THE professional pollsters may have got the election result wrong, but award-winning columnist Trevor Lautens - swimming against the stream as he often does - correctly predicted a win for Christy Clark in this newspaper on May 10 (read it at nsnews.com).
Perhaps Lautens will share the secrets of his crystal ball with us in his May 24 column, but Angus Reid, chairman of Angus Reid Public Opinion, has already gone public with why the predictions were so far off-base. It was his company's poll, released the day before the election, that muted the B.C.-Liberals-close-the-gap story. It put the NDP some nine points up in the race.
Reid acknowledges that pollsters will have to do a better job of ensuring their samples reflect the shifting demographics of ethnic communities in Metro Vancouver.
That should certainly be done. But it is also now clear that online polls do not accurately capture life in the polling station. An online poll is usually a series of questions, but the ballot lacks that context. A single "X" really does mark the spot.
More importantly, perhaps, "old-fashioned" telephone polls capture a greater percentage of those who actually go out and vote: the retired. Those willing to participate in an online poll may not participate on election day.
It would appear the B.C. Liberals knew this. The much more interesting question is why the NDP did not. Or, if they did, why they adjusted their campaigning too late in the election to be effective.
