Information Technologies and Omnivorous News Diets over Three U.S. Presidential Elections

ABSTRACT

Technology convergence and rising expectations for interactivity have had a significant impact on the news diets of U.S. voters. While television may appear to be the most important single media in this system of political communication, for a growing portion of the population, news diets are defined by combinations and permutations of secondary media. What explains the changing distribution of primary media choice and the dramatic rise in secondary media? We offer a theory of omnivorous information habits to help explain the rising number of people who make active choices to get political news and information from several media technologies, sourced from multiple news organizations, and then engage with news and information through varied interactive tools. Data from 2000, 2004, and 2008 demonstrate not just the growing importance of secondary media, but the importance of the Internet in particular. Indeed, elections have become occasions in which people make significant changes in their information diets.

Keywords:

• Elections
• information habits
• Internet
• multimedia information retrieval

Acknowledgments

For access to datasets, the authors are grateful to Lee Rainie, Director of the Pew Internet and American Life Project. For feedback on this manuscript, the authors would like to thank Lance Bennett, Andrew Chadwick, David Domke, Deen Freelon, John Gastil, Patricia Moy, Michael Schudson, Keith Stamm, and Jennifer Stromer-Galley.

Notes

1. Authors' calculations based on Pew Research Center for the People and Press data from 2000, 2004, and 2008.

2. For replication data, readers are invited to visit the author's Web page at http://faculty.washington.edu/pnhoward/.

3. This trend may have been subject to interviewer effects, as some interviewers would have probed more deeply than others. However, interviewers were given the same set of instructions each year, and the survey was organized by the Princeton Research Associates each year. Interviewers were required to note a primary media choice, and were asked to probe for a secondary media choice. They were not instructed to rotate options, however, and offering television first may have actually positively biased the response rate for that answer option.

4. See, for example, Bennett and Iyengar (2008) or the special collection of Political Communication edited by Michael Schudson on the contributions of cultural sociology to the study of political communication.

5. See http://www.pewinternet.org.

6. See http://www.people–press.org

7. Questions for each year of data collection changed, sometimes slightly, and sometimes more significantly. In constructing the indices, we attempted to include equivalent questions wherever possible. The Media Choice index for 2000 consisted of the answer to this question: “How have you been getting most of your news about the presidential election results and questions about who won? From television, from newspapers, from radio, from magazines or from the Internet.” In 2004, the Media Choice index consisted of responses to the following question: “How have you been getting most of your news about the presidential election campaign? From television, from newspapers, from radio, from magazines, or from the Internet?” In 2008, the question became: “How have you been getting most of your news about the November elections—from television, from newspapers, from radio, from magazines, or from the Internet?” See the Appendix for lists of the Source Choice Index and Interactive Index questions.

8. Data about voting habits and Internet use is known to suffer from an over-reporting problem due to the social desirability of affirming good habits with survey interviewers. This research assumes that such over-reporting is present but constant between elections.

9. In this example from the election year in 2000, the odds = 0.070(Constant) * 0.964(Age)* 0.739(Female) * 0.774(African American) * 0.527(Asian American) * 0.991(Other) * 2.151(Income $50k or More) * 2.711(Education BA or More) * 1.551(Republican) * 1.402(Democrat) * 14.957(2004 Election Year) * 18.477(2008 Election Year) and since e⁽⁰⁾ = 1, the odds = 0.070(Constant) * 0.964(30) * 0.739(0) * 0.774(1) * 0.527(0) * 0.991(0) * 2.151(0) * 2.711(1) * 1.551(0) * 1.402(1) * 14.957(0) * 18.477(0).

10. The survey evidence presented here was collected around election time, when public attention turns to politics. Along with the growing omnivorousness among a portion of the public is a segment that reports getting no news. On an average day—during the rest of the year—some 19 percent of survey respondents in 2008 reported getting no news (Pew Research Center for the People and the Press, 2008).