Tuesday, August 4, 2015
Election 2015: Ignore the Polls; Focus on Policy
Polling is a kind of science that is somewhat like economics. It involves a snapshot and a projection based on that snapshot, but the real world is not predictable. Things can change quickly, and considering the length of this campaign, the polls in the early stages are going to be particularly unreliable because any number of factors could change things.
There was an interesting panel I watched in the lead up to the election in which a number of pollsters discussed the shifts in the way people vote today. They explained that voters are less attached to parties than they were in the past. Because of this, the tendency for shifts, particularly in the last few days of a campaign is quite possible. Take for example, the UK election back in May. The polls right leading up to the election had a tie between the Labour Party and the Conservatives and the result was a Conservative majority.
There are a number of reasons that the polls might be off, and part of this is a tendency to underestimate Conservative parties, although this can be debated. There is also the fact that people could possibly be making up their minds very late in the campaign, or even as late as in the voting booth. So when we see the trend now with 11 weeks of campaigning to go, we should be careful how much stock we put in it.
The latest polls put the NDP ahead, but not by much, and the Conservatives have a tendency to outperform their numbers. This combined with the new ridings which overall will favour the Conservatives, it could be that the Conservatives may pick up more toward the top half of their estimated seat count than the lower. However, the trend overall appears to be the NDP with the momentum, with a slight Conservative uptick, both at the expense of the Liberals, who are lagging behind.
Now what could change the polls? The debates are going to be very important, and perhaps a better indicator of where the winds are blowing than the polls. The media also have a tendency to underestimate the impact of policy decisions in public opinion, favouring a focus on the gaff or the "knock out punch" in the debate. As we can see by what happened to the Liberal Party when they supported the controversial Bill C-51, we should not be too quick to assume that the public don't pay attention to things like this.
Another example of this is the Conservatives assuming that their decision to have an extended campaign and increase the costs won't be an election issue, but it very well could play into the narrative that the Conservatives do not play fair, and the question "why are we having an 11 week election" may not go away.
Polls do matter, but the overall sentiment of the voting public on the ground, and the reaction to party and government policy is a far better indicator of outcome than the snapshots given to us by polls. Polls cannot pick up on everything, and something that may seem insignificant at first may turn out to be the deciding factor.