On 10-20 I published two partial-count snapshots of Alberta's referendum (earlier piece) and wrote that every one of the nine non-separation questions had moved toward Yes between them. Later counts changed that picture. Unofficial totals are still moving, and official results are due 10-29.
Question 10: it moved away from the poll
I compared the count with the CBC/Janet Brown poll (72 remain, 23 Option 2, 4 unsure), which is 75.8% / 24.2% with the unsure dropped. Last time I called the exact match at 5:10 a.m. a coincidence of a partial count. It was.
| Snapshot | Remain | Option 2 | |---|---|---| | ~3:00 a.m. EDT 10-20 (Vista Radio) | 76.3% | 23.7% | | 5:10 a.m. 10-20 (Calgary Citizen) | 75.8% | 24.2% | | 7:30 a.m. 10-20 (Elections Alberta table, as copied by Todayville): 1,379,884 vs 447,271 | 75.5% | 24.5% | | Late Tuesday (Global News, about 1.9 million ballots on Q10) | "roughly 75%" | "nearly half a million" |
Option 2 rose at every step. The poll's decided split understated it by roughly a point or less at this stage, which is inside the poll's stated ±2.8 points. I computed the 7:30 shares from the table's raw counts.
Where Option 2 votes came from
On 10-16 I wrote that, by my arithmetic, most Option 2 support in the CBC poll sat among UCP voters, and that I couldn't test this because Elections Alberta doesn't report by party. The Globe and Mail has now compared district results with the 2023 outcome. This is not a test of individual voters, but it points the same way:
- Two of 87 divisions voted for Option 2: Rimbey-Rocky Mountain House-Sundre and Central Peace-Notley, both at 51% and both represented by UCP cabinet ministers.
- The next 42 divisions by Option 2 share were all UCP-won in 2023.
- Divisions the UCP won in 2023 averaged 33% Option 2. NDP-won divisions averaged about 15%.
- Edmonton and Calgary were at roughly 86% and 83% remain, and the rest of the province at 60%.
A district average can't tell you which voters in a district chose Option 2. The Globe's own text only says where the votes were, not who cast them. A separate Substack/Todayville analysis I read (Yakiwchuk) also found Option 2 roughly double in UCP-held constituencies, but that is a commentator's tally, and I haven't checked it.
The nine other questions: the lean faded
Percent Yes of votes counted on each question. The first two columns are from my last piece. The 7:30 column is computed from Elections Alberta's table as copied by Todayville. The Wednesday column is from The Hub, which said Q1 was about 90% counted and the others 63–90% complete. The Hub gave a single range for Q6–9 (48.5–52.8), so I left those cells blank, and a Hub column by Trevor Tombe puts Q8 at 48.6.
| Question | 5:10 a.m. | 7:30 a.m. | Wed. (The Hub) | |---|---|---|---| | 1. More control over immigration | 58.6 | 55.9 | 53.4 | | 2. Programs for citizens, PRs, approved status | 57.2 | 56.6 | 51.9 | | 3. 12-month residency for supports | 58.4 | 58.4 | 54.9 | | 4. Fees for health care and education | 57.1 | 57.1 | 55.2 | | 5. Proof of citizenship to vote | 63.1 | 63.2 | 62.4 | | 6. Provinces select judges | 50.9 | 51.2 | within 48.5–52.8 | | 7. Abolish the Senate | 52.5 | 53.2 | within 48.5–52.8 | | 8. Opt out of federal programs | 48.3 | 48.7 | 48.5 / 48.6 | | 9. Provincial laws prioritized | 51.5 | 50.6 | within 48.5–52.8 |
What this shows:
- Questions 1 to 4 lost 1 to 5 points of Yes as the count went on. Question 2 fell about five. Question 5 barely moved.
- So "every question moved toward Yes" was true only for the first two snapshots. I read a trend off the early part of a count that had a known order (Q10 first, then Q1 to 9) and early returns that were not representative. The CBC reported that the early results suggest many rural Albertans support the immigration policies while the two big cities are more divided; urban ballots arriving late would fit that. That is an explanation I haven't verified.
- Question 8 is the only one that failed in every account I read. The Globe has it at 51% No, and The Hub's analysis puts it at 48.6% Yes to 51.4% No. Question 9, which flipped to Yes in my first piece, has stayed within a point and a half of 50% on the snapshots I could compute.
- The Hub's analysis says Yes led on eight of nine, and that the yes margins were thin. Premier Smith says she has a clear mandate on separation, which is a claim about Question 10 and not about these.
Conflicts and gaps
- The Hub's analysis says the constitutional questions ranged from 48.5% to 52.8%, and also says support for constitutional change "sits no higher than 54 percent." Both can hold, but I'd go by the first, since it is a range from the same count.
- The Toronto Star analysis I cited last time called all nine questions majorities for change while calling Question 8 a close fight. The count above has Question 8 below 50% at every snapshot I found, so I'd take the Star line as inaccurate on that point. I only saw an excerpt of it.
- Elections Alberta's results page loaded for me again as an empty template, so the 7:30 column is a third party copying the official table. The time zone is unlabelled in some of the sources, and I assumed Alberta time for the 5:10 and 7:30 snapshots.
- I haven't seen a final count for any question. Mobile ballots were 209 of 326 locations at 7:30. I also haven't read the Elections Alberta statement on when counting finished.
What comes next
- Official results: 10-29.
- Smith said in her Tuesday address that she will not hold another separation referendum while she is premier, and that she would act on immigration reforms over coming months. Per The Hub, much of that work may wait until after the 2027 election.
- Whether the final Q1–9 numbers settle near the Wednesday ones or keep falling.
Sources
Elections Alberta table as copied by Todayville (7:30 a.m. 10-20); Global News, Smith's address (10-20); CBC, preliminary counts (10-21); The Globe and Mail, "by the numbers" (10-21); The Hub, analysis of the nine questions and a Trevor Tombe column (both 10-21); Calgary Citizen and Vista Radio snapshots from my earlier piece.