The Effect of the Market on the Framing of Elections at the Subnational Level in Canada

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40

The Importance of Context: The Effect of the Market on the Framing of Elections at the Subnational Level in Canada

Shannon Sampert Department of Political Science, University of Winnipeg (on leave) Perspectives and Politics, Editor, Winnipeg Free Press– Email Address: [email protected]

Adelina Petit-Vouriot Department of Political Science, University of Toronto – Email address: [email protected]

Abstract: In Canadian media studies, very little research has been done on how elections at the subnational level are covered by newspapers. This paper analyzes the way major newspapers in six Canadian capitals covered provincial elections between 2011 and 2012. The findings were interesting in that they underscored the importance of context in news coverage. Rather than a formulaic game frame focus, with an adherence to who’s winning and who’s losing, our study determined that each newspaper had a distinct voice and focus in their news coverage. This adds a new and interesting dimension to the study of election coverage in Canada.

Keywords: Election studies, media analysis, content analysis, framing, game frame, agenda setting.

Résumé: Dans les études de media canadiennes, très peu de recherches ont été effectués sur les élections au niveau régional qui sont couvertes par les journaux. Ce document analyse comment les journaux régionaux de six capitales canadiennes on couverts les élections provinciales entre les années 2011 et 2012. Les résultats étaient intéressants, ils insistaient sur l’importance du contexte pour la diffusion des actualités. Plutôt qu’un jeu stéréotypé, avec un gagnant et perdant, notre étude a déterminé que chaque journal avait une présentations et diffusion particulière. Cela ajoute une dimension nouvelle et intéressante aux études des élections régionales au Canada..

Mots-clés: étude d’élections, étude de media, étude du contenue, formulation, formulation du jeu, ordre du jour

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40

Theoretical Framework (Strömbäck et al. 2012, 720). Landerer notes that the normative logic of media The literature on election media deals with concerns about what “they coverage in Canada tends to focus on should do in order to contribute to the national elections, overlooking the democratic process” (2013, 245). This possibility of comparative election clashes with market logic and is analysis at the subnational level. Based subordinated to concerns about the on the existing federal-level research, “maximization of audience… in order to media outlets have received considerable generate profit” (2013, 244). criticism for their increased adherence to Election coverage in Canada framing elections as “games” – contests in clearly illustrates this conundrum. For which the focus is on winners and losers, journalists working on elections, their polls, strategies and campaign gaffes. “public service function suggests they With that type of framing, the concern is should focus on issues and policies to guide voters through a myriad of that news consumers are not provided promises. Yet elections are horse-races. adequate information about policies and The story of who is winning and who is issues; as a result, their ability to make losing often overshadows the issues” (Cox informed decisions at the ballot box is 2010, 97). stunted. This chapter expands on the The ubiquity of a contest frame current literature by evaluating election fulfills the market side of media logic coverage at the subnational level using which news media rely on in order “to be local newspapers. We are interested in competitive in the ongoing struggle to further examining the concepts of capture people’s attention” (Strömbäck, mediatization and media logic – the news 2008, 233). As Cox writes, “money values and story telling techniques used determines everything” (2010, 105) and to capture media audiences (Strömbäck, the need to entertain is seen as necessary 2008; Strömbäck and Dimitrova, 2011). to maintain audiences and attract readers. Is there a standard format to news While issues “are supposed to be at the coverage of elections or does the market core of any debate over choosing a in which the newspaper is located have government… who will win and how is the an effect on how the election is story” (Cox 2010, 101, emphasis added). portrayed? Covering elections is expensive. Not every event that happens Newspaper election coverage does not becomes a news story (Bain 1994). provide revenues for newspapers but Mediatization literature holds that story instead creates an added expense, with selection relies on a number of different papers having to pay for reporters to factors that are both normative – relating travel alongside political leaders (Cox to format and news values – and market 2010, 107). Media-sponsored polls during oriented – relating to market and elections are also very expensive (Adams audience competition (Landerer 2013, 2010 162). Canada’s newspapers 243). The more competitive the media struggling with falling circulation rates market, the “more the media have to take and advertising revenues, particularly in into consideration what audience saturated media environments, must segments advertisers are interested in make decisions about what to cover and reaching and what kind of news these how to do so in order to maintain or build audience segments are interested in” audiences while preserving the budgetary

