketch out which groups of people should vote together to elect each member of Alberta’s Legislative Assembly.
The latest commission’s findings were eventually delivered in March in a detailed report running several hundreds of pages long.
But that, of course, is not where the story ended.
The five-member commission’s final report included, for the first time, two very different maps from the majority and the minority members of the commission.
The boundary commission’s chair, Justice Dallas Miller, suggested the legislature should adopt the majority report of the commission.
But if it didn’t, Miller said it should increase the number of seats in the legislature to 91 from 89. The Opposition NDP have suggested anything short of adopting the majority report would constitute “election rigging,” while the UCP have said its goal is to preserve rural representation.
With that representation in mind, the UCP formed an all-party committee of the Alberta legislature to redraw the boundaries once again.
Why the process of redrawing electoral boundaries in Alberta is heating up
That committee’s work hit a snag last week after Alberta’s acting chief justice declined an all-party legislative committee’s request for help finding retired and sitting judges to lead the latest effort to redraw the boundaries.
So the debate is far from over and is certain to animate Alberta’s legislature for months to come, in advance of an expected report in November.
But in the meantime, what Albertans have is the report submitted by the commission — and its examples of the electoral impacts that play out depending on which line goes where.
To better understand the reports, CBC News’ elections unit and data journalists, including the CBC’s Naël Shiab and Meg Banks, conducted what’s known as a transposition of votes. In essence, it’s an exercise that seeks to estimate how a previous general election would’ve looked had it been held with the new boundaries.
It’s a routine exercise that team undertakes whenever there is a redistribution in the country, in order to compare apples to apples on election night.
This doesn’t mean that the next election will deliver these results. But some voting habits are long-standing, and examining proposed boundaries this way illustrates the potential impact of changing lines.
The previous election in Alberta saw the United Conservative Party winning 49 seats to the Alberta New Democrats’ 38.
Using boundaries proposed under the majority report, submitted by Miller and NDP appointees Greg Clark and Susan Samson, the UCP would have won as many as 48 seats.
Under maps drawn by minority commissioners Julian Martin and John Evans, the UCP would have won as many as 56 seats.
Many of the changes play out in “battleground Calgary.”
In 2023, the NDP took 14 seats to the UCP’s 12, some by extremely tight margins. For instance, after a recount, Calgary-Acadia was won by the NDP by only 22 votes.
Because the city was so closely contested — of the city’s 26 ridings, half had a margin of victory of less than 10 per cent — the transposition shows that moving lines around had the effect of flipping multiple ridings, in both directions.
In 2023, the UCP carried Calgary-Cross by 3.4 percentage points. The majority map changes boundaries slightly, making it a notional NDP win by 4.8 percentage points.
Calgary-Edgemont was won by the NDP in 2023 by 1.2 percentage points. The minority map would shift boundaries, leading to a notional UCP win by 0.4 percentage points.
Under the majority’s report, two seats would be added to Calgary, lifting the city’s total seats to 28. Those two seats would go to the NDP.
The minority’s report, meanwhile, divides Calgary into 29 ridings, and the UCP would have carried most of them — 18 to the NDP’s 11 — most of them being so-called “rurban” or hybrid ridings.
The province changed the rules for this iteration of the commission, no longer requiring it to align ridings with municipal boundaries. The minority report would extend 11 ridings outside of the boundaries of Calgary, compared to the last election.
Take Calgary-Foothills, won by the NDP in 2023 by 261 votes. The minority’s report extends that riding north into Airdrie, renaming it Calgary-Foothills-Airdrie West, and leading to a notional UCP win by eight points.
Calgary-Nolan Hill-Cochrane, another new hybrid riding, takes a slim part of Calgary-Foothills and expands to the west farther than Cochrane. It becomes a UCP riding, too.
The majority report also includes four hybrid ridings.
Moving now to Alberta’s capital city, considered an NDP stronghold. In 2023, the party swept all of the city’s 20 seats.
The majority report would add one net seat to Edmonton, which the transposition shows would go to the NDP.
The minority report would add two hybrid ridings to the city’s west.
The minority report would take part of Edmonton-South West, won by the NDP in 2023 by 14.3 percentage points, and draw a new riding west into the old Drayton Valley-Devon riding.
Under its new name, Edmonton-Enoch-Devon, the UCP would win the riding by 148 votes, or 0.92 percentage points (the closest margin in the city), the transposition shows.
All told, the two added ridings in Edmonton in the minority proposal would potentially see two seats added to the UCP’s tally, although one (Edmonton-Enoch-Devon) would be a toss-up.
