Narrative in the 2015 Federal Election: Some Initial Results

During the recent federal election campaign, a research assistant and I recorded and analyzed all the English-language ads posted online by the three major parties. I am studying how the parties constructed their narratives for the campaign, what they had in common, and how they differed.

We went to the parties’ websites and YouTube channels daily to look for ads. We found some that were being frequently broadcast on television and others that were only posted online. By the end of the campaign, we had 52 Liberal ads, 43 NDP, and 40 Conservative, for a total of 135. Many of these ads still remain online, but others have been taken down. Thus, it was essential to study the ads as they appeared throughout the campaign.

We were looking for evidence that would address concepts in both political science and narratology. In this post, I’ll discuss the complete set of 135 ads.

Political scientists often refer to the presidentialization of Canadian politics, because of the vast, and ever-increasing, powers of the prime minister. In a campaign, presidentialization would refer to the extent that voters’ attention is focused on the backgrounds and personalities of the party leaders. It is therefore striking that 134 of 135 ads mentioned one or more of the party leaders. A narratological interpretation of this finding would be that a story needs a protagonist and the most compelling protagonist is not the party, but rather the person who leads the party. For the record, the one ad that didn’t mention party leaders at all was a satirical NDP ad in which an actor portraying a senator urged voters who wanted to see senators retain their power and perks not to vote NDP.

Narratives always have a narrator, a person or persons who tells the story. We found that 43 percent of the ads – by far the largest proportion – were narrated by the party leaders. Next, 23 percent were narrated by an unseen and unidentified but authoritative voice (often referred to by media scholars as a “voice of God.”) These unseen voices were only slightly more often male (18 instances) than female (14 instances). 13 percent of the narrations were produced graphically, by flashing story boards rather than a voice. 9 percent were narrated by politicians other than the party leader and another 9 percent by unnamed but opinionated citizens (or, as we can’t know for sure, by actors portraying said citizens).

Another question of interest to scholars, commentators, and the general public is whether politics is getting nastier, as measured by the balance between ads extolling the virtues or ideas of a party or its leader – what I call advocacy ads – and attack ads, which criticize another party’s ideas and/or its leader. 21 percent of the ads were pure advocacy ads while 33 percent were pure attack ads. The largest group – 45 percent – both advocated on behalf of one party and criticized one or more of the other parties. Put another way, three-quarters of the ads were at least partially attack ads. Therefore, attack ads continued to play a major role.

Returning to the narrative theme, standard components of moving image stories are people, lighting, and music. Politicians, and opinionated citizens (or actors representing them) are the stars. But there are also crowds. In 44 percent of the ads citizens were present in crowds enthusiastically listening to party leaders’ speeches or shaking their hands. Conversely, 56 percent of the ads did not have crowds.

In the area of colour, I was interested in the difference between brightly-lit and dark settings. The traditional belief is that an advocacy ad will be brightly-lit and an attack ad dimly-lit, especially if it refers to nefarious activities. We found 31 percent of the ads had a brightly-lit setting, 45 percent had a dark setting, and 24 used both (because some scenes were bright and others dark). While there were some attack ads that used classic dimly-lit settings, other factors were at play. Many ads were filmed indoors, for example at speeches or rallies, and used the available lighting and clothing colours of the crowd. In some ads, the party leader is in shirt-sleeves, with his white or blue shirt standing out against a crowd wearing darker colours.

Finally, music. We coded background music according to the mood it was intended to create. Our categories included upbeat and inspiring, foreboding and threatening, ironic and humorous, and calming. My research assistant, who has studied music at the university level, was instrumental (npi) in operationalizing the categories. Thus we identified upbeat music as being in a major key, with its pace accelerating and volume increasing during the ad. We classified 43 percent of the ads as having upbeat or inspiring musical accompaniment, 15 percent foreboding or threatening, 15 percent humorous or ironic, and 15 percent calm. The upbeat and calm music (58 percent) was generally associated with advocacy ads, while the foreboding, threatening, humorous, and ironic music (30 percent) with attack ads. Finally, 12 percent used no musical accompaniment.

Looking at these statistics overall, there seem to be some contradictions, or at least paradoxes. Party leaders were omnipresent and most often the narrators. There were frequent attack ads and the colours were often dark. On the other hand, the musical accompaniment was more often positive (upbeat or calm) than negative (foreboding or ironic).

The essential next step is to break the data down by party, so that each party’s narrative strategy becomes visible. This should enable us to see the difference between the Conservatives’ “not ready” attack strategy and the Liberals’ “sunny ways.”

As part of our coding, we noted the date each ad was first posted and the number of times it was viewed on YouTube up to Election Day. Thus we have the total views and average views per day for each ad, which can be considered measures of the extent to which it resonated with the electorate. There were considerable variations both in terms of how frequently individual ads were viewed and how frequently the different parties’ set of ads were viewed. Viewcounts thus enable us to comment on the success, or failure of the different parties’ narrative strategies. But this too is a discussion for future posts.

1 comment

Leave a Reply

Your email address will not be published. Required fields are marked *