Justin Trudeau’s Liberals started the summer recess with their lowest share of committed voters since they were elected in 2015. They trailed the Conservatives in all age groups, with men and women, and in all provinces but Québec. A month later and things remain much the same.
On most top-of-mind issues, including the economy, housing and immigration, Pierre Poilievre and the opposition Conservatives have the backing of many more of those surveyed than the governing Liberals do.
The poll tracker is a product of the ongoing partnership between Abacus Data and the Toronto Star, launched last year to continually take the pulse of Canadian voters on the political issues, problems and personalities that shape our national debates.
* Although opt-in polls cannot be assigned a margin of error, the margins listed here are for a comparable random sample of this size.
Methodology
Abacus Data uses online, opt-in panels to survey a nationally representative sample of Canadian adults. Participants are recruited from multiple panels to reduce the bias any single panel might have. Results are then “weighted” against census data to ensure the findings effectively represent the larger population.
While online polls cannot be assigned a margin of error, they are regarded as an accepted, industry-standard method of polling. Phone call response rates have drastically declined with the advent of caller display, and many firms have turned to opt-in polling as a way to get a significant sampling of people.
Polling website 338 ranks several pollsters that use opt-in panels – Leger, Ipsos and Abacus – as among the most accurate in predicting the last three elections. One study, in Cambridge University Press, found online polling to be a likely “mode of choice for subsequent election studies” over telephone polling, due to the methodology’s relative accuracy and cost-effectiveness. Another study comparing telephone polling to online polling found neither survey mode was superior to the other.
No polling method is perfect. In a study from the Pew Research Center, for example, respondents’ answers to the same question varied depending on whether they were surveyed by phone or web (people were more likely to give positive responses over the phone). While online surveys can be cheaper and reach a wider audience faster than phone surveys, they also offer less personalization, which can lead to less detailed responses. A study from Pew found the errors in six opt-in online surveys it examined were due to “bogus respondents” or those who answered “yes” regardless of the question.
Abacus Data has tools to help prevent bots from infiltrating its surveys. It also seeks to ensure respondents are thoughtfully answering by including several questions that gauge whether participants are actually engaged in the research.
Mark Ramzy is an Ottawa-based general assignment reporter for the Star. Reach him via email: mramzy@thestar.ca
Stephanie Levitz is an Ottawa-based reporter covering federal politics for the Star. Follow her on Twitter: @StephanieLevitz.
Raisa Patel is an Ottawa-based reporter covering federal politics for the Star. Follow her on Twitter: @R_SPatel.
Alex Ballingall is an Ottawa-based reporter covering federal politics for the Star. Follow him on Twitter: @aballinga.
Susan Delacourt is an Ottawa-based columnist covering national politics for the Star. Reach her via email: sdelacourt@thestar.ca or follow her on Twitter: @susandelacourt
Cameron Tulk is the Graphics Editor at the Star, based in Toronto. Reach him via email: ctulk@thestar.ca.