The UCP has historically dominated in the rural parts of the province. In the 2023 election, the UCP carried 37 of the 41 seats outside the two major cities, including Airdrie, Red Deer, Fort McMurray and Grand Prairie.
If the 2023 election were held with the majority report’s boundaries, the NDP would win St. Albert, but the margin would slightly tighten, the transposition found. Margins in Sherwood Park and Lethbridge-West would also tighten, but would remain close to the last election.
Banff-Kananaskis, meanwhile, would be renamed Canmore-Banff, and would become much more favourable to the NDP.
Overall, the proposed majority map wouldn’t change the seat count outside the big cities from 2023.
Under the minority report’s boundaries, the 2023 election results play out differently.
The NDP would still have won St. Albert, while Sherwood Park would remain a close race.
But Banff-Kananaskis, an NDP win by 303 votes in 2023, would see its boundaries moved south. The northern part of the riding would join Rimbey-Rocky Mountain House-Sundre and become Rocky Mountain House-Banff Park, and the southern part of the riding would be named Canmore-Kananaskis. The UCP would win both.
A big change would come to Lethbridge. In 2023, the NDP won Lethbridge-West by more than 11 percentage points, while the UCP won Lethbridge-East by just 2.9 percentage points.
The minority report would see the city redrawn into four large ridings: Lethbridge-Cardston, Lethbridge-Fort Macleod-Crowsnest Pass, Lethbridge-Little Bow and Lethbridge-Taber-Warner.
The transposition shows the UCP would pick up all four, and the closest margin in any of the four new ridings is 17.8 percentage points, in Lethbridge-Little Bow.
In Red Deer, the seats don’t change hands, but the map becomes less competitive.
In 2023, the UCP won Red Deer North by 18.9 percentage points, and Red Deer South by 14.6 points.
Those margins remain the same in the majority map as the ridings are largely unchanged.
But under the minority’s proposed boundaries, Red Deer, like Lethbridge, is drawn into quarters that expand into rural Alberta. Of the proposed Red Deer-Blackfalds, Red Deer-Innisfail, Red Deer-Lacombe, and Red Deer-Sylvan Lake, the tightest margin is in Red Deer-Innisfail, where the UCP would win by more than 28 points.
Keith Archer, who served on a previous Alberta commission and was chief electoral officer in British Columbia, spoke to CBC News after the release of the initial majority and minority reports.
He said the minority report, authored by the government-appointed members, reintroduced partisanship into a process that historically has been depoliticized from legislative assemblies to more of an administrative function.
“[If the government adopts] something that looks like the minority report, I think that’s very unfortunate. It’ll be a disservice to the commission as a whole,” Archer said, “and to Albertans as a whole, who I think are looking for an opportunity to adjust electoral boundaries in a way that doesn’t provide a kind of explicit partisan advantage to one party in the legislative assembly over another.”
Methodology
Transposition performed by Naël Shiab.
Here’s how we did it.
Previous electoral results:
We used the detailed 2023 general election results. For each district, we have the number of votes for a small subset of voting areas. For example, Zone A for District 1 had voting areas 001, 002 and 003 assigned. Unfortunately, results at the voting area level are not available.
To produce results at the voting area level, we redistributed the votes from a given zone to each of its assigned voting areas. For instance, if Zone A had 99 votes for party X, we split them between 001 (33 votes), 002 (33 votes) and 003 (33 votes). We did this for all districts, zones, parties and voting areas.
For advance votes, mobile votes and special ballots, we kept them at the district level since they do not have specific voting areas assigned.
Previous electoral boundaries:
We used the 2023 voting area boundaries. We used the 2019 electoral district boundaries. We converted them to the WGS84 projection. We clipped both geospatial datasets using the 2020 Land Cover of Canada dataset to focus only on urbanized areas. The rationale is that we do not want to redistribute votes to areas where no one lives.
We then merged the boundaries with the corresponding voting area results. When no voting area results were available, we merged them at the district level.
Redistributing the votes:
For the EBC and minority maps, we calculated the intersection of each proposed district with the general election voting areas and districts. We then redistributed the votes of a given voting area or district to a new proposed district proportionally. For example, if 20 per cent of a voting area’s surface was within a new district, we assigned 20 per cent of the votes for each party to that new district.
Final results:
We then summed all the redistributed votes per party for the EBC and minority maps. We determined the winning party for each district in both cases. From there, we were able to count the number of MLAs for each map.
