Polls During Election Campaigns

I have a problem with polls during election campaigns. To be blunt, I don’t think they should be published. Some say that polls do not influence how people vote on election day. I say those who say that have their head up their ass if they truly believe that.

As you know, Nova Scotians will be going to the polls in 8 days — the day after the August long weekend, perhaps the slowest work day of the year. No one is in the mood for an election, except the incumbent Tories who are hoping that widespread ambivalence in the political process at this time of year will sweep them back to power. And their strategy might work very well. Unfortunately, say I.

Enters my problem with polls. I believe that for many who sit on the fence (i.e., those with no set allegiance and whose vote could go to any of two or three parties), polls become a substitute for looking closely at the issues and the incumbents’ record. If one party’s platform doesn’t strike the fence sitters as spectacularly odious on the surface and that party seems to be leading the polls, then why not just be “agreeable” and go with what seems to be the flow? And if the lead is sizeable, why not just vote for what seems like a foredrawn conclusion (or not vote at all)? Besides, most people derive some satisfaction when, on election night, they find that they voted for the winning party.

I believe polls played an important role in sweeping the Conservatives to power in Nova Scotia in 1999. The 1998 general election had yielded a minority Liberal government: 19 seats for the Liberals, 19 seats for the NDP, and 14 seats for the Conservatives. Since the Liberals were the incumbents, they were asked to form the government. The gulf in political philosophy between the NDP and the Conservatives is such that the two parties could never have formed a majority coalition. However, even if they could have overcome their differences, I don’t believe they would have been allowed to form such a coalition under the current system, even though such a joining would have reflected the will of the majority of Nova Scotians.

Intent on governing as though it had a majority, the minority Liberal government elected in 1998 fell the following year. A summer election followed, but it had to happen: the province couldn’t go without a government just because it was summer. Polls published during the 1999 campaign showed a dead heat between the Liberals and NDP and the Conservatives trailing, suggesting Nova Scotia was heading towards another minority government. The thought of annual provincial elections was probably too much for most people to contemplate. With dissatisfaction towards the Liberals still in the air and what seems to be a belief that the NDP is better suited to be an effective opposition, Nova Scotians woke up the morning after the election a bit surprised to find themselves with a majority government headed by the Conservatives, which the polls had touted as the underdogs. The end result, though, was that Nova Scotians could set aside the notion of going back to the polls for yet another provincial election in 2000 or 2001.

Perhaps a huge chunk of the undecided voters in those 1999 polls opted for a devil they already knew from the late-’70s to early-’90s rather than the then-disliked Liberals or the NDP that never held power in this province. Perhaps they will pull the same stunt in 2003 and propel to power the supposed underdogs, the NDP, who’ve demonstrated great political maturity by staging an effective opposition in the last 4 years. (I doubt that’ll happen, though.) But with a strange twist, only one conclusion can be reached about the 1999 polls: they were wrong. Very wrong.

I don’t consider “who seems to be winning” to be a valid bit of information during an election campaign, at least not for the public. In fact, I see it as a way for fence sitters to avoid considering the issues. As such, polls are more than a snapshot of public sentiment. They can influence the outcome of an election. I truly believe that. Fence sitters end up relying not on their heart and mind but on often unreliable numbers, not to mention the desire to “do like everybody else.”

There’s one experiment I’d like to see done in a subsequent election: a moratorium on publishing poll results the moment an election campaign begins. Each party should, of course, be allowed to conduct polls to see if their campaign strategy is working. Then, only when the election is over, publish those poll results. If the aggregate data closely reflect what happened on election day, then I’ll shut up …after admitting that I was dead wrong in suspecting that poll results influence, rather than simply reflect, the outcome of an election. But I would like to see at least one election fought strictly on the issues, free of speculative number crunching.