The Root of Media Bias

The media. Some complain certain news outlets are too liberal, others too conservative. Regardless of which way a news outlet leans, is it the media themselves that drive bias or is it the demands of their audience?

“I frequently hear people complain about how biased the media is, and I have always wondered why the news outlets that are perceived to be the most biased also enjoy the largest viewership/readership,” said Cagdas Agirdas, assistant professor of economics.

Could it be, Agirdas wondered, that people enjoy receiving their news from a source that agrees with their own views? And if that is true, wouldn’t media companies maximize profits by appealing to their readers’ political preferences?

These were the questions Agirdas explored in his study “What Drives Media Bias? New Evidence from Recent Newspaper Closures,” which was published in the Journal of Media Economics.

Aware that since the advent of the Internet many newspapers have been forced to close due to declines in revenue, Agirdas chose to look at the behavior of surviving newspapers.

“I thought that media bias should not change if the bias was driven by the political preferences of the owners/editors of these newspapers,” he said.

On the other hand, if media bias is demand-driven, a surviving newspaper could expand its reader base by moderating its bias to reach out to the readers of the closed rival newspaper.

What did he find?

“I found significant moderation of bias after the closure of a rival newspaper,” said Agirdas. “This is interesting, because many people blame media outlets for their bias, instead of the appetite for bias among readers.”

For the study, Agirdas used a large panel data set of newspaper archives for 99 newspapers from 1990–2009, which included 24 newspaper closures.

To measure bias, he looked at the amount of reporting on unemployment news for two reasons. First, studies have found that a change in the unemployment rate is a strong predictor of the success of an incumbent president’s political party. And second, there is a correlation between a newspaper’s partisan endorsement pattern and its reporting of unemployment news.

He found that conservative newspapers reported 17.4 percent more unemployment news when the president was a Democrat before the closure of a rival newspaper in the same media market. This effect was 12.8 percent for liberal newspapers. After the closure, these numbers were 3.5 percent and 1.1 percent respectively. (Moderate newspapers did not display statistically significant bias either before or after closure.)

Agirdas said these finding support the idea that media bias is demand-driven, as the surviving newspapers appeared to temper their bias to gain the readers of a closed newspaper.

“From time to time, there are calls to split up ownership of media outlets to avoid monopolization,” said Agirdas. “However, if the media outlets are simply serving the preferences of their viewers, such an approach would not create better outcomes. Two newspapers owned by the same company might disseminate completely different political views if they are trying to appeal to readers of opposing views.”