Continued from Page 1…
Into the Homestretch: A Rough Ride for Party Leaders in the Horserace, June 21
As we head into the last week of the 2004 federal election campaign, similarities and differences in coverage by the country's leading newspapers are coming into sharper relief.
Thus far, the papers have been unanimous in giving a rough ride to Liberal leader Paul Martin. The 'net tone' (per cent of positive mentions minus per cent of negative mentions) of his coverage has been negative in all seven papers content analyzed by McGill's Observatory on Media and Public Policy, and particularly in Le Devoir (-25), the National Post (-20), and the Calgary Herald (18). Likewise, net tone of coverage for the Liberal Party is negative in all seven papers, although most negative in Le Devoir, the Post, and the Herald.
Conservative leader Stephen Harper's coverage has been mixed. Throughout the campaign, he has received positive press in the Calgary Herald (+4) and the National Post (+16). Even in papers where Mr. Harper's coverage has been negative on the whole, it has still been better than Mr. Martin's. The same is true of stories about the Conservative Party: favourable coverage from the Herald and the Post and less negative coverage than the Liberals in the other papers.
These figures reflect cumulative coverage through the first three weeks of the campaign. In recent weeks, Mr. Martin's coverage has softened somewhat, while Mr. Harper's has become more negative. Still, Bloc Quebecois leader Gilles Duceppe is the only leader with positive press in at least three papers.
The papers have also tended to print more horserace than issue coverage. This is especially true of the Globe and Mail and La Presse: the gap between horserace and issue coverage (per cent horserace articles minus per cent issue articles) is 30 per cent and 20 per cent of election articles, respectively.
The two outliers are the Calgary Herald and the Vancouver Sun. Issues have dominated election stories in both papers, but only by small margins.
As the campaign moves into the homestretch, the dominance of horserace coverage will increase. Issues will be overpowered by a frenzy of leader travel and stories detailing leader gaffes. With missteps on Air Canada and strongly-worded press releases on child pornography from the Conservatives plus accusations aimed at NDP leader Jack Layton that he is willing sacrifice principle for power by bedding down with a Conservative minority, weekend election coverage certainly provides ample fodder for horserace stories.
Debates and the media, June 18
Most voters don't watch debates, they learn about them through newspapers and newscasts. But that doesn't mean that the effect of debates is entirely determined by media.
Coverage of party leaders in the aftermath of the 2004 French and English language debates demonstrates this markedly. Following the French debate, Bloc leader Giles Duceppe's 'net coverage' (per cent positive mentions minus per cent negative mentions) increased from 0 to +15. He was clearly the winner, at least where tone of news coverage is the measure. The other primary beneficiary of the French debate was Liberal leader Paul Martin, whose net tone moved up almost as much, albeit starting at a much lower point (-12 to 0). Conservative leader Stephen Harper's coverage was essentially unchanged following the French debate; NDP leader Jack Layton's net tone slid from -8 to -10.
These trends continued the second night, enhanced by an increase in horse race coverage from 50 per cent to 70 per cent of all election articles this week. Mr. Duceppe's coverage remained positive; Mr. Harper's coverage was generally neutral; Mr. Layton's coverage was slightly more negative; and Mr. Martin's net tone continued to rise. In fact, Wednesday was the first time since the campaign began that Mr. Martin's coverage was positive rather than negative. As far as media were concerned, Mr. Martin was in the ascendancy while Mr. Harper and Mr. Layton were falling.
This changed dramatically the next day. Coverage of leaders on Thursday returned almost to the way it was before the debates. In particular, this meant a sharp drop in the tone of Mr. Martin's mentions while Mr. Layton and Mr. Harper improved substantially.
Why did this happen? Before the second day's coverage an IPSOS poll was released showing that voters disagreed with media assessments of leader performance. Voters were relatively less impressed with Mr. Martin, and more approving of Mr. Harper and Mr. Layton. One strong possibility is that media corrected its initial impressions in light of public opinion.
Contrast this with what we see in trends in party coverage (of which there is twice as much than leader coverage). Here, there has been a relative lack of movement over the week. This illustrates the gap that often exists between leader and party assessments. It also fits with recent trends in public opinion, where Mr. Harper's lead remains small but apparently stable.
The debates had a clear effect on the coverage of leaders, then, but relatively little effect on parties' coverage. And even shifts in leaders' net tone appear to have been temporary. Only two days after the debate, the campaign is continuing almost where it left off.
