59%
Mark Carney government
Abacus Data · Oct 21, 2026
Original poll (opens in a new tab)CANADA / FEDERAL VOTING INTENTIONS
The public pulse.
Individual polls tell part of the story.
See how the reviewed numbers come together.
AVERAGE AS OFNov 6, 2026Experimental average · v1
01 / POLITYIQ AVERAGEAS OF NOV 6, 2026
The national picture.
15.4 point leadEstimated share among decided / leaning voters
Other combines PPC and other parties so releases with different party groupings can be compared. Values may not add to 100% after display rounding. This combined group has no single leader or party logo.
Portrait & logo sources
Leader identities reviewed October 22, 2026. Portraits are cropped from official party or leader websites; official marks retain their original artwork.
- Liberal leader (opens in a new tab) · Photo (opens in a new tab) · Logo (opens in a new tab)
- Conservative leader (opens in a new tab) · Photo (opens in a new tab) · Logo (opens in a new tab)
- NDP leader (opens in a new tab) · Photo (opens in a new tab) · Logo (opens in a new tab)
- Bloc Québécois leader (opens in a new tab) · Photo (opens in a new tab) · Logo (opens in a new tab)
- Green leader (opens in a new tab) · Photo (opens in a new tab) · Logo (opens in a new tab)
BEHIND THE AVERAGEWhose polls count?
Whose polls count?
And how much?
Newer fieldwork carries more weight. Larger total samples receive a capped boost. Each firm contributes its latest eligible release.
Abacus Data43.2%
Liaison Strategies28.9%
Léger27.9%
Contribution to this snapshot’s average. These are model weights, not pollster accuracy ratings.
CANADA / BEYOND THE BALLOT
Confidence & leadership.
Follow the government, the prime minister and the people who want the job.
LATEST SOURCE CHECKOct 25, 11:34 a.m. PDTSee coverage
64%Mark Carney
Liaison Strategies · Oct 17, 2026
Original poll (opens in a new tab)61.3%Mark Carney
Nanos Research · Oct 16, 2026
Original poll (opens in a new tab)Latest collected national polls, shown separately by measure. Dates show fieldwork end. Expand a row for every response, the respondent base and original source. Approval, satisfaction, favourability and preferred PM measure different things.
24 collected measurements · page 1 of 2Download these measures
Federal voting intentions · CANADAAbacus DataFederal party supportn = 1,500LPC48%CPC32%NDP9%+4 responsesExpand full results and survey details
LPC48%
CPC32%
NDP9%
BQ6%
GPC3%
OTH1%
PPC1%
Question summary: Federal party choice if an election were held now (summary)
Respondent base: Committed voters; undecided removed
Method: Online · blended double-opt-in partner panels
Source checked: Vote intention interactive: Overall table; article methodology. Released October 25; fieldwork ends October 21.
Publisher release and interactive tables checked. Total survey n=1,500; separate unweighted question base not published. Weighted to census population benchmarks. Comparable probability-sample margin of error is not a design-based margin for this online panel.
Original release (opens in a new tab)Source checked
Government approval · CANADAAbacus DataMark Carney governmentn = 1,500Approve59%Disapprove26%Neither approve nor disapprove15%Expand full results and survey details
Approve59%
Disapprove26%
Neither approve nor disapprove15%
Question summary: Overall approval of the federal government led by Mark Carney (summary)
Respondent base: Canadian adults; overall government job approval
Method: Online · blended double-opt-in partner panels
Source checked: Government Approval: Overall table. Published aggregate approve=59; rounded strongly/somewhat components total 58, not substituted for the aggregate.
Publisher release and interactive tables checked. Total survey n=1,500; separate unweighted question base not published. Weighted to census population benchmarks. Comparable probability-sample margin of error is not a design-based margin for this online panel.
Original release (opens in a new tab)Source checked
Leader favourability · CANADAAbacus DataMark Carneyn = 1,500Positive56%Negative27%Neutral15%+1 responsesExpand full results and survey details
Positive56%
Negative27%
Neutral15%
Do not know enough2%
Question summary: Positive or negative impression of Mark Carney (summary)
Respondent base: Canadian adults; personal impression
Method: Online · blended double-opt-in partner panels
Source checked: Leader Impressions interactive: Overall Comparison card for Mark Carney.