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40 bottom line. Previous studies in Canada the media to select and frame election focused on the news selection and stories in this way (2009). framing of election coverage in The use of the contest frame and newspapers, yet none have examined the the reliance on polling information to context of the market logic used by media drive election coverage appears to be organizations during election coverage increasing at the federal level. Pétry and and none have examined the impact of Bastien found that in the 2008 federal market factors in how elections are election, media outlets sponsored over covered. This chapter seeks to do this by 200 national surveys on voter intentions providing a comparison of election over a 37-day period. This compares to coverage through a variety of newspapers just 22 national polls reported over a 51- and provinces, positing that the market day period in the media in 1988 (2013, 1). context and political environment in A more general analysis of the 2004 and which the papers operates influences the 2006 federal elections in seven daily type of coverage provided. Canadian newspapers determined that The specific critique of the role of the contest focus was featured in more the media in framing elections concerns than half of the newspaper coverage in its potential effects on democracy. 2004 (55%), up from 45% in 2006 Studies in North America and Europe (Soroka and Andrew 2010, 116). In their have determined that media tend to study of the headlines in Canada’s two frame elections as contests and with national newspapers (The Globe and Mail limited substantive discussion regarding and the National Post), Trimble and policy or policy outcomes (see for Sampert determined that the contest example Capella and Jamison, 1997). frame was frequently used in the Framing is a necessary story telling coverage of the 2001 federal election. technique and a way “of processing and Anderson studied the election coverage of packaging information so it can be quickly the 1997 federal election in eleven conveyed by reports and easily Canadian daily newspapers and three interpreted by the audience” (Trimble television networks and found that and Sampert 2004, 52). Contest coverage polling information and the horse-race refers to the framing of elections as “a frame dominated news coverage at the political contest between political actors expense of information about political with limited attention for the issues at issues (Anderson 2000, 294). There are stake” (Takens et al. 2013, 280). Inherent concerns that the use of the contest frame in that construct is an emphasis on the is a “continuing obsession” for news game, a “strategic meta-narrative” that organizations and that the creation of relies on horse-racing, sports or stories which are “easy to report” allows battlefield metaphors (Trimble and news organizations to appear neutral and Sampert 2004, 53) and is enhanced by the authoritative and absolve them of the responsibility to probe more deeply, use of campaign polls and poll reporting explore issues more thoroughly, and in the media. As Strömbäck and Esser describe the consequences of party point out, the emphasis on the “game” or positions for the country (Taras and the contest frame in election coverage Waddell 2012, 75). points to the market logic of media because it is competitiveness that drives Pétry and Bastien are very critical of the use of polling information, calling

