COFFEE BREAK – Are polls representative of elections?

10 January 2020
Our Coffee Break with Leger video clip series is designed to provide a brief overview of answers to key questions in the market research world. If you would like further information about the research services we provide, click the “Contact Us” button at the bottom of this page to get in touch with our team.

Over the past 50 years in all Western democracies, polls have been very close to election outcomes pretty much everywhere. However, asking whether polls predicted election outcomes is actually the wrong question.

When polls are shown in the media in the crucial last weekend before the election, the results reported are for questions that were asked to Canadians the week before. On the last weekend before the election, we’re not out there asking questions to Canadians, so the poll does not include this information. The last weekend prior to the election is a crucial time, when many Canadians make up their minds on who they will support. More and more Canadians vote strategically and change their minds at the last minute, which is something that polls cannot see.

Polls are not there to predict the election outcome- they’re there to tell you how trends are evolving, how public opinion is shifting, and how this tells a story of who we are as voters. However, they still remain very close to the actual election outcome.

Christian Bourque has over 20 years of experience in public opinion and market research. He heads the team of researchers and advisors of Leger’s Montreal office. As company spokesperson, he regularly comments on provincial and federal politics and is a consumer trends columnist for Radio-Canada. Christian continues to teach research methods at the Université de Montréal.

Related Posts

Lana Porter Meets AIMI: A Human Take on AI-Moderated Interviews

As AI becomes more common in market research, qualitative teams are asking a practical question: how can new tools help us move faster and hear from more people, while still protecting the nuance and interpretation that make qualitative insights useful? For Leger,...

The Biggest Survey Design Mistakes in Academic Research

A well-designed survey can produce reliable, meaningful data. A poorly designed one can compromise an entire study. And yet, survey design is often underestimated. It’s more than just a series of questions; it’s one of the key pilars for a strong thesis. Whether...

Get the latest in your inbox

Stay up to date on cutting-edge research, news and more.