Pew Report On Bias Among Journalists

by on May 25th, 2004

Editor & Publisher reports on a Pew Research Center poll which supposedly shows an inordinate liberality among media correspondents.

However, a consultation with the actual report should counsel against the glib conclusions drawn by many bloggers that this report confirms the existence of a liberal bias in the mainstream media.

First, anyone who wants to claim this is even proof of a liberal bias among reporters are confronted with obvious methodological deficiencies–small sample size, selection bias, dependence on political self-ascription rather than more objective indicators, and so forth.

Second, even if you accept the report as proof of a bias among reporters, it does not follow that there is a bias in reporting. The attempt to draw such an inference simply ignores the many other structural pressures that might tend to bias news content in a conservative direction.

Consider just a couple of obvious economic pressures along these lines:

1. Media corporations are businesses and as businesses they will seek to maximize profits. One way for, say, a newspaper to goose profits is to create news content that audiences want to read (and therefore buy). If the majority of news consumers are conservative, one would expect the majority of news content would be geared toward attracting such readers.

2. Media companies make most of their money through advertising revenue. But advertisers generally won’t want their product associated with news content that tweaks the political views of those in its target market. Media companies know this, and will adjust their news content accordingly.

In short, to argue that there is a reporting bias merely by appeal to bias among reporters is to ignore a vast number of institutional structures that are likely to influence news content.

The claim that mainstream media exhibit some kind of reliable political bias in one direction or the other is an incredibly strong claim that is itself politically charged and obviously subject to political bias. Getting a handle on what objective phenomena would be entailed by such a hypothesis would be a pretty serious undertaking, one ill served by impressionistic musings, glib inferences drawn from cherry-picked examples, and methodologically suspect studies. So far that’s about the only kind of thing those who holler “media bias!” have given us. The question deserves more careful attention.

Tadlow Windsor II