What will the important issues be? June 15th 2004
There is much speculation about the issues that will be important in tonight's English-language debate. The sponsorship scandal, and government accountability more generally, are sure to emerge as issues, as in the French debate. But what other issues are on the agenda? What are the policies we can expect party leaders to address?
Recent news coverage may provide an answer, or at least some clues. McGill's Observatory on Media and Public Policy has been monitoring election coverage since before the writ was dropped. The data show that coverage of issues has changed dramatically over the campaign. They also suggest which issues may be most important in this evening's debate.
While government accountability dominated election coverage in the first few days of the campaign, it has received less and less attention over the campaign. In the first week, health care and taxes emerged as salient issues. These were followed though to a lesser extent by international affairs and environmental policy.
Last week, we noted the apparent rise in social issues (same sex marriage and abortion) on the campaign agenda. Data show that over the past week these issues were receiving the most coverage, displacing health, taxes, and accountability. Indeed, about one in every three campaign-related articles dealt with social issues.
While much commentary has focused on Conservative momentum leading into the debate, then, media coverage points to another possible dynamic. The rise of social issues on the campaign agenda was a function of the recently-successful Martin campaign, trying to paint Conservative policy as socially regressive. The decline of government accountability, conversely, suggested that the Harper campaign was temporarily faltering. At least, it was not leading the campaign as it did in the prior week.
What does this mean for tonight's debate? This debate may play a critical role in defining the issue agenda for the second half of the campaign. We can expect to see Martin talking about social issues, and Harper talking about accountability. Indeed, coverage over the past weekend shows an increasingly tight competition between accountability and social issues.
The extent to which one issue dominates the other in the debate may have a significant effect on the issues voters use to judge parties, and as a consequence the party they vote for on June 28th.
Martin: Less coverage, less negativity, June 11
Polling data into the third week of the 2004 election campaign has indicated the Liberals losing their lead and the Conservatives gaining ground, but a contrary trend is emerging in media coverage of the major party leaders.
After receiving great volumes of heavily negative attention at the start of the race, Liberal leader Paul Martin is now being mentioned in fewer stories, and fewer negative stories. He appeared in 67 per cent of all election stories during the first week of the campaign, and the 'net tone' of his coverage (per cent of positive mentions minus per cent of negative mentions) was -20. In the third week, he has appeared in just 34 per cent of articles with his net tone improving to -12.
Interestingly, Mr. Martin's party is not faring as well as its leader. The Liberals have been mentioned in more than 80 per cent of election stories every week since before the writ was dropped. The tone of its coverage, after improving in week two (when the party released its platform), is now back at -24, almost as negative as it was at the start.
The gap between party and leader may be a result of Mr. Martin's absence from the campaign when he attended G-8 meetings and D-Day ceremonies.
Conservative leader Stephen Harper, meanwhile, has been mentioned in 76 per cent of election stories in the third week of the race compared to 59 per cent in the first week, and the tone of his coverage has deteriorated from neutral to -6.
The Conservative party is on the same track as its leader, receiving more and increasingly negative attention. That may forecast a downward shift in the polling fortunes of the Conservatives. Mr. Martin's numbers may herald a slight leader-led recovery for the Liberals.
McGill's Observatory on Media and Public Policy has been analyzing election coverage in seven major Canadian newspapers since the week before the writ was dropped.
The Evolving Issue Agenda, June 8
The contest between Canada's two leading parties, Liberals and Conservatives, has developed in unexpected ways as the federal election campaign ends its second week. Not only are issues becoming increasingly important in what many expected would be a content-free campaign, but the battles lines between the parties are sharply drawn.
In recent Canadian electoral history, it is rather rare to see clear distinctions between competing parties on major issues. In this race, at least so far as the media is concerned, we have classic divisions between the parties on the two most fundamental sets of issues.
According to analysis of seven leading Canadian newspapers by McGill's Observatory on Media and Public Policy, the Conservatives are increasingly portrayed as fiscally conservative (favoring tax cuts and lower government spending) and socially conservative (opposing recognition of same-sex marriage and abortion, among other issues). The Liberals, promising new social programs such as national day care, are increasingly viewed as fiscally and socially liberal. The actual policies of the leading parties may not always warrant the sharp contrast evident in the newspapers. Liberal and Conservative positions on abortion, for instance, are rather similar. But the perception of fundamental differences is unmistakable.
