Polls are not about telling the future, but about influencing it!

To: news@hilltimes.com, hill times <kmalloy@hilltimes.com>

Dear Ed, Eric Grenier summarizes his Did Quebec’s election polls lie?; HillTimes Sept 17/12 with Polls do need to be treated with more caution and with increased recognition of their limitations.   ... the results of the elections in Alberta and Quebec should provide an impetus for a greater understanding of what polls can tell us, what they cannot, and what that means in the context of an ever-changing election campaign.  During Elections, polls commissioned for publication (ie not all polls) by Parties/Campaigns are intend to influence voters. The idea is to show a wavering supporter that the "enemy" is gaining support and encourage the on-the-fence voter to vote against that apparent trend. The best example was the May 2/11 General Election when polls were published showing huge national-wide gains by the NDP - the numbers were true ... except in these so-called "national polls" Quebec-based support is adjusted to show it as part of the national picture (regional is goosed up). The late Mr Layton's ebullient public rallies after the Orange Surge poll - boldly speaking about becoming Prime Minister Layton - actually cooked his own goose. The middle class in the Rest of Canada, who distrusted the NDP more than they worried about the Ref/Cons in majority and  who had planned to contain the HarperTeam to minority again (they were right -look at the arrogance subsequently), decided the "enemy of my enemy is my friend" and flipped to Blue in the final days. Without talk radio & the right-of-centre press spreading the implications of the "apparent Orange Surge" presented in that (misunderstandable) poll 7-10days before election day ... Mr Harper would not be in the position of "Unstoppable and Unchecked Autocrat" aka King-for-Four-Years rce