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40 the reliance on pollsters “a threat to McFadyen offered a new vision for the democracy” because the inaccuracy of province after twelve years of NDP rule. polling results coverage “deceives more In Ontario, Dalton McGuinty’s Liberals than it informs voters” (2013, 20-21). were lagging in popular support. Indeed, Indeed, as McGrane points out in this a Toronto Star poll released in June of edited collection, polling accuracy is of 2011 suggested that McGuinty was one of considerable concern particularly in Canada’s most unpopular premiers pollsters ability to accurately forecast (Benzie 07 June 2011). The other three election outcomes. provinces in our study, however, could An additional component to the also view the election as a game, mediatization literature outlining market particularly with a focus on potential logic is an emphasis on opinion writing changes in the number of seats being held and columns. As Sampert et al. by each party in the provincial determined in their analysis of the legislatures. While PEI, Newfoundland coverage of leadership campaigns in the and Labrador (NL), and Saskatchewan Globe and Mail over a 37-year period, the had governments that were still quite use of simplified and personalized media popular, improving their seat count was a coverage has increased over time – focus. particularly after 2000 – as a direct result As each election progressed, the of market competition (Sampert et al. parties and their leaders were forced to forthcoming). There are concerns that deal with a number of controversies that newspapers are moving away from arose, even in the provinces with reporting news stories and are instead relatively certain election outcomes. For providing opinion that is “often example, in NL, a controversial hydro- interpretive, critical, provocative and electric project in Muskrat Falls was an partial, standing in stark in contrast to the issue highlighted throughout the more thorough and balanced approach of campaign, with concerns about cost the regular news story” (Sampert et al. overruns. In PEI, immigration became a forthcoming). central problem for the Liberals. A front- In at least three of the six elections, page Globe and Mail article referred to it would seem that framing the election as “unsubstantiated allegations of fraud and a game would make sense. Based on early envelopes of money changing hands polls, government turnovers were between PEI government bureaucrats and possible in Alberta, Ontario and Manitoba. well-heeled Chinese immigrants, under In Alberta and Manitoba, the incumbent PEI’s Provincial Nominee Program premiers were facing their first election (McKenna and Desserud 2013, 68). In since winning the party leadership – Saskatchewan, while there was no Alison Redford and Greg Selinger. specific scandal, there was considerable Redford’s Progressive Conservatives were discussion about program funding and as thought to be confronting their first real the weeks progressed, concerns about the competitive election in decades. In profitable third-quarter earnings report Manitoba, Selinger’s New Democrats were from the Potash Corp increased, and facing the electorate for the first time centered on the NDP argument that these since the departure of the extremely earnings should be used to pay for social popular Gary Doer. Some polls suggested spending (McGrane et al. 2013, 4). Thus, the NDP was in trouble as PC leader Hugh even in provinces in which the race’s

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40 outcome was expected, there were still the Atlantic provinces have the lowest potential points of contestation and circulation numbers at 25,645 for The contention on which the media could Telegram and 17,724 for The Guardian hook. (Canadian Newspaper Association). The Guardian and the Telegram publish a Methodology newspaper six days a week. The Leader Post, Star, and Journal publish as a In order to compare newspaper broadsheet seven days a week. The coverage across provinces, we conducted Winnipeg Free Press also publishes seven a manual content analysis. In total, we days a week, but its Sunday paper is only examined six newspapers published in available in stores or online thus we did the capital city of each of the provinces in not include it as part of the overall which elections were held in the study’s analysis as it was not widely available. time frame of 2011-2012. We analyzed The ownership model and media coverage in The Telegram in St. John’s, market in which the papers operate is Newfoundland and Labrador; The also important to understand. The Star is Guardian in Charlottetown, Prince owned by Torstar, which also owns the Edward Island; the Toronto Star in Metro, a free daily distributed in nine Toronto, Ontario; the Winnipeg Free Press cities, and it also operates in one of the in Winnipeg, Manitoba; the Leader Post in most highly saturated and competitive Regina, Saskatchewan; and the Edmonton media markets (see for example Androich Journal in Edmonton, Alberta. In cities in 02 November 2012). The Toronto Star is which there was more than one Canada’s largest daily newspaper in newspaper available, we chose the Canada’s largest city. The Winnipeg Free newspaper with the highest circulation Press is an independent newspaper numbers. Where possible, we relied on owned by FP Canadian Newspapers, the physical copy of the newspaper; which also owns the Brandon Sun. however, some days were missing. We Winnipeg’s media market is viewed as then used microfiche and as a last resort, quite small, with the Free Press competing the electronic versions of the newspapers. with the Metro and the Sun for newspaper Three of the papers in this study advertising revenue, but Winnipeg is could be considered “large papers,” not considered the only major city in only because of their circulation size but Manitoba, so all major newspaper also because of the media market in advertising revenue will be spent in that which they operate, while three of the market first. The Edmonton Journal and newspapers could be considered “small Leader Post are part of the Postmedia papers”. The Toronto Star, Winnipeg Free chain, which has 9 daily papers including Press and Edmonton Journal have the top the National Post. In terms of population, three circulation numbers of the papers in Edmonton is larger than Winnipeg, while the study. The Toronto Star has the Regina is smaller. The Guardian and the greatest average daily circulation Telegram are both owned by TC. Media, a numbers at 357,612. The Winnipeg Free company which has nine newspapers in Press reaches 113,251 households and the PEI, NL, Nova Scotia and Saskatchewan as Edmonton Journal 101,950. By well as one in Montreal. Both of these comparison, the Leader Post has 41,598 newspapers service population areas that subscribers and the papers that publish in are quite small. In terms of economies of

