Negative vs. Positive

What also emerged, in debate coverage and elsewhere, was a clear predilection toward negativity. More than half of all stories contained twice as many negative assertions as positive. This finding stands out especially because during this election cycle, citizens have repeatedly told the media they are tired of the negative nature of campaigns. Further, studies of press coverage in general rarely find such great disparity between positive and negative coverage.

For a story to be considered anything but neutral, the positive or negative statements within it must outnumber each other by at least two-to-one. For example, in a story assessing how a candidate performed in a debate, there would have to be eight negative statements for every four positive for the story to be considered negative.

For this study, the 29% of stories considered straight news accounts of the candidates' actions or statement were not measured for tone, though a sample test suggested that they, too, were not particularly neutral. Nevertheless, even if all of these straight news stories were added to the neutral column, the number of negative stories would still be extraordinarily high and outweigh positive by nearly three to one.

Interestingly, the negative coverage did not simply occur because both candidates have struggled to pull ahead. The press was actually more negative when covering the candidates' character (77% negative) and policy issues (51%) than when covering campaign internals (45%).

Perhaps one reason is that this study, unlike many in the past, includes the Internet—which accounted for nearly a third of the stories—and many web sites are far harsher in their coverage than traditional print and broadcast outlets. For instance, the three Internet sites studied that include mostly web-only content—National Review Online, Slate and Salon—were notably more negative than average. At least six-in-ten of their stories carried a negative tone (63% for Slate, 61% for the National Review, 60% for Salon). Contrast that to 41% for televisions and 53% for newspapers.