It is also interesting to note that these stark oppositions may not serve voters. Many are socially liberal and fiscally conservative, or vice versa. They may feel cross-pressured and frustrated by the current alignment of parties and issues. It is not at all clear where these voters turn on election day if the current alignment holds.
Our analysis shows as well that issues are far more important now in the middle of the campaign than they were in the pre-campaign period. The sponsorship scandal and stories about government accountability have virtually disappeared. Health care, which was also expected to be crucial in the campaign, is no longer receiving much attention. On the other hand, environmental issues are moving to the fore, an indication of the relative success of the Green Party.
In sum, as we approach the mid-point of the campaign, the 2004 race is shaping up as a potentially defining moment in Canadian electoral history.
A question of bias, June 3
The issue of media bias is seldom hotter than during an election campaign. Accusations of bias usually arise in response to particular stories in particular papers. But a better sense of the tendencies of print outlets can be had by close comparisons of their content over time.
McGill's Observatory on Media and Public Policy has studied positive and negative mentions of the major parties and leaders in the news and commentary of seven leading dailies through the first 10 days of the campaign. The results show marked differences among the papers.
The Liberals are getting beaten up everywhere, but especially in the National Post where the average net coverage is 37 per cent. This percentage is based on the per cent of positive mentions minus the per cent of negative mentions -- a score of 0 per cent would be perfectly neutral. Le Devoir (-32 per cent), La Presse (-30 per cent), and the Calgary Herald (-25 per cent) were close behind the Post. The Toronto Star and the Globe & Mail were at -21 per cent and 19 per cent respectively. The Vancouver Sun was kindest at 10 per cent.
The Conservatives are getting better press than the Liberals across the board. The Calgary Herald and the National Post led the way with +12 per cent and +6 per cent respectively. The Toronto Star (-12 per cent) and Le Devoir (-11 per cent) were most critical of the party while the Sun, La Presse and the Globe were basically neutral.
As one would expect, coverage of the leaders closely mirrored coverage of the parties. Le Devoir was particularly hard on Liberal leader Paul Martin (-32 per cent) and the Post was both particularly warm to Conservative leader Stephen Harper (+12 per cent) and hard on NDP leader Jack Layton (-32 per cent).
Mr. Martin's coverage was more negative than Mr. Harper's across the board. The greatest differential in coverage of the leaders was at the Post (Martin 20 per cent and Harper +16 per cent). The Globe has distinguished itself with mildly negative treatment of all four leaders. The Vancouver Sun is close to neutral or mildly positive about parties and leaders.
Campaign coverage differs across newspapers, then. Whether these figures are evidence of bias is another matter. While the word 'bias' has negative connotations, differences across newspapers can be seen to represent a healthy diversity of opinion among media outlets. Whether one paper is right or wrong in giving a particular party positive or negative press is a judgment best left to the reading public.
Note that these results include editorials and opinion pieces, and that the majority of news coverage in these papers is neither positive nor negative, but neutral. To the extent that positive or negative coverage exists, however, it is clear that important differences exist.
The campaign takes a turn, June 2
The election campaign has taken a turn in recent days. We have seen a shift in campaign coverage, and in the way in which the election is being framed.
For the first week, the campaign was essentially about Paul Martin and the Liberal Party. Would Martin be able to deal with the sponsorship scandal? What were his major policies? How are other parties reacting to the Martin campaign? These were the questions that dominated coverage coverage of a campaign that was essentially about Martin winning or losing. In recent days, however, there has been less coverage of how other parties are reacting to Martin, and more coverage giving other party leaders prominence of place.
The shift appears to have taken place with the release of the Ipsos poll at the beginning of this week, a poll suggesting that Conservatives are gaining in Ontario. Media discussion of minority governments has consequently increased. So too has the emphasis on other parties and leaders.
The shift can be seen by examining the number of articles in which party leaders are mentioned, and the proportion of those articles in which each party leader is mentioned first. In the first week of the campaign, 64 per cent of these articles mentioned Paul Martin first. When other leaders were mentioned, they came after Martin.
This week, the proportion of articles mentioning Martin first has dropped to 50 per cent. Conservative leader Stephen Harper, on the other hand, has moved from 18 per cent to 30 per cent of first mentions.
The data illustrate a shift in the structure of campaign coverage, a campaign that is now framed less in terms of the governing Liberals. This is not to say that Martin does not still dominate coverage: in half of the articles mentioning a leader first, Martin is that leader. Nevertheless, a shift away from Martin appears to be underway.
