The second phase of this study was a survey of the public by the Pew Research center for the People and the Press to see if the narrative themes about the candidates in the press were having an impact on public attitudes. While the public perceptions of the candidates character traits do not match precisely what the press has published about them, further analysis of the data by the University of Missouri suggests the coverage is, apparently, having an impact.1
Despite the negative coverage, the strongest associations people have with President Bush are positive.
About half of the public (48%) associates the idea of a "strong and decisive leader" with Bush rather than Kerry. And even a slightly higher percentage (53%) say the positive theme Kerry has most tried to project, that of a courageous person who won't back down from a fight, best describes Bush, (just 15% of the public identified this more with Kerry.)
The character theme that the press has honed in on the most, that Bush is stubborn and arrogant, does seem to have registered. Nearly half of the public (44%) say that trait best describes Bush (just 13% associate it more with Kerry).
The other negative character theme that the press has focused on, that the President lacks credibility, did not register much with the public by mid June. Overall, just 33% of the public associated this more with Bush. As it turns out, nearly as many (28%) feel it describes Kerry best, even though in the press coverage it was clearly tagged to Bush.
What else did the public think of Kerry?
Overall, its impression of him was indistinct. The only theme that more of the public saw as best describing Kerry rather than Bush was that he was a flip flopper. Interestingly, this is the one theme about Kerry that came at the public from all three of the major media categories that we examined-press coverage, advertising and late night comedy. In all, a little more than a third, 36%, attached this theme to Kerry more than Bush.
In most cases, partisanship dominates these specific perceptions of the candidates. For example, 60% of Democrats say the phrase "he twists the facts and misleads people" better describes Bush while a similar number of Republicans (55%) say it better applies to Kerry.
But even many Republicans are willing to acknowledge Bush's stubbornness. About as many Republicans say that negative trait applies to the president (22%) as say it better describes Kerry (24%). On the other hand, a significant minority of Democrats (21%) say the phrase "strong and decisive" better describes Bush; that is not much fewer than the number of Democrats who say it applies more to Kerry (31%).
The character themes are mostly lost on swing voters-those who are either undecided at this point or say they still may change their vote. Majorities of swing voters said most of the themes did not particularly apply to either candidate. The exceptions are traits-both positive and negative-associated with Bush: 44% say the phrase strong and decisive applies more to Bush; 44% say the phrase "personally tough" better describes the President; and about the same number of swing voters-43%-believe the phrase "he is stubborn and doesn't admit mistakes" also better characterizes the President.
Having trouble sticking with a decision is that only character trait that at least 30% of swing voters apply to Kerry. At the same time, however, other negative aspects of Kerry's personality seen in the coverage have little to no resonance among swing voters. For instance, just 9% of swing voters say the phrase "he is a wealthy elitist" better describes Kerry-more than twice as many (24%) say it applies more to Bush.
The Media's Influence
Why does the public largely associate positive character traits with Bush if coverage has been so negative?
To assess that, researchers at the University of Missouri took a closer look at the polling data. First they looked at responses to the character themes overall to see if paying more attention to news media was having an effect. Next they looked at each character theme see which ones in particular were impacted by news consumption.
What they found was that, first, news consumption indeed was influencing how people saw the candidates. The more attention they paid to the news, the more they tended view the candidates they way they were portrayed in the press.
What's more looking at the themes individually, there is statistical evidence the coverage is having a mixed effect theme by theme.
First, the press coverage so far is not having any measurable effect on people's perception of President Bush.
For Kerry, the coverage tends to persuade people both that Kerry is a tough guy who won't give up and that he is a rich elitist.
The biggest factor influencing opinions of the candidates, the analysis shows, was naturally people's political orientation and candidate preferences.
But when these political preferences were controlled for-along with demographic variables like race, gender, education and age-the researchers were able to isolate that people who consume more press coverage about the race are more likely to think Kerry is both tough and an elitist. There is no measurable effect on their perceptions at this point of the President on his major character themes.
The influence of news coverage, moreover, may be more important in the states that will play the biggest role in the election, the so-called "battleground states," those where the electorate is fairly evenly divided and thus where the candidates are spending most of their time and money.
The reason is that people in these states are paying more attention to campaign coverage than elsewhere, according to the survey data. And while the candidates are spending more on ads in these places, the ads here are failing to have a measurable impact on voter attitudes about the character themes, with one exception: people who see more Bush ads in battleground states are more likely to agree with the theme that Bush is strong and decisive.
For now, it is less clear what impact the press coverage or even ads are having on undecided voters. They are less engaged, but that will likely change as Election Day draws nearer.
Incidentally, in non-battleground states, the only statistically significant effect of ads is that those who see more Bush ads tend to say Kerry flip flops.
Some critics may leap to the conclusion that these findings confirm their suspicions of a liberal bias in the press.
Others may argue that the coverage of the President has been more negative than positive since March because the news events have turned against him, from Richard Clarke challenging the President's pre-war thinking to the revelations of prisoner abuse and the continuing instability of the situation in Iraq. An incumbent president naturally reflects the news of the day.
Had this study been done during the war itself or in the months leading up to it, these observers might argue, the coverage would have been more positive for Bush.
We cannot answer this debate.
A similar study four years ago, however, found that the coverage was more positive for Bush than for his Democratic rival, Al Gore. That raises several possibilities, including that the tenor of coverage simply follows events. Or maybe there is a tendency for the press to be more skeptical about the devil it knows best than the devil knows least.
1. While seven threads were analyzed in the press coverage, we surveyed only six. The idea of Kerry as an extreme liberal was omitted, as the public would naturally attach that more to Kerry than to Bush because of their Party identifications.
2. To test what features were associated with people identifying each narrative theme with the candidate it had been linked to in the press, the following analysis was performed: The six character themes were analyzed individually. After controlling for demographic variables and voter preference, researchers looked at the impact of ad exposure and attention to news for each using a hierarchical multiple regression. See methodology for more details.