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40 scale, the expense of covering elections in holidays (some newspapers do not Edmonton and Regina may be publish on a statutory holiday). ameliorated somewhat by the presence of Headlines and stories were coded “sister” publications in the same province. separately because the headline function For example, the cost of covering the is different than the story (van Dijk 1991, leaders’ tours in Alberta can be shared 50). Headlines are meant to capture the between the Edmonton Journal and the reader’s attention and deliver information Calgary Herald. The same could be said in a short, pithy format, thus aligning with about coverage of the Saskatchewan the market logistics of entertainment. By election with Regina and Saskatoon comparison, stories are meant to provide sharing election costs. The “news traffic” more in-depth information and detail. in each city would be substantially Two coders were trained to manually different as well. Toronto is the centre for code each story. An intercoder reliability Canada’s financial and business news, test conducted on a random sample of ten making competition for attention by the percent of the news stories found there media intense. In comparison, smaller was agreement on coding individual cities like Charlottetown and St. John’s variables at least 70% of the time.1 would have less “news traffic” making the This paper examines the nature of possibility of getting published in either the coverage of six provincial elections the Telegram or the Guardian easier. across Canada and explores two main The decision to study newspapers hypotheses. First, we argue that elections rather than television stories was would be considered important news deliberate. First, newspapers at both the stories particularly in cities in which the national and subnational level are still government had the potential to fall. viewed as important agenda setters, Therefore, we anticipated more attention particularly by political strategists (Taras to be paid to the election in Toronto, 1990; Trimble and Sampert, 2004). Winnipeg and Edmonton. In addition, we Second, newspapers provide more in- reasoned that the contest frame would be depth coverage of elections when ubiquitous in all of the provinces, but compared to television stations, which more prominent in the three provinces in are limited by their format. Finally, given which the governing party was in that this study incorporates six provinces, jeopardy. Explored below, our findings it was most feasible to obtain copies of indicate there were very clear differences newspapers stories when compared to in how the provincial election stories attempting to access video transcripts. were told and how they were framed as The period of analysis was four contests. Instead of a media logic weeks from the day after each election influencing journalistic standards that was called, excluding Alberta since its were replicated in each election, we election date was not fixed. In the Alberta determined interesting nuances that case, we studied the newspapers from the suggest that market logic considerations, day after the writ was dropped (March such as community and size of 26, 2012) until the day after the election newspaper, were significant factors in (April 24, 2012). The actual number of newspaper coverage. days studied varied depending on publication days (some newspapers only publish six days a week) and statutory

