Paul Fairie is a political scientist at the University of Calgary, where he studies voter behaviour.
Canada is currently in the midst of one of the longest elections in its history, and research suggests that this could be a good thing: it might result in better-informed voters.
So far, discussion about this unusually long campaign has focused on whether the Conservatives might use the more than $50-million spending limit to bankrupt their opponents in the final days of the election period, or if it might limit spending by outside groups.
These reasons might have been behind Stephen Harper's strategy to start the election early. But political science research suggests that there might be an unintended consequence for voters during this long two-and-a-half-months election: given longer campaigns, voters are likely to learn more about the economic and political realities of their community, and are also more likely to make their vote choice based on these factors.
This effect was the focus of research by political scientists Randolph T. Stevenson and Lynn Vavreck. They looked at 113 election campaigns in 13 different parliamentary democracies, including Canada. They measured a variety of aspects of each election, including campaign length, the incumbent party's vote share, the unemployment rate and rate of inflation, as well as measurements associated with several other common political variables to rule out rival explanations.
Using a technique known as linear regression, Stevenson and Vavreck found that the impact of economic factors on the incumbent's share of the vote depended on the length of the campaign. In longer campaigns, the state of the economy was more important for the incumbent party's vote share than it was during shorter campaigns. Put differently, when the economy is weak, governments should want a shorter campaign so voters do learn as much about the state of the economy and have their votes be affected by this.
Several different studies indicate that this learning process begins about six weeks prior to voting day. A study by statistician Andrew Gelman and political scientist Gary King, which popularized this notion of a six-week learning period, demonstrated that in American campaigns, one's pre-existing beliefs about politics become more important for determining party preference over the length of the campaign.
This doesn't mean that voters become more ideological during elections, but that voters' ideas about politics match up better with their party preference as the campaign grows longer. This learning process is particularly important in the last 40 to 50 days.
This line of research suggests that campaigns might not be about persuasion as much as they are about convincing voters that they already agree with a particular party. On top of this, elections are important for informing voters of the state of the economy and political world. Even if the facts that voters receive are very much mired in the spin provided by political parties, we can still think of this as voters becoming politically knowledgeable.
Political scientist Kevin Arceneaux found that this learning process can work either for or against the governing party as voters learn more about the state of the economy. In the last few weeks of the campaign, he showed that voters' evaluations of whether the economy is growing or shrinking become more and more important for determining their vote choice as election day nears.
Thinking about elections as opportunities for voters to learn about politics might explain why polls seem to have a high level of day-to-day variation, especially early on in a campaign. Polls can often seem to respond strongly to campaign events. Gelman and King showed that these fluctuations might not just be the result of an electorate being convinced to change their minds by brilliant campaign managers, but the consequence of voters learning more and more about the parties, eventually better aligning their beliefs with their votes.
What does this mean for this election? A key date will be September 7 – Labour Day – six weeks before the election. If evidence is a guide, many voters with normally low levels of engagement will start paying attention at this point. If opposition parties can convince this group that the economic and political state of Canada is negative, they can capitalize on this newly-attentive electorate. To remain in power, the government must convince these voters that everything is fine.
In the context of five consecutive months where the Canadian economy has shrunk, this fact must be making the incumbent Conservatives nervous because a reasonably large proportion of the electorate are not yet deciding their vote preferences based on the economic and political state of the nation. Given the longer campaign, research suggests that more Canadians than ever will be doing so this year. What voters learn in the last six weeks of the campaign and how well that aligns with their pre-existing beliefs will determine the outcome of this election.