Publisher release and interactive tables checked. Total survey n=1,500; separate unweighted question base not published. Weighted to census population benchmarks. Comparable probability-sample margin of error is not a design-based margin for this online panel.
Original release (opens in a new tab)Source checked
Leader favourability · CANADAAbacus DataPierre Poilievren = 1,500Negative45%Positive37%Neutral14%+1 responsesExpand full results and survey details
Negative45%
Positive37%
Neutral14%
Do not know enough4%
Question summary: Positive or negative impression of Pierre Poilievre (summary)
Respondent base: Canadian adults; personal impression
Method: Online · blended double-opt-in partner panels
Source checked: Leader Impressions interactive: Overall Comparison card for Pierre Poilievre.
Publisher release and interactive tables checked. Total survey n=1,500; separate unweighted question base not published. Weighted to census population benchmarks. Comparable probability-sample margin of error is not a design-based margin for this online panel.
Original release (opens in a new tab)Source checked
Leader favourability · CANADAAbacus DataAvi Lewisn = 1,500Neutral32%Negative27%Do not know enough24%+1 responsesExpand full results and survey details
Neutral32%
Negative27%
Do not know enough24%
Positive17%
Question summary: Positive or negative impression of Avi Lewis (summary)
Respondent base: Canadian adults; personal impression
Method: Online · blended double-opt-in partner panels
Source checked: Leader Impressions interactive: Overall Comparison card for Avi Lewis.
Publisher release and interactive tables checked. Total survey n=1,500; separate unweighted question base not published. Weighted to census population benchmarks. Comparable probability-sample margin of error is not a design-based margin for this online panel.
Original release (opens in a new tab)Source checked
Federal voting intentions · CANADAAbacus DataFederal party supportn = 3,197LPC46%CPC34%NDP9%+3 responsesExpand full results and survey details
LPC46%
CPC34%
NDP9%
BQ7%
GPC3%
OTH1%
Question summary: Federal party choice if an election were held now (summary)
Respondent base: Committed voters; undecided removed
Method: Online · partner panels
Source checked: National figures, total sample and fieldwork checked against the publisher's overall table. Decided sample size not supplied.
Only the latest release from this firm enters the average. PPC 1%; other parties round to 0% in the published table.
Original release (opens in a new tab)Source checked
Government approval · CANADAAbacus DataMark Carney governmentn = 3,197Approve57%Disapprove29%Neutral14%Expand full results and survey details
Approve57%
Disapprove29%
Neutral14%
Question summary: Approval of the job of the Carney-led federal government (summary)
Respondent base: All Canadian adult respondents
Method: Online · partner panels
Source checked: Article: Government Approval Falls Three Points to 57%; sample/methodology sections.
Published 29% disapproval retained despite rounded components summing to 30%.
Original release (opens in a new tab)Source checked
Leader favourability · CANADAAbacus DataMark Carneyn = 3,197Positive55%Negative29%Neutral13%Expand full results and survey details
Positive55%
Negative29%
Neutral13%
Question summary: Overall impression of the named federal leader (summary)
Respondent base: All Canadian adult respondents
Method: Online · partner panels
Source checked: Article: Carney’s Personal Rating Remains Strong, but Has Slipped from Its September High.
Only explicitly published numeric categories are included; missing categories are not inferred. This is impression, not job approval.
Original release (opens in a new tab)Source checked
Leader favourability · CANADAAbacus DataPierre Poilievren = 3,197Negative44%Positive37%Expand full results and survey details
Negative44%
Positive37%
Question summary: Overall impression of the named federal leader (summary)
Respondent base: All Canadian adult respondents
Method: Online · partner panels
Source checked: Article: Carney’s Personal Rating Remains Strong, but Has Slipped from Its September High.
Only explicitly published numeric categories are included; missing categories are not inferred. This is impression, not job approval.
Original release (opens in a new tab)Source checked
Leader favourability · CANADAAbacus DataAvi Lewisn = 3,197Negative29%Not enough information23%Positive18%Expand full results and survey details
Negative29%
Not enough information23%
Positive18%
Question summary: Overall impression of the named federal leader (summary)
Respondent base: All Canadian adult respondents
Method: Online · partner panels
Source checked: Article: Carney’s Personal Rating Remains Strong, but Has Slipped from Its September High.