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40

Findings Guardian reported the second highest average, followed by the Journal, the Free Just by looking at the newspapers, Press, the Telegram and finally, the Leader it becomes clear that each election was Post. These findings make it difficult to covered in the paper’s unique way. For make any assertions about newspaper example, the Edmonton Journal provided coverage of elections based on size of a large amount of graphics to showcase circulation, competitiveness of the information about the Alberta election. newspaper market, or even volatility of Additionally, it provided a specific section the election. Instead, it would appear that called “venting” in which readers were the decision to dedicate time and invited to write comments about the resources to covering the provincial election. While we did not include coding election is dependent on the newspaper of the venting segment in our following managerial team’s dedication to the data, it is important to note that there subject. were 59 of these vents published in the Another way to gauge the level of Journal, many of which were extremely importance given to the provincial negative towards politics in general. election is its placement in the Conversely, the Winnipeg Free Press newspaper. The front-page is where the launched the “Democracy Project” a year most important stories of the day will be ahead of the election date, in a bid to get found and it is used to capture the more people interested in voting. During reader’s interest (see for example Gibson its election coverage, several stories were 2011). Thus, the use of the front-page is written about voter apathy alongside indicative of the market logic of the stories that attempted to educate media. It is used to “sell” the paper. constituents about voting (Goodhand 02 Clearly, the Toronto Star did not see it as a October 2010). In addition, the Free Press’ part of its market logic, because it had the front page does not contain any editorial lowest number of front-page stories than content. Instead, it offers graphics and any other newspaper in our study – only photos, with editorial content continued 15. The Star opted instead to place on subsequent pages. Coverage of the election stories in its “Insight” and election in the Atlantic papers was much “Greater Toronto” sections. The Guardian more traditional, with a focus on news had the highest number of front-page stories rather than graphics or stories with 43. The Telegram was third photographs. with 38, the Edmonton Journal followed As a means of determining the with 33, the Leader Post had 27 and the importance of the election, we looked at Free Press had 21. The Free Press the overall number of stories given to numbers are also interesting because it is election coverage in each newspaper and devoid of editorial content on the front- determined the average number of stories page, making room for more headlines that ran per day. Not surprisingly, the than in a conventional front-page, yet it largest newspaper – the Toronto Star – had the least front-page headlines overall. had the highest average number of stories per day over the election period under study, publishing slightly more than ten stories a day. However, the much smaller

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40

Table 1: Number of election stories per paper Star Free Press Journal Leader Telegram Guardian Post Total 288* 172** 264* 148 160 222 number of stories Average 10.3 7.2 8.8 6.2 7 9.3 stories per day Front-page 15 21*** 33 27 38 43 stories

* indicates the inclusion of Sunday publication ** does not include headlines from front-page with no editorial comment *** the Free Press front-page has no editorial content, thus these are headlines and not stories

Table 2: Story type by newspaper Star Free Press Journal Leader Post Telegram Guardian

Hard News 165 110 196 92 125 151 57.3% 57.0% 74.2% 62.2% 78.1% 68.0%

Editorial 12 12 16 6 9 12 4.2% 6.2% 6.1% 4.1% 5.6% 5.4%

Columns 54 26 35 40 23 13 18.8% 13.5% 13.3% 27.0% 14.4% 5.9% Briefs 43 15 8 0 2 37 14.9% 7.8% 3.0% 0% 1.3% 16.7% In-Depth 1 1 1 0 0 0 Feature .3% .5% .4% 0% 0% 0% Personality 10 1 0 2 1 8 Profile 3.5% .6% 0% 1.4% .6% 3.6%

Voter Info 0 6 7 4 0 1 0% 3.1% 2.7% 2.7% 0% .5%

Coverage of the election was not more intensive election coverage. The consistent over the campaign period, Guardian, for instance, published a either. Examining the weekly number of relatively greater number of election articles in each newspaper reveals stories in Week 1 of the PEI campaign, different styles in how newspapers used while the remaining papers started

