A Matter of Language or Culture: Coverage of the 2004 U.S. Elections on Spanish- and English-Language Television

Abstract

This article fills a gap in the communication and political science literature by comparing how Spanish- and English-language television stations cover U.S. elections. A content analysis of more than 400 national network news stories and nearly 3,000 local news stories reveals that local and network Spanish-language stations provide less election coverage than their English-language counterparts. Although Spanish-language stations are more likely to focus election coverage on “Latino” issues or interests, the results indicate only moderate differences in how stations in each language frame their election stories, with stations in both languages concentrating more coverage around campaign strategy and the horse race than substantive issues.

ACKNOWLEDGEMENTS

The authors would like to thank Roberto Suro of the Pew Hispanic Center, Ken Goldstein of University of Wisconsin–Department of Political Science, and Martin Kaplan of the Norman Lear Center at the University of Southern California's Annenberg School for Communication for their generous support of this research. The authors would also like to thank the MCS editor and reviewers for their expert help and guidance in preparing this manuscript.

Notes

¹Throughout this article we use the term Latino as opposed to Hispanic. The term Hispanic carries a potentially negative association as it was created by the U.S. government as a way of defining a population; in contrast Latino is seen as a term originating from the population. In addition, Hispanic is linked directly to colonial Spain, whereas Latino is more broadly focused on Latin America.

²To be clear, we refer to Spanish-language news throughout this article to mean Spanish-language news programs aired and produced in the United States. We are not examining Spanish-language news programming produced in other countries but (via satellite or through a special subscription) available within the United States. DeSipio (2003) found satellite usage to be rare among Latino viewers in his survey sample.

³This view is not universal; see, for example, Bonilla (1998), Fox (1996), and Torres (1998), who argued that the media is in fact following, rather than leading, the move toward a Pan-Latin identity.

⁴The Pew Hispanic Center (2004) study suggests that as much 77% of those watching Spanish-language TV are immigrants, whereas DiSipio (2003) placed the figure at 61%. Even if a sizable number of these viewers have or are in the process of obtaining citizenship, the results still support the point that many viewers of Spanish-language television are not eligible to vote.

⁵To extend the argument we might expect that Spanish-language news would cover more foreign policy, immigration, or economics outside of the campaign context in more substantive way than English-language stations. This, however, is a subject for another article.

Note. Market information from Nielsen Media (see http://www.nielsenmedia.com).

^a Source: U.S. Census Bureau (2000); all numbers based on Metropolitan Statistical Area.

^b Source: Cook Political Report (2004, October 26). The Cook Political Report is a nonpartisan publication that is highly regarded for its independent analysis of election campaigns.

⁶An election story is defined to be about (a) a single/multiple race story or (b) a voting issue/process problem, ballot initiative, or bond issue. See Appendix B for an explanation of all relevant coding definitions.

⁷Although these three markets are obviously not the only local markets with Spanish-language television, they are the three of the largest Latino markets, and as a result, local Spanish-language television is well developed.

⁸The dates were chosen to coincide with the final month of the 2004 election in response to a recommendation by the so-called Gore Commission (see http://www.ntia.doc.gov/pubintadvcom/piacreport.pdf). The choice of 29 days was made to maximize capture of weekday (as opposed to weekend) news programming, which is less likely to preempted by sporting events.

⁹These data were collected as part of a larger project, which captured news in 11 markets: New York, Los Angeles, Philadelphia, Dallas, Seattle, Miami, Denver, Orlando, Tampa, Dayton, and Des Moines.

¹⁰Our decision to focus on broadcast channels as opposed to either cable news channels or satellite programming was driven by the current public policy debates regarding the effects of increased access to the digital spectrum and broadcaster public interest obligation. In addition, although there is evidence that cable news stations (e.g., CNN, MSNBC) may actually have higher ratings than traditional nightly news broadcasts though both rank below local television (Project for Excellence in Journalism, 2004), it is still true that more people have access to over the air broadcasts than cable.

¹¹The basic structure of local newscasts is actually quite similar regardless of language. Local news broadcasts in both languages are generally half-hour programs that air during the early evening (5:00–6:30 p.m.) and at 11:00 p.m. The exception is the hour-long 10:00 p.m. newscast aired by the English-language FOX news in all three markets, which makes up for the lack of early-evening news broadcasts. Our data contain overall percentages and averages, which, given the high capture rate, are unlikely to be significantly or systematically affected by either the small amounts of missing data or differences between news broadcast times. Television news broadcasts are often preempted or replaced by late-running sporting events, particularly on weekends. As a result, the number of broadcasts for each station is based on regular news programs actually aired between 5:00 p.m. and 11:30 p.m., not the number of broadcasts a station would have aired without being preempted or replaced.

¹²In this analysis, we used Scott's Pi for nominal variables and Krippendorff's R for ordinal variables. As there is no standard in the literature, we compared both statistics to other widely used measures and found no notable difference in the results.

¹³See Appendix B for the complete coding definition of this question.

¹⁴Interestingly, 80% of English-language network stories about Latino interests aired on NBC, whose parent company, General Electric, also owns Telemundo.

¹⁵For this section, we excluded non-race-related coverage and “voting process” stories, as they are not related to particular races or candidates. See Appendix B for coding definitions.

¹⁶Mentions of early voting had to be in the context of coverage of a particular office, generally presidential or multiple race coverage, not about the mechanics of voting early, which would be included as voting process stories and therefore excluded from this analysis.

¹⁷As Benoit et al. (2004) indicated, people in nonbattleground states are less informed about issue differences between candidates than people in battleground states. Two of the three local markets studied (Los Angeles and New York City) were not in battleground states.

Note. All but 2 of the 18 local stations captured are owned and operated by their respective parent networks. The exception is in Miami, where the ABC and FOX affiliates are owned by smaller ownership groups. DMA = Designated Market Area.

Additional information

Notes on contributors

Matthew Hale

Matthew Hale (Ph.D., University of Southern California, 2002) is Assistant Professor in the Department of Public & Healthcare Administration at Seton Hall University. His research interests include media coverage of the public and nonprofit sectors and the impact of information technology on nonprofit organizations and local governments.

Tricia Olsen

Tricia Olsen is a Ph.D., Candidate at the University of Wisconsin–Madison. Her research interests include political economy, political communication, and the politics of developing countries.

Erika Franklin Fowler

Erika Franklin Fowler (Ph.D., University of Wisconsin–Madison, 2007) is Robert Wood Johnson Scholar in Health Policy Research at the University of Michigan. Her research interests include political behavior, public opinion, and media.