Only explicitly published numeric categories are included; missing categories are not inferred. This is impression, not job approval.
Original release (opens in a new tab)Source checked
Government approval · CANADAAbacus DataFederal party supportn = 2,744Approve60%Disapprove25%Neither / neutral15%Expand full results and survey details
Approve60%
Disapprove25%
Neither / neutral15%
Question summary: Approval of the job of the Carney-led federal government (summary)
Respondent base: All Canadian adult respondents.
Method: Online; partner panels via PureSpectrum
Source checked: Overall Government Approval section, combined approve/disapprove/neither totals; methodology section.
Historical backfill from the publisher's national tables. Fieldwork dates taken from methodology, publication date from article dateline. Published rounded components retained; no missing response inferred.
Original release (opens in a new tab)Source checked
Leader favourability · CANADAAbacus DataMark Carneyn = 2,744Positive56%Negative25%Neither / neutral17%+1 responsesExpand full results and survey details
Positive56%
Negative25%
Neither / neutral17%
Don't know enough to say2%
Question summary: Overall impression of Mark Carney (summary)
Respondent base: All Canadian adult respondents.
Method: Online; partner panels via PureSpectrum
Source checked: Overall/Comparative Party Leader Impressions, named leader's four component percentages; methodology section.
Historical backfill from the publisher's national tables. Fieldwork dates taken from methodology, publication date from article dateline. Published rounded components retained; no missing response inferred. The release notes differences between printed net scores and rounded components/trend endpoints. Only directly published positive, negative, neutral and unknown components are stored, not net scores.
Original release (opens in a new tab)Source checked
Collection coverage 12 / 34 firms successfully checked in latest recorded attempts ↓
The collector researches original releases on a schedule. The list includes every firm in our historical registry; a listed firm is not proof of a working source. 2 conflicting records held for review. This collection is growing and does not claim every Canadian poll has been captured.
| Polling firm | Latest attempt | Source status | Coverage note |
|---|---|---|---|
| Abacus Data | Oct 25, 11:34 a.m. PDT | Checked | October 25 release reviewed in original publisher page and interactive Overall, Subgroups and Methodology tabs: October 16–21, n=1,500. Added national vote, government approval and three leader impressions. Regional vote percentages are visible but separate unweighted regional question bases are not published in the reviewed tables; held out of model. No PM job approval question substituted from personal impressions. |
| Angus Reid Group | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Angus Reid Institute | Oct 23, 7:18 p.m. PDT | Checked | Read federal index and linked latest 'Trump Bump' release. No newer qualifying wave found. Prior vote-base/date holds remain unresolved. |
| Campaign Research (Ontario PC internal) | Oct 22, 9:40 p.m. PDT | Needs review | Official insights index reviewed; visible federal reports predate the two-year window. Discovered official domain is not in the approved writer registry; no source-code change made. |
| Canada Pulse Insights | Oct 23, 7:18 p.m. PDT | Checked | Read insights index and linked Oct18 economic-changes article. Latest Toronto/BC material excluded; economic-plan preference not substituted for general approval. No new qualifying federal wave found. |
| Cardinal Research | Oct 22, 9:40 p.m. PDT | Unavailable | Homepage returned HTTP 502. No data published. |
| Deltapoll | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| EKOS | Oct 23, 7:18 p.m. PDT | Unavailable | Original index returned406. Release-level coverage unavailable; accepted data retained. |
| Empirical Intelligence | Oct 22, 9:40 p.m. PDT | Unavailable | Official domain was not accessible through the research reader. No data published. |
| Environics Institute | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Focaldata | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Forum Research | Oct 22, 9:40 p.m. PDT | Unavailable | Official polling index returned HTTP 502. No data published. |
| Innovative Research | Oct 23, 7:18 p.m. PDT | Checked | Read homepage/latest-research list: Oct20 BC provincial and Oct16 Alberta referendum. No newer qualifying federal wave found in index review. Prior August fieldwork conflict unresolved. |
| Ipsos | Oct 23, 7:18 p.m. PDT | Checked | Read politics index, Dec16 2025 release and linked tables (pages2–3 visually checked). Backfilled seven vote and three regional government-approval records with explicit unweighted bases. Suppressed/dash values omitted. Newly discovered Oct23 release is BC provincial and excluded. |