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40 unhurriedly, steadily increasing their events including where leaders or campaign coverage until Election Day. candidates were to appear and contained The higher quantity of stories by the little substantive information. The largest Toronto Star and Edmonton Journal came paper (the Star) and one of the smallest in the final week of the election, when papers (Guardian) had the highest overall their coverage leapt in comparison to percentage of “in brief” stories, suggesting other papers. that this may have been a managerial The type of story used to describe decision to use briefs to fill news holes in election events also provides insight into the newspaper. election coverage. As discussed, there is Next, we examined the prevalence evidence that opinion pieces or columns of the contest or game frame in the are replacing news stories as part of election coverage. As the literature newspapers’ market logic. This means suggests, the contest frame should be that consumers are being provided with ubiquitous in all the papers, but in fewer opportunities to read about events particular, we expected to see its use that are presented as regular news – entrenched in provinces in which there news that follows journalist guidelines of was an anticipated potential change in fairness and balance. Instead, the news is government. It did not happen. Only two mediated through editorials or columns of the three papers in which the race was that hold a particular perspective. There viewed as competitive used the game are variations on the level of mediation frame in over 50% of the stories: the Star across the board. For example, hard news and the Journal (67.7% and 54.5%) while comprised less than 60% of the Toronto the Free Press used a contest frame in just Star’s coverage (57.3%). Almost 20% of 35.5% of its stories. The Free Press’s its coverage appeared in columns, and concerted effort to increase voter turnout 4.2% of its election stories were through its Democracy Project, a clear editorials. The Regina Leader Post also decision to focus on the public interest or had a relatively low percentage of hard normative logic in media, accounted for news stories (62.2%). By contrast, the some of this variance. In the smaller Leader Post’s columns were 27% of its papers, the contest frame was used in overall coverage and editorials made up under 50% of the stories. It was found in 4.1% of the news coverage. Coverage in 45.3% of the Leader Post stories, 47.5% f the Telegram was found predominantly in the Telegram stories and 32.4% of the hard news stories (78.1%), and less than Guardian’s. This is remarkable given that 20% of its stories were mediated through the analysis of news coverage at the columns and opinion pieces. federal level indicates that the contest Interestingly, the Guardian had the frame should be the common frame for highest percentage of election news the story. Using a chi-square test, we stories in the “in brief” format – that is, determined a relationship between short election stories with limited newspaper and the use of the game information. For the Guardian, “in brief” frame.2 For example, both the Star and stories made up 16.7% of its overall the Journal had higher percentages than coverage. The Star also had a higher expected for game frame stories, while percentage of “in brief” stories at 14.9% the Free Press and the Guardian had much (43 stories). Such articles usually lower percentages than expected. Both provided brief information about election the Leader Post and the Telegram’s

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40

Table 3: Story Frame by Newspaper

Star Free Press Journal Leader Post Telegram Guardian

Game 196 61 144 67 76 74 67.7% 35.5% 54.5% 45.3% 47.5% 32.4% Issue 75 100 109 71 78 117 26.0% 58.1% 41.3% 48.0% 48.8% 52.7% Other 17 11 11 10 6 33 5.9% 6.4% 4.2% 6.8% 3.8% 14.9% Total 288 172 264 148 160 222 100% 100% 100% 100% 100% 100%

Table 4: Top five categories of game frame by newspaper

Star Free Press Journal Leader Post Telegram Guardian

Strategies 52 23 24 21 17 9 26.5% 37.7% 16.7% 31.3% 22.4% 12.5% Leader 47 10 27 16 15 15 24.0% 16.4% 18.8% 23.9% 19.7% 20.8% Candidate 34 3 40 7 14 10 17.3% 4.9% 27.8% 10.4% 18.4% 13.9% Party 17 15 33 9 16 15 8.7% 24.6% 22.9% 13.4% 21.1% 20.8% Polls 14 5 9 5 9 11 7.1% 8.2% 6.3% 7.5% 11.8% 15.3%

percentage of stories that contained a 55.6% in the last week, and in the Journal, game frame were close to what was it climbed from 40.4% to 59.5%. In sum, expected. there was a clear “race to the finish” Framing appeared to evolve over mentality among all newspapers, the course of the election. The fourth and although some used the horse-race frame final week of each election saw each more than others. paper’s share of the game frame surpass Remarkably unlike federal the issue frame, with most newspapers elections, polls did not push election nearly doubling the use of game frame coverage in the provinces under over the course of the campaign. For examination. In stories that contained the example, in the Guardian, use of the game game frame, we coded to determine what frame increased from 24.5% of the total was the impetus behind the contest frame stories in Week 1 to 64% in the final and determined the top five indicators to week; in the Telegram, from 47.6% to be (in order of overall frequency) 62.5%; in the Star, from 69.8% to 71.5%; campaign strategies, leader performance, and in the Free Press, from 20% to 48.1%. candidate performance, party In the Leader Post, the game frame performance, and polls. Polling stories increased from 44.8% of the stories to only made up only 8.6% of stories framed