Note also that these data only partly reflect the gap between coverage of Martin and other party leaders, particularly Harper. To date, 20 per cent of Martin mentions have been negative and only 2 per cent positive (others are neutral). For Harper, 9 per cent have been negative and 10 per cent have been positive. So not only is Harper currently gaining coverage, he's getting better coverage.
A campaign about the issues, May 31
It appears as though the 2004 federal election campaign may be a war of ideas after all.
When the writ was dropped, it was by no means certain that the contest would generate substantive discussion about policy and issues. Indeed, Canada's major newspapers were initially preoccupied with the Liberal sponsorship scandal and questions of political accountability. According to content analysis of leading Canadian dailies by McGill University's Observatory on Media and Public Policy, these themes accounted for 50 per cent of all election articles at the start of the race.
Since then, coverage has shifted from process to policy -- from how parties and governments do things to what they do. In the middle of the first week, health care emerged as a significant theme in the campaign, followed closely by taxes and municipal concerns. Those issues remained front and centre throughout the week.
By last Friday, only 28 per cent of stories dealt primarily with process while 47 per cent focused on policy. Process articles -- the sponsorship scandal, leadership and accountability issues, voter turnout -- made a slight comeback over the weekend, hitting 31 per cent compared with 38 per cent for policy pieces, but the campaign today remains one of policy and ideas.
While leadership and accountability are obviously important concerns to voters, the predominance of policy articles is a positive development. If elections are supposed to give governments mandates to make policy, platforms need to be aired and analyzed. The emerging salience of health care and taxes is especially intriguing. The trend began with Liberal leader Paul Martin's introduction of the Liberal heath-care package and continued with all parties taking stands on taxes. The result is a classic and important competition between taxes and spending. It is a debate similar to and perhaps foreshadowed by the recent Ontario election. It is no coincidence that Ontario Premier Dalton McGuinty's health premium has played an instrumental role in the federal campaign.
Methodology and early observations, May 26
Media is generally assumed to be an important factor in election campaigns. Journalists, academics, and politicians routinely cite coverage as influential in the rise and fall of leaders, the polling performance of parties, the emergence of particular issues during campaigns, even the outcomes on voting day.
Unfortunately, media influence is more often presumed than demonstrated. There has seldom been any systematic investigation of how Canadian news outlets cover campaigns.
To help fill this gap, McGill University's Observatory on Media and Public Policy will analyze the contents of seven major Canadian dailies for the duration of the 2004 federal race.
Each day, a team of coders will scrutinize the main news section of The Globe and Mail, National Post, Toronto Star, Calgary Herald, Vancouver Sun, La Press and Le Devoir. They will note all articles relating to federal politics, including reportage, analysis, opinion, and editorials. This content will be coded for mentions of issues, parties, and leaders, as well as positive or negative (or neutral) tone. The precedence of these mentions and their prominence within the paper will also be noted, along with other factors.
This method of coding newspapers has proven effective in demonstrating trends in coverage in other jurisdictions. It can indicate which parties or leaders are getting positive and negative treatment, and which issues are dominating the campaign. It can show which parties are associated (positively or negatively) with leading issues. It can suggest which parties are driving coverage of the campaign and which are responding to the initiatives of their opponents.
The coding will also demonstrate how treatment of parties, leaders, and issues varies from newspaper to newspaper, and from region to region.
Finally, by comparing trends in positive and negative coverage of parties with opinion polls, it is possible to get a sense of whether the press is leading or following public opinion in the campaign. The net result of this content analysis should be an accurate and valuable perspective on the role of print media and individual print outlets in a major election campaign.
Coding began on Monday May 17th, a week before the election was called, and results already provide an early indication of how the campaign is shaping up. Based on 523 articles published between the 17th and 26th of May (inclusive), two noteworthy trends emerge.
Second, this negative coverage is linked to the significance of certain issues in pre-election coverage. Government accountability and the sponsorship scandal have been mentioned in 21% of all articles; indeed, this was the leading issue before Martin's health care proposals were laid out yesterday. Healthcare and taxes are the other two big issues, with 23% and 14% respectively.
These trends will change over the next few weeks as the campaign progresses and party platforms take shape, of course.
For now, the overall picture is one of a campaign that began with a largely negative focus on the Liberal Party. The success of Paul Martin's campaign may be dependent in large part on its ability to change the focus of media and public attention towards issues the Liberals can win. Yesterday's health policy announcements may reflect exactly this strategy, successfully making healthcare a more prominent issue than government accountability.