| JL Partners | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Kolosowski Strategies | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Liaison Strategies | Oct 23, 7:18 p.m. PDT | Checked | Read press index and latest Oct19 federal article. No newer eligible wave listed. Latest table read failed at storage.ghost.io redirect, so regional/question bases remain unverified; no regional figures inferred. |
| Léger | Oct 23, 7:18 p.m. PDT | Checked | Read homepage and linked Oct16 federal release, matching saved Oct9–12 national/regional wave. Newer listed releases concern BC/Quebec provincial politics. No newer qualifying federal wave discovered. |
| MQO Research | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Mainstreet Research | Oct 23, 7:18 p.m. PDT | Checked | Read original polling index: latest municipal/BC provincial/Alberta referendum items excluded. No new eligible federal release found. Index review only; subscriber reports not accessed. Prior Oct5 Ontario federal date/base hold retained. |
| Nanos Research | Oct 23, 7:18 p.m. PDT | Checked | Read latest-research list; latest federal release remains Oct20, already stored with unknown fieldwork start. Newer Toronto municipal items excluded. No newer eligible federal wave found. |
| Narrative Research | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Oracle Poll | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Pallas Data | Oct 23, 7:18 p.m. PDT | Checked | Read official polls index: latest Mississauga/Toronto/BC/Quebec releases municipal or provincial. No new federal wave identified in index review. Existing May numeric/method-date ambiguity retained. |
| Pollara | Oct 23, 7:18 p.m. PDT | Checked | Read latest-insights listing: Oct23 BC provincial, Oct20 Toronto and Oct17 Alberta referendum excluded. No newer general federal poll identified in index review. Prior historical date ambiguity retained. |
| Probe Research | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Relay Strategies | Oct 22, 9:40 p.m. PDT | Unavailable | Publisher index returned HTTP 429; transient rate limit. Existing data retained. |
| Research Co. | Oct 23, 7:18 p.m. PDT | Checked | Read homepage: latest BC provincial/hunting/Alberta items excluded. No new qualifying federal wave identified in index review. No issue-specific approval substituted for general approval. |
| Rubicon Strategy | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Segma Research | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Spadina Strategies | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Spark Insights / Advocacy | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
| Synopsis Research | Oct 22, 9:40 p.m. PDT | Needs review | Registry source remains unmapped. Publisher-domain discovery search did not verify an original release; no figures ingested. This URL identifies the registry, not a primary poll source. |
| YouGov | Oct 22, 9:40 p.m. PDT | Needs review | Official homepage reviewed. No qualifying recent federal release verified in this pass; further report-index/history review remains pending. |
Checks are research attempts, not guarantees of completeness. Inaccessible or unclear reports remain pending. Existing accepted figures survive failures. Duplicate releases do not add extra weight; conflicting revisions are held. The headline vote average recalculates from eligible collected records. Approval and leadership measures remain individual polls until a comparable, validated model is available.
Source data & status JSONCANADA / TWO YEARS OF FEDERAL POLLINGThe polls.
The polls.
The full context.
October 22, 2024 – October 22, 2026. Search the archive, compare firms and inspect the source behind each record.
547national poll records31 firms across all geographies
Canada · national
0 matching records · 0 firmsDots are individual releases; lines are an experimental descriptive trend. No confidence bands or seat projections are implied. Reference-date records appear as dots but do not enter the trend. Values above 60% remain in the table. Select a dot for its source and metadata. Flagged records remain in the ledger but are excluded from chart dots and lines.
The poll ledger.
Download matching CSVPercentages as published or transcribed. A dash means not separately reported. The darkest cell marks the highest listed party support, not a winner call. Dates labelled “reference” may be survey midpoints. Open a firm name for full metadata.
| Polling firm · date · n | Source check |  LPC |  CPC |  NDP |  BQ |  GPC | PPC | OTH | Numerical lead |
|---|---|---|---|---|---|---|---|---|---|
| No records match these filters. Try another geography or clear the polling-firm filter. | |||||||||
Every polling firm in this archive 31 firms ↓
Firm names are reconciled across registries. A firm appears here if a qualifying federal voting-intention record was found in the two-year window. Counts include regional breakdowns from the same survey; they are not all independent polls.
