The hostile media effect refers to the tendency of partisans to view the media as biased against their side. Psyblog reviews a groundbreaking 1985 study that demonstrated the effect.
Robert P. Vallone and colleagues from Stanford University invited 144 Stanford undergrads who held a variety of views on the continuing Arab-Israeli conflict to watch some of the news coverage of the Beirut massacre (Vallone et al., 1985). The Beirut massacre was the killing of between 328 and 3,500 Palestinian and Lebanese civilians by Lebanese militia forces in September 1982...
Some of the participants recruited for the study were moderate in their initial views, others were specifically recruited from both the pro-Arab and pro-Israeli student associations. Each was asked for their views about the conflict, its history and where their sympathies lay...
All the participants then watched a series of news segments taken from US networks (NBC, ABC and CBS). Afterward they were asked to rate whether overall it was for or against Israel.Vallone found that the pro-Israel viewers judged the reports to have a marked pro-Arab bias. The pro-Arab viewers judged the videos to have a marked pro-Israel bias. The moderate viewers perceived a slight pro-Arab bias.
One possible explanation for the hostile media effect is that it results from a combination of two cognitive biases. The first bias, confirmation bias, is the tendency to prefer evidence that supports our own beliefs and hypotheses. The second bias, attribution error, refers to the tendency to overvalue dispositional explanations for the behavior of others, while undervaluing situational factors. For example, we make mistakes while driving; other people drive like jerks.
In Vallone's study, partisans on both sides valued information that confirmed their own views over information that undermined their views. Rather than treating the challenging information as an accurate reflection of facts on the ground, partisans assumed that reporter bias accounted for information that contradicted their opinions.
This isn't to say that media bias doesn't exist. Economist Jeremy Burke (2008) says that increasing competition among news outlets actually increases biased reporting as news agencies try to maximize viewership. Burke also reports that even non-partisan viewers actually prefer biased reporting.
I wonder if the preference for biased reporting, even among non-partisan viewers, reflects a preference for certainty over uncertainty and ambiguity. This is just a casual observation, but it does seem that political bloggers who offer a steady diet of certainties and stacked deck arguments are among the most popular bloggers.