Imagine the following dilemma.
Suppose that during an election campaign you have decided on a party you are very keen to vote for, and another party you are very keen to see defeated.
Then suppose a poll is published showing your preferred party is falling behind, stuck in third or fourth place, while your least-favoured party is in a close race at the top.
Do you change your vote to help the main rival of the unwanted party, even though that’s not your first choice? Is that desirable behaviour by a voter?
What happens if the poll on which you are relying is flawed? Would you know the difference?
Given some recent missteps, it’s fair to say Albertans and political polls have had a rocky relationship of late due to projections that really seemed to miss the boat.
The current provincial election campaign has offered additional food for thought.
For Albertans trying to use the information, it’s a muddled mess that raises questions as to whether polls are providing more harm than help to the democratic process.
And if that’s the case, are there moves we as a society should consider to limit their influence?
But before those questions can be answered, it’s important to try to understand how voters use polling information.
Political scientists I chatted with, including Mark Pickup at Simon Fraser University and Melanee Thomas at the University of Calgary, suggest polls typically have minimal weight over how people vote.
The most susceptible are voters who remain undecided late into a campaign. But even among this group, while some may cast a ballot based only on a perception of who is winning or losing, evidence suggests many will catch up on the relevant issues by the end of the campaign.
Polls can actually help in this regard by keeping the electorate interested, and by bringing attention to parties, candidates and controversies that may have otherwise escaped attention, said Ian Large of the Leger polling firm.
And though parties will try to encourage so-called strategic voting — in which people vote for their second choice to keep another party out of power — there is no evidence this is widespread.
That said, polls can make some difference, if not so much on how people vote, then on whether they vote at all. Survey results showing a close race can drive up the turnout, while polls showing a landslide can often keep people home.
Regardless, to any degree in which voters make decisions on what could be very defective information, it’s worth discussing if something can be done to mitigate it. Which leads me to the idea of setting standards or limits around polling.
In Canada, no polling results are published on the day of an election. Other countries have longer blackout periods, ranging from three days in the Czech Republic to two weeks in Italy to up to 30 days in some Latin American nations.
Several countries also have mandatory “reflection” periods in which election advertising and even media coverage of a campaign is prohibited or severely restricted. The idea, in part, is to hopefully get voters to focus on issues rather than dodgy calculations of the horse race.
While Pickup and Thomas both suggest polling bans have minimal effect on voting intentions, I think the idea of a week-long blackout in Canada is worth considering.
For me, strategic voting is not generally a healthy thing, and I think democracy is better off by encouraging voters to pick the politicians who most closely align with their values.
That said, an extended blackout period comes with challenges. A potential infringement on free speech is one, and enforcing a blackout may be nearly impossible in the digital age.
As well, such a ban may unduly hurt smaller parties who can’t afford to do their own polling.
For those and other reasons, I ultimately think a better approach is to improve polling literacy among the electorate, including the media.
For one thing, voters should pay attention not just to the numbers but how parties, advocates and media report them. A five per cent deficit for one party could be portrayed as either a huge failure or a positive surprise.
Moreover, polls that tend to get the most attention are those that show the most change or movement. That’s a problem, Pickup said, because those polls are also the most likely to be outliers and subject to error.
Following aggregators, which average out the results of numerous polls, usually provides better information.
As well, voters could benefit from researching the reputations of various pollsters, and by gaining a basic understanding of methodology, response rates and weighting — so that they know, for example, that robo-call surveys generally aren’t as good as other methods.
In that vein, it would be helpful if polling firms were required to provide such information as a matter of law or professional standard.
I don’t pretend to suggest this is an easy issue to address.
Instead, all I can do is express my hope that when Albertans head to the ballot box next Tuesday, they vote with their hearts instead of their calculators.
Note from WSOE.Org : This content has been auto-generated from a syndicated feed.