| Polling firm | National | Regional | Riding | First record | Last record |
|---|---|---|---|---|---|
| Abacus Data | 59 | 339 | 0 | 2024-10-22 | 2026-10-13 |
| Angus Reid Institute | 23 | 126 | 0 | 2024-10-25 | 2026-09-07 |
| Campaign Research (Ontario PC internal) | 0 | 1 | 0 | 2025-03-26 | 2025-03-26 |
| Cardinal Research | 1 | 7 | 49 | 2024-11-15 | 2025-04-23 |
| Deltapoll | 1 | 0 | 0 | 2025-04-09 | 2025-04-09 |
| EKOS | 39 | 168 | 0 | 2024-10-26 | 2026-10-02 |
| Empirical Intelligence | 1 | 0 | 0 | 2025-04-07 | 2025-04-07 |
| Focaldata | 1 | 0 | 0 | 2025-04-23 | 2025-04-23 |
| Forum Research | 1 | 0 | 0 | 2025-04-27 | 2025-04-27 |
| Innovative Research | 26 | 102 | 0 | 2024-12-15 | 2026-08-25 |
| Ipsos | 16 | 72 | 0 | 2024-12-08 | 2026-08-27 |
| JL Partners | 1 | 0 | 0 | 2025-04-16 | 2025-04-16 |
| Kolosowski Strategies | 3 | 0 | 0 | 2025-04-04 | 2025-11-18 |
| Liaison Strategies | 107 | 578 | 7 | 2025-03-12 | 2026-10-17 |
| Léger | 34 | 214 | 0 | 2024-11-02 | 2026-10-12 |
| MQO Research | 5 | 26 | 0 | 2025-03-27 | 2025-04-26 |
| Mainstreet Research | 53 | 54 | 2 | 2024-11-24 | 2026-10-01 |
| Nanos Research | 132 | 148 | 0 | 2024-10-25 | 2026-10-16 |
| Narrative Research | 0 | 3 | 0 | 2025-04-13 | 2026-03-09 |
| Oracle Poll | 0 | 0 | 5 | 2025-02-16 | 2025-04-24 |
| Pallas Data | 14 | 78 | 1 | 2025-01-06 | 2026-07-18 |
| Pollara | 7 | 36 | 0 | 2025-02-19 | 2026-02-24 |
| Probe Research | 0 | 6 | 0 | 2024-12-10 | 2026-03-13 |
| Relay Strategies | 4 | 0 | 0 | 2024-12-06 | 2025-01-08 |
| Research Co. | 11 | 60 | 2 | 2025-01-04 | 2026-08-07 |
| Rubicon Strategy | 0 | 1 | 0 | 2025-04-10 | 2025-04-10 |
| Segma Research | 0 | 0 | 1 | 2026-08 | 2026-08 |
| Spadina Strategies | 0 | 0 | 5 | 2025-04-11 | 2025-04-26 |
| Spark Insights / Advocacy | 6 | 24 | 0 | 2025-05-28 | 2026-09-02 |
| Synopsis Research | 0 | 4 | 0 | 2025-04-11 | 2026-08-25 |
| YouGov | 2 | 0 | 0 | 2025-04-22 | 2025-04-25 |
Coverage, sources & trend method Open the full audit trail ↗
Reconciled public-registry inventory. All firms found in the listed sources are retained; unpublished, unindexed or inaccessible releases cannot be ruled out. This snapshot contains 547 national, 2,047 regional and 72 riding records. The archive is manually collected, not a live poll-discovery feed. Read the October 22 data audit .
The collection reconciles the 2025 and next-election national and regional tables, their constituency archives, current and historical federal registries, and additional original publisher releases. All firms found in those source tables are retained regardless of size or rating. Registry transcription is not independent verification of every original release. 19 records have a direct primary-source check recorded.
- wiki45 source archive (opens in a new tab) · 314 qualifying input rows before reconciliation
- wiki46 source archive (opens in a new tab) · 259 qualifying input rows before reconciliation
- riding45 source archive (opens in a new tab) · 61 qualifying input rows before reconciliation
- riding46 source archive (opens in a new tab) · 11 qualifying input rows before reconciliation
- 338 source archive (opens in a new tab) · 1,185 qualifying input rows before reconciliation
- 338past source archive (opens in a new tab) · 1,313 qualifying input rows before reconciliation
Wikipedia compilation attribution: contributors to the linked pages, retrieved October 22, 2026; CC BY-SA 4.0 (opens in a new tab). Factual registry values have been normalized and reconciled; third-party ratings, projections and chart artwork are not imported.
