![Sample our Politics & International Relations journals, sign in here to start your FREE access for 14 days]({"type":"image" , "src" : "/sda/5947/TOC-Subject-Sample-2021-Politics-International-Relations.jpg"})
Abstract
Can presidents influence news coverage through their press conferences? Scant research has explored this question leaving two possible answers. On the one hand, presidential news management efforts, combined with norms of journalistic professionalism and the cost of producing news, suggest that the nightly news will cover presidential press conferences. On the other hand, the costs of delivering press conferences espoused by some scholars insinuate that press conferences will have little impact on news coverage. To determine whether the press conference influences news coverage, I use plagiarism detection software to assess the propensity of television news to incorporate the president's rhetoric into stories that cover the president's press conferences. I find that news reports on the press conference rely heavily on the president's words, indicating that it is an important event for presidential influence of the news media and perhaps eventually the public.
Acknowledgments
An earlier version of this article was delivered at the 2011 annual meeting of the American Political Science Association in Seattle, Washington. Thanks to Paul Collins for help with the WCopyFind software and to Jun-Deh Wu and David Kottwitz for assistance with data collection.
Notes
1. Barrett (2004, p. 350) excludes responses to questions in press conferences from his measure of going public because, he asserts, they are reactive, not necessarily part of his policy agenda, and that others “will probably be less concerned with these remarks because they are made at the urging of another individual, not as part of a prepared presidential statement.”
2. Many thanks to an anonymous reviewer for pointing this out.
3. Although a likely indicator of the president's top priorities, CitationSmith (1990, p. 84) cautions that “we cannot automatically assume that the opening statement is the presidential agenda.” Indeed, there are numerous instances where presidents did not offer a lengthy opening statement, but still had a policy focus in those press conferences.
4. We found each press conference and copied and pasted the text of the press conferences from the American Presidency Project (http://www.presidency.ucsb.edu/). We searched only “Oral: News Conferences” and “Oral: News Conferences—Joint.” This approach is less inclusive than others, such as CitationRottinghaus (2010), who code a public remark as a press conference “if press officials who were able to ask any questions attended the event” (p. 260). I do not consider informal, brief question and answer sessions to be press conferences. My numbers appear to differ slightly from Kumar's (2007). Given that her data are not monthly, and this analysis begins in March, I cannot determine precisely by how many.
5. Our sample included two primetime press conferences (out of 13 total for our time period) and 17 press conferences that were televised on CBS. When CBS televised the press conference, our Lexis-Nexis search produced a transcript of the actual press conference, along with opening remarks by the anchor or White House correspondent.
6. Obviously, I am speculating here as I have no evidence. This could be a good paper for another to write.
7. See http://plagiarism.phys.virginia.edu/Wsoftware.html. The analysis uses the default settings that, among other things, match on a minimum of six consecutive words. This avoids the possibility that the plagiarism detection software is matching on more common words or phrases, such as President Obama, the president of the United States, or health care reform. To counter the concern that the phrase matches are solely coincidental, I collected plagiarism scores for 10% of the data that match on a minimum of 10 consecutive words (see CitationCorley, Collins, & Calvin, 2011, footnote 7). These scores correlate at .983 and do not alter the substantive results presented in this article. The average score drops by just over 1 percentage point nevertheless.
8. It is also a more conservative estimate. There are times when the topic of the president's press conference generates news coverage, even though it did not produce a plagiarism score. For instance, the president could talk about the Persian Gulf war in a press conference, and the news story might reference the president's news conference as a lead-in to a story on the Persian Gulf war but not quote the president or repeat the president's words directly. As a point of comparison, my sample produced 116 (out of 135) press conferences that were referenced on the CBS Evening News, but only 84 (out of 135) generated a comparison score.
9. This discussion and presentation of dependent variables is attributable to that of CitationCorley, Collins, and Calvin (2011) and CitationCorley (2008), who further justify the value of using WcopyFind.
10. Average presidential approval ratings are 53% in the present sample; the average approval rating for the universe of Gallup poll results for the entire study time frame is 54%.
11. Of these, two Saturday afternoon press conferences were not covered by the news because there was no news broadcast that evening due to PGA golf tournaments.
12. A larger number (47) of the president's opening statements made the news (see Note 7).
13. Even so, 21 of these 33 press conferences were still mentioned on the next day's news.
14. I modeled a number of alternative specifications, including measures for reelection years, the misery index, whether a press conference's opening statement addressed foreign or domestic policy, the effects of lame duck years (last 2 years of a second term), and whether or not a president was Republican. None of these variables produced interesting results or altered any of the coefficients’ significance or direction.
15. Because I did not hypothesize a conditional relationship, I do not present the model. Nevertheless, modeling an interaction between opening statements and solo press conferences reveals that as presidents deliver more solo domestic press conferences, the impact of the opening statement on news coverage declines, by about 8 percentage points.
16. Memo from Marlin Fitzwater to George H. W. Bush, December 20, 1988, John Sununu Files, Office of the Chief of Staff to the President, Formerly Withheld, NLGB Control Number 11317, George Bush Presidential Library, provided by CitationHan (2011).