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40 as a contest. The focus on strategies and The role of market logic is insider’s looks at leaders, candidates and underscored further in the examination of party performances speaks to the framing. The two large papers operating prominence of market logic. Such stories in saturated media markets in the context represent a potential “scoop” for of a potential government upset – the newspapers and the possibility for Edmonton Journal and the Toronto Star – scandal since they reveal “behind-the- clearly used a contest frame to portray scenes information that the parties do not the election. The smaller papers covering really want public” (Cox 2010, 105). elections that were not likely to see a change in government focused on issues. Implications Perhaps more telling, in an election in which there was a potential for an upset – This is the first time in Canada that the Manitoba election – the story frame a comparative analysis of newspaper was predominantly issue-based. This coverage of provincial elections has been may be in part due to the commitment conducted, adding more dimensions to made by the newspaper to provide the growing study of media in the information about the election as part of country. What can be made of the its Democracy Project but it also may be findings? First, provincial election in part the result of the market. The coverage is not like federal election Winnipeg Free Press dominates its market coverage. It becomes obvious that when throughout the province and unlike it comes to writing about elections, the Edmonton and Toronto does not have as role of market logic, including much competition. For the smaller competitiveness and market size, shape papers, even when there was a potential news coverage. For example, the Star had for scandal, the focus remained on the the resources and the newspaper space to issues rather than on the contest. For provide coverage of the provincial example, the Provincial Nominee Program election, and so it is no surprise that it in PEI and the Muskrat Falls story in NL had the highest average for number of could have easily been framed as stories per day at 10.3 stories a day. controversies, playing up the conflict and However, the smallest newspaper in the employing a game frame replete with study, the Guardian had the second boxing or war metaphors. Neither paper highest average at 9.3 stories a day resorted to this. Instead, the coverage suggesting that management considered provided an overview of the issues with the story important. The two smaller limited mediation. papers – the Telegram and the Guardian More exploration needs to be done had the most front-page stories, more to fully understand the differences. The than the competitive races in Toronto and next step would be to conduct interviews Calgary. This points to the smaller media with editors and journalists at these and news market influencing what is papers to determine the community published, suggesting that in markets contexts in which they operate. Smaller where the newspaper is not competing communities who feel a close affinity with for advertising revenues, the normative – their newspapers may not react well to or public service function of the news – is stories that are overly negative or that given primacy. ignore issues for a contest frame. As well, more elections need to be studied. For

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40 example, would a small market 97-112. Toronto, ON: Pearson newspaper covering an election in which Canada. the government was facing a potential upset still rely on normative rather than Canadian Newspaper Association. http://www.newspaperscanada.ca/ca market logic? Are there interesting tegory/canadian-newspaper- comparisons we can make between association. provincial and federal elections by one newspaper? There are many questions Goodhand, Margo. 02 October 2010. “Giving that remain unanswered, signaling that democracy a helping hand.” Winnipeg more research needs to be done at the Free Press. subnational level. http://www.winnipegfreepress.com/ ______special/democracy-project/giving- democracy-a-helping-hand- Works Cited 104199839.html.