Date and sample checks: fieldwork-end dates take precedence where available; reference dates are explicitly labelled. Some regional entries list the whole survey’s n because a regional base is not supplied. Unknown n values stay unknown. Rolling releases may overlap; tables retain releases while trend lines use at most one release per firm.
Trend method: a trailing 30-day window, latest eligible release per firm, sample weight √min(n, 2,000), 14-day weight half-life, sampled weekly. When fieldwork ends on the same day, primary-reviewed records take precedence; within the same review status, the larger stated sample is selected. At least three firms with a reported value are needed for a party line. Reference-date, release-date, month-only, missing-sample, flagged and riding records are excluded from lines. The chart is descriptive and has no calibrated uncertainty interval.
Remaining review: 19 cross-registry discrepancies were flagged during reconciliation; 29 same-firm/date/geography groups also await review. Unresolved records stay in the ledger but are excluded from chart dots and trend lines; reviewed counterparts remain eligible. Exact duplicates and matching midpoint/end-date entries were merged; different samples and conflicting releases may remain separate. The displayed ledger retains provenance and review notes. Missing-party values are not inferred from remainders, and historical leader names are not retroactively changed.
Seat projections, majority probabilities and coloured riding forecasts require a separately validated model; they are not calculated from the screenshots or presented as poll data here.
03 / THE METHODNo mystery
No mystery
in the maths.
Method polityiq-federal-v1
Dataset scheduled-polling-v1
How the average is calculated +
We use reviewed Canadian national federal voting-intention polls released by the snapshot date, with fieldwork ending in the previous 30 days. The most recent fieldwork from each firm is retained; older releases from that firm are excluded to limit overlapping trackers and repeated influence.
Each poll receives this raw weight:
√min(total sample, 2,000) × 2^(−age / 14)Age is the number of days between the end of fieldwork and the snapshot date. A poll’s weight halves every 14 days. We divide by the combined weight to obtain the contributions shown above, then take each party’s weighted mean. At least three eligible firms are required.
These are initial design choices, not parameters fitted to past election accuracy. The model does not yet apply pollster ratings or house-effect corrections.
Sample sizes, rounding & uncertainty +
The sample weight uses total respondents consistently across firms because a decided-voter base is not available for every release. A total sample is not an effective sample size for voting intentions. Published decided bases are shown where available.
Individual releases retain their published percentages. For the calculation, six comparable party groups are rescaled to sum to 100 to handle rounding. A national Bloc share describes its share across Canada, not its support within Quebec.
Online panels, telephone surveys and differences in question wording can produce different results. Polls can share systematic errors. This first version has no calibrated confidence interval; displayed decimals do not imply that level of accuracy.
What happens after a new poll? +
We check the original release, field dates, total and decided bases, geography, voting-intention question and party categories. A reviewed record is added to the versioned dataset, the average is recalculated, and this page and its data downloads update from stored records without a new deployment.
A scheduled research task checks original publishers and submits structured records. Validation checks dates, geography, response categories and duplicates. Conflicting waves are held for review while accepted data is retained. Consult collection coverage for actual source checks; this is not a guarantee of exhaustive discovery.
From vote intentions to riding forecasts +
A seat model needs regional polling, past riding results, current boundaries and evidence for local candidate effects. National support alone is insufficient: the same vote share can produce very different seat totals depending on where voters live.
The next modelling stage must account for shared national and regional polling errors, simulate many possible elections and test past forecasts using only information available at the time. Riding probabilities and seat ranges will need that validation before publication.
Official Elections Canada riding results (opens in a new tab)
How to read the data
- Live
- Unofficial counts arriving now.
- Official preliminary
- Accepted official preliminary counts, subject to verification and final count.
- Projection
- Model output, not an election result.
- Simulated
- A rehearsal or what-if scenario, not a result.
- Archived
- Saved past results; the original source status still applies. Archived does not mean certified.