How Interest Group Mobilization Explains Media Bias

In nearly every campaign and policy debate, at least one of the sides (and often both) make the claim that the news media is biased toward the other side. As previously discussed here, some political scientists measure media bias by comparing the citations of think tanks and advocacy groups in different media outlets with mentions of the same organizations in Congress. The idea is that reporters and legislators reveal their ideology by citing liberal or conservative groups.

In my new book, The Not-So-Special Interests: Interest Groups, Public Representation, and American Governance, I find that the patterns of media citations of interest groups reflect the composition of the advocacy community in Washington. One side in a policy debate gets more of an airing for their views when there are more mobilized groups on that side. Both the Washington print media and the national television news tend to rely most on the largest, oldest, and broadest organizations, whether they are liberal or conservative.

How does this help explain claims of liberal media bias? There are simply many more public interest groups representing liberal issue perspectives in Washington than those speaking on behalf of conservative issue perspectives (by my count, four or five times as many). The liberal groups are also, on average, larger and older. On some issues, there is a liberal issue group but no equivalent conservative one (although some see the cacophony of specialized voices as a weakness for liberals).

The large ideological difference in the population is reflected in organizational citations in media coverage. Liberal issue groups (including environmental and consumer groups) account for one-quarter of all advocacy group mentions on the television news; the conservative equivalents account for only 3.5% of the mentions. (Groups that represent occupations, identity groups, or other issue perspectives account for the remainder).

My counts do not include corporations and their associations, which vastly outnumber public interest groups and often have conservative views. When corporate interests are included, they represent a large share of media citations in policy debates (although one comparison found advocacy groups more prominent).

To see how this works in practice, look back at a USA Today story previewing 2011. The story cites Third Way, the Brookings Institution, the Heritage Foundation, the U.S. Chamber of Commerce, the AFL-CIO, the Committee for a Responsible Federal Budget, Environmental Working Group, and the Bipartisan Policy Center. The Groseclose/Milyo measure of media bias would compare its citations to those of members of Congress, perhaps finding it more likely to cite liberal groups.

Yet looking for differences across media and policymakers misses the larger picture: some organizations consistently gain more attention from both. I find that policymakers and reporters both amplify the voices of the same types of large, institutionalized organizations. A previous study of think tanks also found that organizational size and resources, rather than their ideology, determines their prominence.

The population of available groups is also consequential. Reporters call a cross-section of the types of groups that are accessible to them. If the organizations that they cite are indicators of media bias, the largest source of biases in coverage are differences in the relative mobilization of each constituency. Some social groups, economic interests, and issue perspectives generate more organized representation and they receive more media coverage.

9 Responses to How Interest Group Mobilization Explains Media Bias

  1. Andrew Gelman April 4, 2012 at 8:01 am #

    I would rephrase your title. Instead of writing, “How Interest Group Mobilization Explains Media Bias,” you could write “How Interest Group Mobilization Biases a Certain Measure of Media Bias.”

    • Matt Grossmann April 4, 2012 at 8:13 am #

      I don’t think that their measure is biased. It reflects the relative tendency of media outlets and members of Congress to cite the same groups. I also believe that which groups get cited in the media is a reasonable indication of the biases in coverage. My point is that the most important source of that bias is which groups most effectively mobilize to influence public policy.

      • Andrew Gelman April 4, 2012 at 2:22 pm #


        I disagree with the claim that “which groups get cited in the media is a reasonable indication of the biases in coverage.” Search blog for Groseclose for further discussion, but the short answer is related to what you wrote above, that citing a source is not the same thing as political alignment with that source.

        • Matt Grossmann April 4, 2012 at 7:06 pm #

          I re-read your post, but still disagree. I will try an analogy. Economists get quoted more than political scientists. I consider this a form of bias, whether it is an indirect measure of the influence of each discipline’s ideas on the news or just a direct measure of which type of experts get to share their views. This is separate from the question of whether the bias arises from the views of reporters or the attributes of economists (which is equivalent to the argument I have with Groseclose). I agree, however, that interest group citations are not the most important form of bias or the best indication of the political ideology of news coverage.

  2. Paul Waldman April 4, 2012 at 11:01 am #

    You’re also mistaken when you say, “some political scientists measure media bias by comparing the citations of think tanks and advocacy groups in different media outlets with mentions of the same organizations in Congress.” That’s not accurate. Two political scientists, in one methodologically laughable study, measured media bias that way. The fact that they got some undue attention for their study doesn’t mean that their study is the way media bias is measured. There have been HUNDREDS of other studies on media bias by political scientists and communication scholars using a variety of methodologies, nearly all of which are more revealing than the one study you cite. I hope that since you decided to address the topic of media bias you did what the authors of that study failed to do. It’s this obscure thing called a “literature review,” which one learns about in one’s first semester of grad school.

  3. Matt Grossmann April 4, 2012 at 11:23 am #

    My book does review some studies of media bias, though I focused on studies of which interest groups generate media mentions. I agree that the Groseclose/Milyo study did not effectively review the previous literature. That is why I cite (in this post) what I believe are the three most relevant previous studies about which think tanks and advocacy groups generate the most media mentions (Rich 2005; Danielian and Page 1994; Berry 1999). Their article does not cite these highly-relevant studies. I did not mean to imply that interest group mentions are the only, or the best, way to study media bias, but I can see from the comments that my title implies this. Obviously, there are many other aspects of media bias which have little to do with interest groups and which interest group mobilization does not explain.

  4. Noumenon April 5, 2012 at 12:48 am #

    Just wanted to say I’m impressed with the OP’s ability to defend himself professionally in the comments.

  5. Paul Rini April 12, 2012 at 8:13 am #

    The subject of media bias is one that I take very seriously, because if you can’t trust the news for straight information , then who can you trust for the truth?

  6. Liberal Bias April 12, 2012 at 10:32 am #

    I have become increasingly concerned about unfair reporting by the mainstream news media. Media have tremendous power in setting cultural guidelines and in shaping political discourse…