Adams, Christopher. 2010. “Polling in Gibson, David R. 2011. “All the News That Canada: Calling the Elections.” In Fits to Print: Desk Competition for Mediating Canadian Politics, edited by Front-page Space at the New York Shannon Sampert and Linda Trimble, Times.” Sociological Forum. 26:2. 151- 168. Toronto, ON: Pearson 287-305. Canada. Landerer, Nino. 2013. “Rethinking the Logics: Androich, Alicia. 12 November 2012. “2012 A Conceptual Framework for the Media Players of the Year Shortlist.” Mediatization of Politics.” Marketing Magazine. Communication Theory. 23. 239-258. http://www.marketingmag.ca/news/ McGrane, David, Stephen White, Loleen media-news/2012-media-players-of- Berdahl and Michael Atkinson. 2013. the-year-shortlist-the-grid-65583. “Leadership, Partisan Loyalty, and Bain, George. 1994. Gotcha! How the Media Issue Salience; The 2011 Provincial Distort the News. Toronto: Key Porter. Election in Saskatchewan.” Canadian Political Science Review. 7:1. 1-12. Benzie, Robert. 07 June 2011. “McGuinty remains highly unpopular.” Toronto McKenna, Peter and Don Desserud. 2013. Star. “The 2011 Provincial Election in http://www.thestar.com/news/cana Prince Edward Island.” Canadian da/2011/06/07/mcguinty_remains_h Political Science Review. 7:1. 65-71. ighly_unpopular_poll_shows.html. Pétry, François and Frédérick Bastien. 2013. Cappella, Joseph and Kathleen Hall Jamieson. “Follow the Pollsters: Inaccuracies in 1997. Spiral of Cynicism: The Press Media Coverage of the Horse-race and the Public Good. New York, NY: during the 2008 Election.” Canadian Oxford University Press. Journal of Political Science. 46:1. 1-26.

Cox, Bob. 2010. “Covering Political Rosenthal, Jack. 22 August 2002. “What Campaigns: A Practitioner’s View.” In belongs on the front-page of the New Mediating Canadian Politics, edited by York Times” New York Times. Shannon Sampert and Linda Trimble, Sampert, Shannon, Linda Trimble, Angelia

Canadian Political Science Review Vol. 9, No. 1, 2015, 28-40

Wagner and Bailey Gerrits. Strömbäck, Jesper and Frank Esser. 2009. Fothcoming. “Jumping the “Shaping politics: Mediatiziation and Shark: Mediatization of Canadian media interventionism” in Party Leadership Contests, 1975 – Mediatization: Concept, Changes, 2012.” Journalism Studies. Consequences. Knut Lundby Ed. New York: Peter Lang. 205-223. Soroka, Stuart and Blake Andrew. 2010. “Media Coverage of Canadian Elections: Horse-Race Coverage and Negativity in Election Campaigns.” In Takens, Janet, Wouter van Atteveldt, Anita Mediating Canadian Politics, edited by van Hoof and Jan Kleinnijenuis. 2013. Shannon Sampert and Linda Trimble, “Media logic in election campaign 113-128. Toronto, ON: Pearson coverage.” European Journal of Canada. Communication. 28:3. 277-293.

Strömbäck, Jesper, Michael Karlsson and Taras, David. 1990. The Newsmakers: The David Nicolas Hopmann. 2012. media’s influence on Canadian Politics. “Determinants of News Content: Toronto: Nelson Canada. Comparing journalists’ perceptions of the normative and actual impact of Taras, David and Christopher Waddell. 2012. different event properties when “The 2011 Federal Election and the deciding what’s news.” Journalism Transformation of Canadian Media Studies. 13:5-6 718-728. and Politics.” In How Canadians Communicate IV: Media and Politics, Strömbäck, Jesper. 2008. Four Phases of edited by David Taras and Mediatization: An Analysis of the Christopher Waddell, 71-107. Mediatization of Politics. The Edmonton, AB: Athabasca University International Journal of Press/Politics, Press. vol. 13(3), 228-246. Trimble, Linda and Shannon Sampert. 2004. Strömbäck, Jesper and Daniela V. Dimitrova. “Who’s in the Game? The Framing of 2011. “Mediatization and Media the Canadian Election 2000 by The Interventionism: A Comparative Globe and Mail and The National Post.” Analysis of Sweden and the United Canadian Journal of Political Science. States.” The International Journal of 37:1. 51-71. Press/Politics, vol. 16(1), 30-49. van Dijk, Teun. 1991. Racism and the Press. London and New York: Routledge. Endnotes

1 Intercoder reliability scores are by no means uniform nor is the method by which they are conducted. Studies have been published citing intercoder reliability rates as low as 56% and as high as 85% (see for example Manganello and Blake 2012; Schnidermann et al. 2005). 2 A chi-square test was performed and a relationship was found between the newspaper and the use of the game frame X2(10, N=1254)=102.49, p=.000.