New poll claims youth vote doubled, too bad it’s a poor survey

Historica-Dominion Institute survey has so many problems, it’s impossible to take its conclusions seriously

Youth voting rates doubled on May 2, when compared with the 2008 election, and more than half of those votes went to the NDP, according to a poll released yesterday by the Historica-Dominion Institute.

I’d be willing to believe that more than half of youth votes went to the NDP, but that might be because I live in Montreal and I don’t think I actually know anyone who didn’t vote for the NDP. And while I wouldn’t be surprised if the youth vote-rate increased, I’m a little incredulous about the claim that it rose to 76 per cent, from 37.4 per cent three years ago.

Now, it may be that the possibility of electing some of their peers to parliament drove youth turnout to unprecedented highs. Or it may be that this is just a shoddy survey.

The Canadian Press story about the poll gives some indication of the latter, saying (right at the end) that, “since it was conducted online, it’s impossible to say precisely how exact the poll is.”

But that’s the least of this survey’s problems.

The Historica-Dominion Institute claims the poll is representative of “youth” but that’s not actually the case. The participants are all members of a website,, which claims to help its members obtain university scholarships.

So, in fact, the survey only questioned students and those who intend to go to university in the near future. There’s a big difference between the voting habits of university students and non-student youth.

But even if this survey were being passed off as representative of students, there would be some big problems. The participants were selected from among a group of people who chose to register with a specific website and who opted-in to receiving surveys of this nature. That’s a pretty limited group and one that certainly has some self-selection bias.

The survey also has problems with sample size. Only 812 people across the country participated and while that might sound like a decent number of people, it means some very low sample sizes at the provincial level. In fact, the suvey doesn’t even provide the number of respondents by province for the Prairie (Saskatchewan, Manitoba and Nunavut) and Atlantic regions. Fewer than 60 people were survived in each of those regions.

There’s another big problem with post-election surveys, people tend to over-report voting. Post-election surveys, commissioned by Elections Canada, have found that the number of people who claim to have voted, when surveyed, is usually around 20 percentage points higher than the actual voting rate.

While I’d certainly like to find out that youth turnout increased in this election, we’re going to have to wait a little longer to find out what actually happened.

Looking for more?

Get the Best of Maclean's sent straight to your inbox. Sign up for news, commentary and analysis.