Poll Tracker: Federal poll averages and seat projections
An aggregation of the latest federal polls by CBC poll analyst Éric Grenier
(Note to mobile users: Charts are more easily read when viewed horizontally.)
The weighted poll averages aggregate all publicly available opinion polls, weighting them by three factors: the date of the poll, the sample size and the track record of the polling firm. A full accounting of the polls included in the weighted average can be found at the bottom of this page. The change in arrows in the averages above shows the shift since the previous update.
The chart above shows the evolution of the weighted polling averages nationwide. | Read the weighted average methodology
- Full poll aggregation and seat projection methodology
- LEADER METER: Check out the federal leaders' approval ratings
The following seat projections make estimates for each of Canada's 338 ridings based on how voting intentions today differ from the results of the last three elections. It assumes that these shifting regional and provincial voting intentions will apply equally to each riding in that region or province — with some local factors being taken into account as well. The seat projections are an approximation of how many seats each party could win if an election were held today. | Read the seat projection methodology
The following chart shows the results of the seat projection model at the 80 per cent confidence interval. The low-to-high projected range reflects all of the seats a party could potentially win or lose if an election were held today, taking into account 80 per cent of past discrepancies between poll and election results.
The following chart shows the probability that each party would have of winning an election held today. These probabilities are based on simulations of 10,000 elections, which were done using the seat projections seen above, and which take into account potential errors in both the polls and the projection model itself. | Read a note about the projection methodology
The following charts show the unweighted average of all polls conducted each month from 2009 to present, with electoral results and the arrival of new permanent party leaders noted.
The following is a list of all polls currently weighted at greater than 0.5 per cent of the total projection.
|Poll / Media Sponsor||Weight||Method||Dates||Sample||MOE||LIB||CON||NDP||BQ||GRN||OTH|
|Nanos Research||18||TEL||Feb. 10-Mar. 9, 2018||1,000||+/- 3.1||36%||33%||19%||5%||7%||-|
|Léger / Journal de Montréal||22||NET||Mar. 5-8, 2018||1,536||+/- 2.5||38%||38%||14%||3%||5%||2%|
|Abacus Data||22||NET||Feb. 23-Mar. 4, 2018||4,023||+/- 1.5||36%||33%||18%||3%||8%||1%|
|Ipsos / Global News||13||NET||Feb. 28-Mar. 1, 2018||1,001||+/- 3.1||33%||38%||21%||3%||-||5%|
|Forum / Financial Post||12||IVR||Feb. 27-Mar. 1, 2018||941||+/- 3.2||34%||46%||14%||2%||3%||1%|
|Campaign / Hill Times||2||NET||Feb. 8-10, 2018||2,227||+/- 2.1||39%||32%||16%||4%||7%||1%|
|Nanos Research||3||TEL||Jan. 11-Feb. 9, 2018||1,000||+/- 3.1||38%||30%||19%||3%||9%||-|
|Abacus Data||2||NET||Jan. 26-28, 2018||1,411||+/- 2.6||39%||32%||18%||4%||6%||1%|
|Forum / Toronto Sun||2||IVR||Jan. 22-24, 2018||1,408||+/- 2.6||38%||43%||12%||3%||3%||0%|
|Innovative Research||1||NET||Jan. 11-17, 2018||2,777||+/- 1.9||39%||35%||14%||5%||7%||1%|
|Nanos Research||1||TEL||Dec. 16-Jan. 12, 2018||1,000||+/- 3.1||37%||34%||20%||4%||5%||-|
|Mainstreet Research||1||IVR||Jan. 3-6, 2018||9,830||+/- 1.0||40%||35%||13%||4%||8%||-|
(If on a mobile device, you may need to tilt your screen horizontally in order to see the full chart above.)
- Poll / Media Sponsor shows the pollster that conducted the survey, and the media outlet (if applicable) that either commissioned the poll or first reported on it. When available, a link to the pollster's report is provided.
- Weight refers to the weight (out of 100) the poll carries in the aggregation. Weight is determined by the date of the survey, the sample size, and the track record of the polling firm.
- Method refers to the mode of contact for the survey. Interactive voice response (IVR) polls feature automated calls, where survey-takers respond via the keypad of their telephone. Polls marked as TEL are those conducted over the telephone by live-interviewers. Online polls (NET) are conducted via internet panels, and do not carry a margin of error.
- Dates lists the field dates of the survey.
- Sample refers to the number of total respondents interviewed, including undecideds.
- MOE lists the margin of error (in percentage points) of a probabilistic sample equal to the size of the poll's sample size.
- LIB/CON/NDP/BQ/GRN/OTH shows the results of the poll among decided (and leaning, when included) voters only, for the Liberal Party (LIB), Conservative Party (CON), New Democratic Party (NDP), Bloc Québécois (BQ), Green Party (GRN) and others (OTH).