As the number of U.S. coronavirus cases and deaths keeps surging to new records, how have the news media — particularly partisan media — done at reporting on public health?
We spent the past several months analyzing the coronavirus coverage of the evening lineups of CNN, Fox News and MSNBC, and the nightly national news programs on ABC, CBS and NBC. In a new report, we document the considerable differences in how these networks covered the virus.
Fox News and MSNBC referred to partisans more than health officials
We analyzed news transcripts from Jan. 21, 2020, the day of the first confirmed case of the coronavirus in the United States, through June 12, 2020, right after the country passed 2 million cases. We conducted our research using a combination of manual coding and language analysis software on transcripts drawn from online sources; more technical detail is below.
Fox News and MSNBC coronavirus coverage discussed partisans more than health experts. For each network, 11 percent of the transcript paragraphs mentioned partisan political figures and organizations, while only 8 percent mentioned health organizations, agencies or officials. Fox News more frequently mentioned Democrats than the other networks, while MSNBC more frequently referred to Republicans. In contrast, CNN and the broadcast networks mentioned health officials more often than partisans.
China sent masks, gloves and gowns to many U.S. states. Here’s who benefitted.
The major cable news networks used language that reflected political divides
Fox News covered the virus’s origins, consequences and treatments differently than the other two cable networks. For instance, Fox News used words related to China at a far greater rate than either CNN or MSNBC. The network was 46.7 times as likely to use the phrase “Chinese communist” as CNN and 22.4 times as likely as MSNBC. This usage coincided with President Donald Trump’s admonitions to China for failing to properly warn the world.
Fox News was more likely than the other networks to use such economic terms as “American worker” and “small business,” consistent with Trump’s emphasis on the economic effects of the coronavirus.
When discussing treatments, Fox News used language in line with Trump’s characterization of azithromycin and hydroxychloroquine as “having a real chance to be one of the biggest game changers in the history of medicine.” The network was 1.5 times as likely as CNN and 2.8 times as likely as MSNBC to mention hydroxychloroquine, and 4.1 and 10.4 times as likely to use the word “azithromycin” as CNN or MSNBC, respectively.
Both MSNBC and CNN used phrases corresponding with topics on which liberals believed that Trump had performed poorly. MSNBC was 1.5 times as likely to use the phrase “protective equipment” as CNN and 2.7 times as likely to do so as Fox News. CNN used the word “testing” 2.2 times more often than Fox News and “get test” 1.3 times more than MSNBC.
Another distinction was in how often different networks mentioned the pandemic’s scale. MSNBC described the virus using such words as “crisis” and “bad” more often than did Fox News and CNN. These words were 1.2 to 1.8 times more likely to appear on MSNBC than the other two cable networks.
Does the public rally behind leaders who get covid-19?
Health facts were covered differently across networks
Factual information was presented differently across outlets, as well. We analyzed claims made about wearing masks and using disinfectants or ultraviolet light on the human body to kill the coronavirus.
On April 3, the Centers for Disease Control and Prevention issued guidelines recommending that people wear masks. All the networks covered the need to wear masks more often after the CDC announcement than before.
For this analysis, we found every mention of masks and examined the surrounding five paragraphs to determine whether the networks shared correct, incorrect or mixed information on the matter. The results show a troubling rate of misinformation, particularly from Fox News, where just over one in 10 sampled segments about mask-wearing suggested said it was unnecessary, even after the CDC released its guidelines.
On April 23, Trump made speculative comments about injecting disinfectants and using ultraviolet light inside the body to combat the coronavirus. Several fact-checking organizations concluded that these techniques should not be used.
Following this, Fox News was less likely than the other networks to cover the president’s statements. CNN and MSNBC mentioned the statement in 3.4 times as many segments as Fox News.
When Fox News did cover the topic, it was the least likely to include correct information. MSNBC was also less likely to pair the incorrect information with correct information than were CNN or the broadcast networks.
Cable news wrapped health information in partisan packaging
Popular cable news programs are presenting coronavirus information in ways that map onto party politics. Fox News and MSNBC, in particular, present starkly different views of the virus.
We did not test whether the coverage politicized public perceptions. However, the Pew Research Center found that coronavirus beliefs varied depending on where people turned for news. Our research does find pronounced political differences across the cable news prime-time lineups, with information that does not appear to be aimed at preserving public health.
Professors! Don’t miss TMC’s new classroom topic guides, now online.
Note on research methods
We downloaded transcripts from Nexis Uni and manually coded 44,643 paragraphs for whether they were about covid-19. We then developed a classifier using the DistilBERT model that we applied to the remaining 486,068 transcript paragraphs. We used software from SpaCy to identify the entities mentioned and then manually coded them into categories related to partisanship and health. We identified words and phrases that appeared more often on one network compared with another and then used the Python Gensim library to create clusters of words based on their semantic similarity. We isolated several health-related claims that were false according to fact-checkers, searched the transcripts for words associated with the claims, and manually coded for correct and incorrect/misleading content. All manual coding involved several people to ensure reliability.
Don’t miss any of TMC’s smart analysis! Sign up for our newsletter.
Ceren Budak (@cerenbudak) is an assistant professor in the School of Information at the University of Michigan.
Ashley Muddiman (@ashleymuddiman) is an associate professor in the department of communication studies at the University of Kansas.
Natalie (Talia) Stroud (@TaliaStroud) is a professor and director of the Center for Media Engagement in the Moody College of Communication at the University of Texas at Austin.